British Journal of Surgery

Greve, J W M; Rubino, F

doi: 10.1002/bjs.6389pmid: 18844263

Obesity is undeniably a new and severe epidemic. It is the result of many factors acting in concert, including poor dietary habits, reduced physical activity and genetic predisposition. With the rapid increase in obesity comes a pronounced rise in obesity-related metabolic disorders, in particular type 2 diabetes mellitus. Worldwide, more than a billion people are overweight. In several states of the USA, obesity (body mass index (BMI) greater than 30 kg/m2) affects 30 per cent of the population. In the country as a whole, the estimated incidence of morbid obesity (BMI greater than 40 kg/m2) has risen to 5–8 per cent. As there is a close relationship between obesity and the development of type 2 diabetes, it is hardly surprising that there has been a parallel increase in the incidence of diabetes, at present estimated to affect 16–17 million people. In Europe, there are about 48 million diabetic patients, representing 7·5 per cent of the population. Even India, hitherto an undernourished nation, has seen a rapid increase in obesity. As a result, at least in part, an estimated 25 million of its people have diabetes, a total expected to rise to a staggering 80 million by 2030. Weight loss achieved through conservative means improves diabetes but is very difficult to maintain over time. In contrast, there is considerable evidence that gastrointestinal surgery can induce long-term remission of type 2 diabetes. The first reports of diabetes remission after gastrointestinal operations date back to the first half of the 20th century, with the publication of anecdotal, but consistent, observations of improvement after gastric resection for peptic ulcer or gastric cancer1. However, it was bariatric surgery that made this observation common, because a population with a high prevalence of diabetes underwent this type of intervention. In a review of the metabolic effects of bariatric surgery, Buchwald and colleagues2 noted that it resulted in resolution of type 2 diabetes in nearly 80 per cent of patients when all types of procedure were considered together. In general, the more complex the operation (and so the greater the risk), the better the effect on diabetes. The safest procedure, carrying minimal risk, was gastric banding, which resulted in a resolution of diabetes in 47 per cent of patients. Gastric bypass produced resolution in about 83 per cent. The most complex procedure, biliopancreatic diversion, achieved well over 90 per cent success, but at a higher risk of perioperative complications and long-term deficiencies of macro- and micronutrients. It has been suggested that diabetes resolves as a result of the massive weight loss after surgery. This explanation is too logical to refute; yet there is much evidence against it. Pories and co-workers3 have pointed out the almost immediate resolution of diabetes after gastric bypass, ahead of any significant weight loss. Furthermore, bypassing the duodenum and jejunum (duodenal jejunal bypass) in lean animals with diabetes has the same positive effects on blood glucose as in obese animals with diabetes: the improvement is not due to weight loss4. These experimental findings have been confirmed in humans by Cohen and colleagues5, who carried out duodenal jejunal bypass in non-obese patients with diabetes. After about a year, these patients had normal HbA1c levels without significant weight loss. Similar results have been described with a non-surgical intervention in which a tube, the Endobarrier™ (GI Dynamics, Lexington, Massachusetts, USA), was deployed endoscopically within the duodenum to mimic a duodenal jejunal bypass. Diabetes was improved without transecting the bowel6. Other experimental studies support the concept that surgical control of diabetes occurs as a consequence of the rearrangement of gastrointestinal anatomy4. Although the exact mechanism remains a mystery, recent clinical investigations suggest that changes in gastrointestinal hormones may play a dominant role. Other factors, including calorie restriction and fat content of the adipocyte, may also be relevant. Reduction in the fat content of the adipocyte reduces inflammatory mediators and improves adipokines, which affect the regulation of insulin sensitivity7. Other operations have been proposed as even better metabolic solutions. These include resection of the omentum to reduce the intra-abdominal fat mass, which is considered to be crucial in the development of metabolic disorders. The interposition of an ileal segment at a proximal site in the jejunum may induce an early gut hormone release. Even combinations of these procedures, including duodenal jejunal bypass, have been suggested8. It is too early to speculate on the long-term effects of such surgery, but preliminary results are encouraging. A compelling feature of bariatric surgery in the morbidly obese is a significant improvement in long-term survival9,10. The Swedish obese subject study, which has a follow-up of more than 10 years, involves mainly restrictive procedures10. Despite the limited effect of restriction on diabetes, the improved survival in patients who have surgery appears principally to be a result of resolution of existing diabetes and prevention of new disease. A similar improvement in long-term survival has been recognized in the USA in a study by Adams and colleagues9, who employed gastric bypass procedures. As a result of the now substantial evidence that surgery improves metabolic disorders, leading societies, such as the International Federation for the Surgery of Obesity and the American Society of Bariatric Surgeons, are paying more attention to the metabolic aspects of bariatric intervention. They have even changed their names to the International Federation for the Surgery of Obesity and Metabolic Disorders and the American Society for Metabolic and Bariatric Surgery respectively. Moreover, in 2007 an International Consensus Conference (Diabetes Surgery Summit) was held in Rome, where it was recognized that gastrointestinal bypass procedures improved diabetes by mechanisms beyond simple weight loss, and a multidisciplinary effort was called for to prioritize research in this emerging field of medicine. A whole new field is developing from bariatric surgery. Manipulation of the gut and, in particular, the foregut appears to be able to reverse abnormalities that cause disturbed energy homeostasis and metabolic disorder. Type 2 diabetes mellitus, among other diseases, can be managed surgically. Further research is needed, however, to evaluate the place of the various procedures that have been proposed. Although it is too early to know the correct indications for surgery in this setting, there is no doubt that the future looks brighter for patients with diabetes. References 1 Rubino F . Is type 2 diabetes an operable intestinal disease? A provocative yet reasonable hypothesis . Diabetes Care 2008 ; 31 ( Suppl 2 ): S290 – S296 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Buchwald H , Avidor Y, Braunwald E, Jensen MD, Pories W, Fahrbach K et al. Bariatric surgery: a systematic review and meta-analysis . JAMA 2004 ; 292 : 1724 – 1737 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Pories WJ , Swason MS, MacDonald KG, Long SB, Morris PG, Brown BM et al. Who would have thought it? An operation proves to be the most effective therapy for adult-onset diabetes mellitus . Ann Surg 1995 ; 222 : 339 – 350 ; discussion 350–352. Google Scholar Crossref Search ADS PubMed WorldCat 4 Rubino F , Forgione A, Cummings DE, Vix M, Gnuli D, Mingrone G et al. The mechanism of diabetes control after gastrointestinal bypass surgery reveals a role of the proximal small intestine in the pathophysiology of type 2 diabetes . Ann Surg 2006 ; 244 : 741 – 749 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Cohen RV , Schiavon CA, Pinheiro JS, Correa JL, Rubino F. Duodenal-jejunal bypass for the treatment of type 2 diabetes in patients with body mass index of 22–34 kg/m2: a report of 2 cases . Surg Obes Relat Dis 2007 ; 3 : 195 – 197 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Rodriguez-Grunert L , Galvao Neto MP, Alamo M, Ramos AC, Baez PB, Tarnoff M et al. First human experience with endoscopically delivered and retrieved duodenal-jejunal bypass sleeve . Surg Obes Relat Dis 2008 ; 4 : 55 – 59 . Google Scholar Crossref Search ADS PubMed WorldCat 7 van Dielen FM , Buurman WA, Hadfoune M, Nijhuis J, Greve JW. Macrophage inhibitory factor, plasminogen activator inhibitor-1, other acute phase proteins, and inflammatory mediators normalize as a result of weight loss in morbidly obese subjects treated with gastric restrictive surgery . J Clin Endocrinol Metab 2004 ; 89 : 4062 – 4068 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Santoro S , Malzoni CE, Velhote MC, Milleo FQ, Santo MA, Klajner S et al. Digestive Adaptation with Intestinal Reserve: a neuroendocrine-based operation for morbid obesity . Obes Surg 2006 ; 16 : 1371 – 1379 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Adams TD , Gress RE, Smith SC, Halverson RC, Simper SC, Rosamond WD et al. Long-term mortality after gastric bypass surgery . N Engl J Med 2007 ; 357 : 753 – 761 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Sjostrom L , Narbro K, Sjostrom CD, Karason K, Larsson B, Wedel H et al. Effects of bariatric surgery on mortality in Swedish obese subjects . N Engl J Med 2007 ; 357 : 741 – 752 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

Gough, M J; on behalf of the GALA Trial Collaborators

doi: 10.1002/bjs.6416pmid: 18844247

The European Carotid Surgery Trial (ECST) and the North American Symptomatic Carotid Endarterectomy Trial (NASCET) have confirmed that carotid endarterectomy (CEA) reduces stroke risk in symptomatic patients. The Asymptomatic Carotid Surgery Trial has also demonstrated a lesser benefit for asymptomatic carotid stenosis. Today, measures to improve the safety and efficacy of carotid surgery have become the focus of investigation. Anaesthesia, monitoring for hypoperfusion and emboli, carotid shunting, patch angioplasty and quality control measures have all been examined, often in small randomized or non-randomized studies and in systematic reviews. One such review1 suggesting better outcomes for local anaesthetic (LA) surgery coincided with publication of data suggesting that LA, rather than general anaesthesia (GA) preserved cerebral autoregulation during CEA2. The senior authors of these publications subsequently conceived the GALA Trial, the results of which have now been presented. GALA is the largest randomized surgical/anaesthetic trial ever performed. It recruited 3526 patients from 95 centres in 24 countries. Although a power calculation suggested that 5000 patients might be required to confirm a one-third reduction (from 7·5 to 5 per cent) in primary outcome events (stroke, myocardial infarction, death) within 30 days of operation, this was a more conservative result than that suggested by Rerkasem et al3. Recruitment terminated after a 1 year extension. One reason for the lower than expected total, particularly in the UK (which contributed 46 per cent of the patients), was that many surgeons developed a preference for LA early on. This was due to the ease of determining the need for a shunt and the continuous intra-operative monitoring of adverse events that LA allows. Furthermore, immediate awareness of neurological status following revascularization is welcomed by surgeons. Indeed, many surgeons in the UK have already decided that LA is the optimum anaesthetic technique for CEA. Are they right? Surprisingly, we do not know for certain. While GALA has not shown the superiority of one technique over the other, neither has it shown equivalence. Outcome events occurred between randomization and 30 days after operation in 84 of 1752 patients (4·8 per cent) having GA and in 80 of 1771 patients (4·5 per cent) having LA. This difference is not significant, with three (95 per cent confidence interval (c.i.): − 11, + 17) events prevented per 1000 patients given LA. The situation was similar when each primary outcome event was considered on its own: stroke—70 (4·0 per cent) GA versus 66 (3·7 per cent) LA, three (−10, + 16) prevented per 1000; death—26 (1·5 per cent) GA versus 19 (1·1 per cent) LA, four (−3, + 12) prevented per 1000; fatal and non-fatal myocardial infarction—nine LA (0·5 per cent) versus four GA (0·2 per cent), three (−2, + 8) more per 1000. If 5000 patients had been recruited with the same proportion of outcome events, there would still be no real difference (95 per cent c.i.: 15 fewer to 9 more events per 1000 patients given LA). Although LA does not seem to offer an advantage over GA, there are some confounding variables. One of these is the frequency with which hypertension was pharmacologically induced in patients having GA: 43 per cent versus 17 per cent for LA (P < 0·001). McCleary et al.2 showed that hypertension after clamping was the mechanism by which autoregulation occurred during LA surgery. Anaesthetists appear to have mimicked this effect in many patients randomized to GA. Further analysis should determine if this influenced the trial outcome. One reason why CEA under LA may prove unsafe is a belief that surgery should be completed quickly to avoid patient distress. Although the mean duration of surgery and the frequency of cranial nerve injuries were similar between the groups, patch angioplasty was used less often under LA (42 versus 50 per cent, P < 0·001). One review4 has estimated that patching prevents up to 30 ipsilateral peri-operative strokes per 1000 operations, and this difference might have adversely affected patients treated under LA. Still, this is unlikely to have masked a true difference and, furthermore, there was no apparent advantage to patch angioplasty in the GALA trial as a whole. A previous meta-analysis5 concluded that there was insufficient evidence to advise either routine or selective shunt use during CEA. Methods for detecting post-clamp cerebral ischaemia also appeared of little value in supporting selective shunt use. GALA data for awake neurological testing and analysis of relevant outcome events may update these views. Although GALA does not show a benefit in favour of LA, some findings are worthy of comment. Patients with contralateral carotid occlusion, who formed 9 per cent of the total, had fewer events with LA (8 of 160, 5 per cent) than with GA (15 of 150, 10 per cent). Contralateral strokes occurred more often with GA (7 of 150, 4·7 per cent) than with LA (2 of 160, 1·2 per cent) (P for interaction 0·098). These data should be interpreted with great caution but may reflect preservation of autoregulation under LA. At 1 year after surgery there were fewer events in the LA group (P = 0·094). There is no obvious reason for this and it may be a spurious finding. LA was also associated with fewer peri-operative deaths and this is consistent with data from other trials3. When data from the Cochrane systematic review are combined with the GALA outcomes, the impact of LA ranges from a 64 per cent reduction to a 7 per cent increase in peri-operative death risk. Surprisingly, none of the secondary end-points (quality of life (QOL), high dependency and intensive care use, overall hospital stay) differed between the groups. It is possible that QOL was assessed too late to detect an early difference, and that length of stay was determined by existing protocols rather than individual patient requirements. It could be argued that a more strictly standardized protocol, particularly for blood pressure manipulation and patching, might have increased the chances of demonstrating a difference between LA and GA. However, such a prescriptive protocol would have compromised recruitment and the chosen, more pragmatic, approach should improve the generalizability of the results. In conclusion, GALA shows that the outcome of CEA has improved since the publication of ECST and NASCET, which makes it increasingly difficult to demonstrate a difference between LA and GA. For the present, units undertaking CEA should ideally be competent in both techniques, as patient preference or medical factors might render one method more suitable than the other for an individual. CEA under LA has proven to be safe, and the cost-effective, reliable and continuous intra-operative cerebral monitoring that is afforded by LA will attract many. The less frequent need for shunting with LA is also worthy of note and it is tempting to suggest that it might confer a benefit in patients with a contralateral carotid occlusion. References 1 Tangkanakul C , Counsell CE, Warlow CP. Local versus general anaesthesia in carotid endarterectomy: a systematic review of the evidence . Eur J Vasc Endovasc Surg 1997 ; 13 : 491 – 499 . Google Scholar Crossref Search ADS PubMed WorldCat 2 McCleary AJ , Watson A, Whiteley S, Dickson DE, Gough MJ. The differing effects of regional and general anaesthetic on cerebral metabolism during carotid endarterectomy . Eur J Vasc Endovasc Surg 1996 ; 12 : 173 – 181 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Rerkasem K , Bond R, Rothwell PM. Local versus general anaesthesia for carotid endarterectomy . Cochrane Database of Systematic Reviews 2004 ; 2 :Art. No.: CD000126. DOI: 10.1002/14651858.CD000126 . Google Scholar OpenURL Placeholder Text WorldCat Crossref 4 Bond R , Rerkasem K, AbuRahma AF, Naylor AR, Rothwell PM. Patch angioplasty versus primary closure for carotid endarterectomy . Cochrane Database of Systematic Reviews 2004 ; 2 :Art. No.: CD000160. DOI: 10.1002/14651858.CD000160.pub2 . Google Scholar OpenURL Placeholder Text WorldCat Crossref 5 Bond R , Rerkasem K, Rothwell PM. Routine or selective carotid artery shunting for carotid endarterectomy (and different methods of monitoring in selective shunting) . Cochrane Database of Systematic Reviews 2002 ; 2 :Art. No.: CD000190. DOI: 10.1002/14651858.CD000190 . Google Scholar OpenURL Placeholder Text WorldCat Crossref Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

Dickinson, K J; Troxler, M; Homer-Vanniasinkam, S

doi: 10.1002/bjs.6359pmid: 18844273

Abstract Background Disordered coagulation complicates many diseases and their treatments, often predisposing to haemorrhage. Conversely, patients with cardiovascular disease who demonstrate antiplatelet resistance may be at increased thromboembolic risk. Prompt identification of these patients facilitates optimization of haemostatic dysfunction. Point-of-care (POC) tests are performed ‘near patient’ to provide a rapid assessment of haemostasis and platelet function. Methods This article reviews situations in which POC tests may guide surgical practice. Their limitations and potential developments are discussed. The paper is based on a Medline and PubMed search for English language articles on POC haemostasis and platelet function testing in surgical practice. Results POC tests identifying perioperative bleeding tendency are already widely used in cardiovascular and hepatic surgery. They are associated with reduced blood loss and transfusion requirements. POC tests to identify thrombotic predisposition are able to determine antiplatelet resistance, predicting thromboembolic risk. So far, however, these tests remain research tools. Conclusion POC haemostasis testing is a growing field in surgical practice. Such testing can be correlated with improved clinical outcome. Introduction Assessment of haemostasis and platelet function is important in surgical practice. Haemostatic dysfunction may complicate numerous surgical pathologies and their management, including major trauma, cardiopulmonary bypass, liver transplantation and ruptured aortic aneurysm. In the assessment of bleeding tendencies, results of routine laboratory investigations provide only a snapshot of the clinical situation at the moment the sample was taken; furthermore, such results are often available only after a considerable delay. Point-of-care (POC) testing involves assays performed outside the laboratory by non-specialist personnel1. With regard to haemostasis, these tests provide a global assessment of thrombosis or information specific to platelet function or both. The major advantages over laboratory testing are that POC-generated results are produced more rapidly and frequently, so providing more contemporaneous data to guide patient care. For example, platelet aggregometry2,3 performed in the laboratory involves the measurement of changes in light transmission through platelet-rich plasma after the addition of an agonist. This technique has been described as the ‘gold standard’ of platelet function testing, but it takes time and trained laboratory personnel, and it is impossible to perform at the bedside. In contrast, a POC test such as the activated clotting time (ACT) requires no specialist staff, takes minutes to complete and may be performed on the ward or in the operating room. Surgical patients may also have thrombotic tendencies and POC tests have been developed to assess platelet function, including response to antiplatelet therapy. Such testing aims to classify patients by their reaction to aspirin or clopidogrel as ‘responders’, ‘non-responders’ (may exhibit antiplatelet resistance) or ‘hyper-responders’ (may be at increased risk of surgical bleeding while receiving antiplatelet medication). It is then possible to tailor medication according to a patient's therapeutic response. This article reviews the clinical situations in which POC testing of haemostasis and platelet function guides surgical practice. It describes the bedside tests that are available. Test limitations and precautions required for data interpretation are also discussed, together with possible future developments. Methods A search of Medline and PubMed was performed with manual searching of bibliographies for relevant key references. The following search terms were used: ‘(point of care or point-of-care) and surgery’, ‘surgery and (haemostasis or platelet or thrombosis or bleeding)’, ‘(thromboelastography or TEG® or thromboelastometry) and surgery’, ‘(VerifyNow® or Ultegra® Rapid Platelet Function Analyzer or Ultegra® RPFA) and surgery’, ‘(Platelet Function Analyzer-100 or PFA-100®) and surgery’, ‘(activated clotting time or ACT) and surgery’, ‘Sonoclot® and surgery’, ‘(Clot Signature Analyzer® or CSA) and surgery’, ‘trauma and (point of care or point-of care)’. Results Tests to identify bleeding tendency: high-risk patients Although POC haemostasis and platelet testing may aid care of any critically ill surgical patient, it is likely to be particularly helpful in certain groups vulnerable to haemorrhagic complications. These include patients undergoing hepatic, cardiac or arterial surgery and victims of polytrauma. Perioperative POC testing of haemostasis or platelet function would be invaluable if it could identify patients at increased risk of postoperative haemorrhage. This might guide the preoperative cessation of antiplatelet medication as well as the perioperative administration of heparin, protamine or blood products, optimizing haemostasis and possibly reducing complications. Hepatic surgery Liver tumours often develop against a background of parenchymal liver disease predisposing to portal hypertension and coagulopathy. The proportion of patients who require transfusion for liver surgery depends on the complexity of the resection and values of 40 per cent are common4. Although rates as low as 20 per cent have been reported, the median transfusion requirement with extended left hepatectomy is still 1·5 units4. Strategies to reduce blood loss and transfusion requirements may be employed. These include manual compression of the hepatic artery and portal vein (Pringle's manoeuvre), in situ hypothermic perfusion to allow hypothermic cellular protection and ex vivo resection, and the administration of recombinant factor VIIa (rVIIa)4–6. POC haemostasis testing should prove beneficial in this situation if it could detect preoperative coagulopathy. This would allow correction with appropriate blood factor transfusion, aiming to decrease subsequent intraoperative blood loss. In a similar fashion, intraoperative haemostasis monitoring might guide optimal use of expensive products, such as rVIIa. Cardiovascular surgery Patients undergoing major vascular surgery often have heparin to prevent thrombosis of stagnant blood. The use of heparin in elective open abdominal aortic aneurysm repair has been demonstrated not to affect the risk of bleeding or thromboembolic complications, but myocardial infarction (MI) is rendered less likely, presumably owing to reduced coronary thrombosis7. Perioperative MI occurred in 1·4 per cent of patients with heparin and 5·7 per cent of those not receiving heparin7. If heparin is used, protamine may be administered to reverse its effect, but this may cause severe non-immunological or immunological reactions8. Non-immunological reactions are thromboxane- and histamine-mediated and can be controlled by slowing the administration of protamine and giving vasopressors. Immunological reactions are unpredictable and can be difficult to manage. A selective policy with respect to protamine use is required to minimize adverse effects. Open thoracoabdominal aortic aneurysm repair is associated with significant haemostatic dysfunction, partly resulting from systemic heparinization, mild permissive hypothermia (32–34 °C) and left-heart bypass with centrifugal pump9. Consequently, significant haemorrhagic complications are common and the need for reoperation for bleeding in large series is 2·5–5·1 per cent10,11. Similarly, the use of cardiopulmonary bypass during coronary artery bypass grafting (CABG) has diverse and complex effects on haemostasis (Table 1)12–20. Following CABG, postoperative bleeding is relatively common, with 20 per cent of patients having perioperative haemostatic abnormalities, requiring reintervention in 2–6 per cent21,22. This has important clinical implications as re-exploration is associated with a three to fourfold increase in mortality23. Table 1 Effects of cardiopulmonary bypass on haemostasis Feature of cardiopulmonary bypass . Effect on haemostasis . Contact with artificial surface of bypass circuit/air bubbles Decreased coagulation factors (primarily because of factor XII activation) Thrombin generation Activation of fibrinolysis Platelet dysfunction (decreased ADP and collagen-induced aggregation) Haemodilution (extracorporeal circuit primed with crystalloid solution) Decreased coagulation factors Thrombocytopenia, platelet dysfunction Activation of fibrinolysis Reperfusion of pericardial blood Activation of extrinsic coagulation Hypothermia Altered platelet function (e.g. aggregation) Heparin administration Binds antithrombin and increases binding affinity for factor Xa and thrombin Risk of heparin-induced thrombocytopenia type II Protamine administration Inhibits coagulation and platelet aggregation in high doses, with a paradoxical increased risk of bleeding Feature of cardiopulmonary bypass . Effect on haemostasis . Contact with artificial surface of bypass circuit/air bubbles Decreased coagulation factors (primarily because of factor XII activation) Thrombin generation Activation of fibrinolysis Platelet dysfunction (decreased ADP and collagen-induced aggregation) Haemodilution (extracorporeal circuit primed with crystalloid solution) Decreased coagulation factors Thrombocytopenia, platelet dysfunction Activation of fibrinolysis Reperfusion of pericardial blood Activation of extrinsic coagulation Hypothermia Altered platelet function (e.g. aggregation) Heparin administration Binds antithrombin and increases binding affinity for factor Xa and thrombin Risk of heparin-induced thrombocytopenia type II Protamine administration Inhibits coagulation and platelet aggregation in high doses, with a paradoxical increased risk of bleeding ADP, adenosine diphosphate. Open in new tab Table 1 Effects of cardiopulmonary bypass on haemostasis Feature of cardiopulmonary bypass . Effect on haemostasis . Contact with artificial surface of bypass circuit/air bubbles Decreased coagulation factors (primarily because of factor XII activation) Thrombin generation Activation of fibrinolysis Platelet dysfunction (decreased ADP and collagen-induced aggregation) Haemodilution (extracorporeal circuit primed with crystalloid solution) Decreased coagulation factors Thrombocytopenia, platelet dysfunction Activation of fibrinolysis Reperfusion of pericardial blood Activation of extrinsic coagulation Hypothermia Altered platelet function (e.g. aggregation) Heparin administration Binds antithrombin and increases binding affinity for factor Xa and thrombin Risk of heparin-induced thrombocytopenia type II Protamine administration Inhibits coagulation and platelet aggregation in high doses, with a paradoxical increased risk of bleeding Feature of cardiopulmonary bypass . Effect on haemostasis . Contact with artificial surface of bypass circuit/air bubbles Decreased coagulation factors (primarily because of factor XII activation) Thrombin generation Activation of fibrinolysis Platelet dysfunction (decreased ADP and collagen-induced aggregation) Haemodilution (extracorporeal circuit primed with crystalloid solution) Decreased coagulation factors Thrombocytopenia, platelet dysfunction Activation of fibrinolysis Reperfusion of pericardial blood Activation of extrinsic coagulation Hypothermia Altered platelet function (e.g. aggregation) Heparin administration Binds antithrombin and increases binding affinity for factor Xa and thrombin Risk of heparin-induced thrombocytopenia type II Protamine administration Inhibits coagulation and platelet aggregation in high doses, with a paradoxical increased risk of bleeding ADP, adenosine diphosphate. Open in new tab Traditional laboratory testing of haemostasis and platelet function, such as platelet aggregometry, is too complex and impractical for use in the management of bleeding after cardiac or arterial surgery. Platelet count, fibrinogen level and international normalized ratio are much more accessible haematological tests; these, unfortunately, correlate poorly with aberrant platelet function and are relatively insensitive for the prediction of postoperative bleeding24. Antiplatelet agents Although effective in preventing thrombotic complications, antiplatelet agents may promote bleeding during surgery. Most vascular surgical patients should be prescribed antiplatelet medication, but this results in increased blood loss and transfusion requirements if they are not discontinued before operation25. Patients undergoing CABG while receiving clopidogrel have an 85 per cent increase in platelet transfusion, owing to increased chest tube drainage, compared with those not receiving the drug26. However, if clopidogrel is withheld, the chance of ischaemic complications such as MI increases; drug-eluting coronary stents are particularly at risk. Continuation of dual antiplatelet therapy (aspirin plus clopidogrel) is associated with the lowest risk of in-stent thrombosis and is generally advised, accepting that bleeding complications will be increased. However, the balance of risk versus benefit must be carefully weighed for each patient27. Massive blood transfusion and trauma Massive transfusion (defined as the replacement of one total blood volume in 24 h28) often occurs after trauma or during emergency surgery for bleeding. Tissue perfusion and oxygenation must be restored by the replacement of blood volume, but it is important to correct any coagulopathy, which may occur for a number of reasons. These include hypothermia (exposure and administration of non-warmed blood products), citrate-mediated hypocalcaemia, dilution of platelets and coagulation factors with transfusion of packed red blood cells, and disseminated intravascular coagulation with further consumption of platelets and coagulation factors29. Geeraedts and colleagues30 suggest that blood loss after trauma cannot be estimated as accurately as during elective surgery. The patient may be subject to ‘blind’ massive transfusion, often deficient in fresh frozen plasma and platelets. POC testing could allow real-time monitoring of haemostasis during such resuscitation to guide blood product use, which might avoid under- or overtransfusion and the associated complications. Tests to identify bleeding tendency: the assays Activated clotting time Activated clotting time (ACT) is a measurement of whole blood clotting potential first described by Hattersley in 1966 (Table 2). ACT allows a quantified measurement of heparin activity, for example during cardiovascular surgery, and so the appropriate and judicious administration of protamine. ACT use has resulted in a reduction in chest drainage, operation length and the need for fresh frozen plasma, platelets and red blood cells in patients undergoing cardiac surgery11. This has the potential to reduce the risk of infection, immunosuppression and other complications of blood component therapy11, 35–37. In addition, the ACT is reproducible with a variability well below 10 per cent38. Table 2 Point-of-care tests to assess bleeding tendency in surgical patients Test . Methodology . Parameter assessed . Surgical use . Activated clotting time Kaolin/celite added to blood accelerates clotting by contact activation. Global assessment of haemostasis Widely available and commonly used to monitor heparin treatment during vascular and particularly cardiac surgery when high doses of heparin are administered (300 units/kg compared with ∼70 units/kg typical of abdominal aortic aneurysm repair)4 Subsequent clot formation is detected by inhibition of plunger movement in a well31 Detection limit is 0·5 unit/ml heparin, so it is suitable for monitoring high-dose heparin treatment, unlike activated partial thromboplastin time41 HemoSTATUS® (platelet-activated clotting test) Uses platelet-activating factor to shorten the kaolin-activated clotting time in whole blood32 Platelet responsiveness and whole blood procoagulant activity Potential for use in cardiovascular surgery Thromboelastography (TEG®) Dynamic measurement of physical properties of blood clot Coagulation and fibrinolysis Cardiac surgery: European Association for Cardio-Thoracic surgery guidelines state that TEG® can be used to guide postoperative transfusion, but further study is required before it can be recommended as the standard care for postoperative transfusion management Measurement of clot formation in an oscillating cylindrical cup at 37 °C. The rotation of a pin suspended within the blood is measured and converted to an electrical signal which is then amplified and charted33. A graph is produced, which represents clot formation and lysis and can be interpreted to determine speed and strength of clot formation (Fig. 1) Speed of clot formation and strength of this clot Liver transplant surgery Trauma Sonoclot® analyser Detects the change in impedance to movement of a vibrating probe within a developing clot. Like TEG®, it produces a characteristic graph of clot formation and fibrinolysis Coagulation and fibrinolysis Limited: not commonly used in cardiac surgery Clot Signature Analyzer® Platelet function is assessed by the time required for a platelet thrombus to occlude two small holes in tubing through which the blood flows. This is the platelet-mediated haemostasis time. A second flow system measures the platelet interaction with collagen under conditions of high shear. The time to occlusion of this pathway is the collagen-induced thrombus formation time Whole blood haemostasis and platelet function Not in routine clinical use Hemostasis Analysis System Measures the platelet ‘contractile force’ (representing the force generated during platelet-mediated clot retraction) to assess platelet cytoplasmic protein function as well as adhesive protein expression Platelet function and clot strength Potential use in cardiac surgery with cardiopulmonary bypass ICHOR/Plateletworks® Number of single platelets in a treated sample (ADP- or collagen-) and a baseline control sample are compared. Platelet aggregation is measured by the loss of single platelets. Percentage aggregation is equivalent to the percentage reduction from the baseline single platelet count to that after the addition of an agonist34 Platelet aggregatory function Potential to predict postoperative bleeding in those taking antiplatelet medication Test . Methodology . Parameter assessed . Surgical use . Activated clotting time Kaolin/celite added to blood accelerates clotting by contact activation. Global assessment of haemostasis Widely available and commonly used to monitor heparin treatment during vascular and particularly cardiac surgery when high doses of heparin are administered (300 units/kg compared with ∼70 units/kg typical of abdominal aortic aneurysm repair)4 Subsequent clot formation is detected by inhibition of plunger movement in a well31 Detection limit is 0·5 unit/ml heparin, so it is suitable for monitoring high-dose heparin treatment, unlike activated partial thromboplastin time41 HemoSTATUS® (platelet-activated clotting test) Uses platelet-activating factor to shorten the kaolin-activated clotting time in whole blood32 Platelet responsiveness and whole blood procoagulant activity Potential for use in cardiovascular surgery Thromboelastography (TEG®) Dynamic measurement of physical properties of blood clot Coagulation and fibrinolysis Cardiac surgery: European Association for Cardio-Thoracic surgery guidelines state that TEG® can be used to guide postoperative transfusion, but further study is required before it can be recommended as the standard care for postoperative transfusion management Measurement of clot formation in an oscillating cylindrical cup at 37 °C. The rotation of a pin suspended within the blood is measured and converted to an electrical signal which is then amplified and charted33. A graph is produced, which represents clot formation and lysis and can be interpreted to determine speed and strength of clot formation (Fig. 1) Speed of clot formation and strength of this clot Liver transplant surgery Trauma Sonoclot® analyser Detects the change in impedance to movement of a vibrating probe within a developing clot. Like TEG®, it produces a characteristic graph of clot formation and fibrinolysis Coagulation and fibrinolysis Limited: not commonly used in cardiac surgery Clot Signature Analyzer® Platelet function is assessed by the time required for a platelet thrombus to occlude two small holes in tubing through which the blood flows. This is the platelet-mediated haemostasis time. A second flow system measures the platelet interaction with collagen under conditions of high shear. The time to occlusion of this pathway is the collagen-induced thrombus formation time Whole blood haemostasis and platelet function Not in routine clinical use Hemostasis Analysis System Measures the platelet ‘contractile force’ (representing the force generated during platelet-mediated clot retraction) to assess platelet cytoplasmic protein function as well as adhesive protein expression Platelet function and clot strength Potential use in cardiac surgery with cardiopulmonary bypass ICHOR/Plateletworks® Number of single platelets in a treated sample (ADP- or collagen-) and a baseline control sample are compared. Platelet aggregation is measured by the loss of single platelets. Percentage aggregation is equivalent to the percentage reduction from the baseline single platelet count to that after the addition of an agonist34 Platelet aggregatory function Potential to predict postoperative bleeding in those taking antiplatelet medication ADP, adenosine diphosphate. Open in new tab Table 2 Point-of-care tests to assess bleeding tendency in surgical patients Test . Methodology . Parameter assessed . Surgical use . Activated clotting time Kaolin/celite added to blood accelerates clotting by contact activation. Global assessment of haemostasis Widely available and commonly used to monitor heparin treatment during vascular and particularly cardiac surgery when high doses of heparin are administered (300 units/kg compared with ∼70 units/kg typical of abdominal aortic aneurysm repair)4 Subsequent clot formation is detected by inhibition of plunger movement in a well31 Detection limit is 0·5 unit/ml heparin, so it is suitable for monitoring high-dose heparin treatment, unlike activated partial thromboplastin time41 HemoSTATUS® (platelet-activated clotting test) Uses platelet-activating factor to shorten the kaolin-activated clotting time in whole blood32 Platelet responsiveness and whole blood procoagulant activity Potential for use in cardiovascular surgery Thromboelastography (TEG®) Dynamic measurement of physical properties of blood clot Coagulation and fibrinolysis Cardiac surgery: European Association for Cardio-Thoracic surgery guidelines state that TEG® can be used to guide postoperative transfusion, but further study is required before it can be recommended as the standard care for postoperative transfusion management Measurement of clot formation in an oscillating cylindrical cup at 37 °C. The rotation of a pin suspended within the blood is measured and converted to an electrical signal which is then amplified and charted33. A graph is produced, which represents clot formation and lysis and can be interpreted to determine speed and strength of clot formation (Fig. 1) Speed of clot formation and strength of this clot Liver transplant surgery Trauma Sonoclot® analyser Detects the change in impedance to movement of a vibrating probe within a developing clot. Like TEG®, it produces a characteristic graph of clot formation and fibrinolysis Coagulation and fibrinolysis Limited: not commonly used in cardiac surgery Clot Signature Analyzer® Platelet function is assessed by the time required for a platelet thrombus to occlude two small holes in tubing through which the blood flows. This is the platelet-mediated haemostasis time. A second flow system measures the platelet interaction with collagen under conditions of high shear. The time to occlusion of this pathway is the collagen-induced thrombus formation time Whole blood haemostasis and platelet function Not in routine clinical use Hemostasis Analysis System Measures the platelet ‘contractile force’ (representing the force generated during platelet-mediated clot retraction) to assess platelet cytoplasmic protein function as well as adhesive protein expression Platelet function and clot strength Potential use in cardiac surgery with cardiopulmonary bypass ICHOR/Plateletworks® Number of single platelets in a treated sample (ADP- or collagen-) and a baseline control sample are compared. Platelet aggregation is measured by the loss of single platelets. Percentage aggregation is equivalent to the percentage reduction from the baseline single platelet count to that after the addition of an agonist34 Platelet aggregatory function Potential to predict postoperative bleeding in those taking antiplatelet medication Test . Methodology . Parameter assessed . Surgical use . Activated clotting time Kaolin/celite added to blood accelerates clotting by contact activation. Global assessment of haemostasis Widely available and commonly used to monitor heparin treatment during vascular and particularly cardiac surgery when high doses of heparin are administered (300 units/kg compared with ∼70 units/kg typical of abdominal aortic aneurysm repair)4 Subsequent clot formation is detected by inhibition of plunger movement in a well31 Detection limit is 0·5 unit/ml heparin, so it is suitable for monitoring high-dose heparin treatment, unlike activated partial thromboplastin time41 HemoSTATUS® (platelet-activated clotting test) Uses platelet-activating factor to shorten the kaolin-activated clotting time in whole blood32 Platelet responsiveness and whole blood procoagulant activity Potential for use in cardiovascular surgery Thromboelastography (TEG®) Dynamic measurement of physical properties of blood clot Coagulation and fibrinolysis Cardiac surgery: European Association for Cardio-Thoracic surgery guidelines state that TEG® can be used to guide postoperative transfusion, but further study is required before it can be recommended as the standard care for postoperative transfusion management Measurement of clot formation in an oscillating cylindrical cup at 37 °C. The rotation of a pin suspended within the blood is measured and converted to an electrical signal which is then amplified and charted33. A graph is produced, which represents clot formation and lysis and can be interpreted to determine speed and strength of clot formation (Fig. 1) Speed of clot formation and strength of this clot Liver transplant surgery Trauma Sonoclot® analyser Detects the change in impedance to movement of a vibrating probe within a developing clot. Like TEG®, it produces a characteristic graph of clot formation and fibrinolysis Coagulation and fibrinolysis Limited: not commonly used in cardiac surgery Clot Signature Analyzer® Platelet function is assessed by the time required for a platelet thrombus to occlude two small holes in tubing through which the blood flows. This is the platelet-mediated haemostasis time. A second flow system measures the platelet interaction with collagen under conditions of high shear. The time to occlusion of this pathway is the collagen-induced thrombus formation time Whole blood haemostasis and platelet function Not in routine clinical use Hemostasis Analysis System Measures the platelet ‘contractile force’ (representing the force generated during platelet-mediated clot retraction) to assess platelet cytoplasmic protein function as well as adhesive protein expression Platelet function and clot strength Potential use in cardiac surgery with cardiopulmonary bypass ICHOR/Plateletworks® Number of single platelets in a treated sample (ADP- or collagen-) and a baseline control sample are compared. Platelet aggregation is measured by the loss of single platelets. Percentage aggregation is equivalent to the percentage reduction from the baseline single platelet count to that after the addition of an agonist34 Platelet aggregatory function Potential to predict postoperative bleeding in those taking antiplatelet medication ADP, adenosine diphosphate. Open in new tab The HemoSTATUS® (Medtronic Blood Management, Parker, Colorado, USA) test is a modification of the ACT, designed to assess platelet function as well as whole blood procoagulant activity (Table 2). Transient platelet dysfunction is the most important defect in haemostasis following cardiopulmonary bypass and cannot be adequately detected by the ACT alone. Thrombocytopenia during bypass can be treated with desmopressin, which stimulates factor VIII production from endothelial cells. Patients who have had cardiac surgery and have received desmopressin or platelet transfusion have significantly improved clot ratios measured by this assay40. This shortening of ACT has been attributed to the indirect or direct platelet activating factor procoagulant action on platelets, and the HemoSTATUS® assay may therefore be useful in detecting patients who would benefit from platelet or desmopressin administration39,40. Despite this theoretical advantage, the HemoSTATUS® test has been shown to correlate poorly with total blood loss and transfusion requirements after CABG with cardiopulmonary bypass42–44. ACT does have limitations. It lacks sensitivity at lower heparin concentrations, but this has relatively little clinical relevance as heparin reversal is more likely to be beneficial when serum concentrations are high (over 1 unit/ml). In addition, although ACT is useful in detecting persistent heparin, it may not be sensitive enough to detect ‘heparin rebound’44 because of its relatively high threshold. This ‘rebound’ effect may contribute to the microvascular bleeding seen after bypass and its associated morbidity. One solution is to give heparinase to degrade heparin to smaller, inactive fragments; this improves the detection of residual low heparin concentrations by ACT. Although ACT is a POC assay in common clinical use, it provides only a global assessment of haemostasis and cannot, without modification, distinguish between the causes of coagulopathy. During cardiopulmonary bypass, in addition to the effect of heparin, the ACT can become significantly prolonged by haemodilution and hypothermia31. This observation has led to development of an individualized heparin management system using the Hepcon® device (Medtronic, Minneapolis, Minnesota, USA)45–47. This measures heparin concentration during bypass and calculates protamine doses based on residual heparin levels. This avoids the use of ‘best guess’ doses of protamine and appears to be associated with reduced haemostatic activation, postoperative blood loss and blood product requirement32,48. However, Slight and co-workers49 have recently suggested that the heparin management system is no more efficacious than ACT as a guide to heparin administration. Viscoelastic methods These POC tests are also global tests of haemostasis that assess fibrinolysis and coagulation (Table 2; Fig. 1). Thromboelastography (TEG®) has classically been used in patients having cardiac surgery and a number of studies support its use. Microvascular bleeding, attributed to transient platelet dysfunction, is the primary haemostatic defect with cardiopulmonary bypass. Mongan and colleagues50 found TEG® to be predictive of risk of increased postoperative bleeding. Other studies have specified the maximum amplitude of the trace (Fig. 1) as a predictor of both postoperative bleeding51,52 and blood product use53. In addition, TEG®-directed desmopressin therapy13 and fresh frozen plasma transfusion54 have led to reduced blood transfusion requirements. Reported benefits are clinically significant, with one study reporting a 2-unit reduction in transfusion requirements with TEG® and a decline in surgical revision from 5·7 to 1·5 per cent33. The benefits of TEG® are not restricted to cardiopulmonary bypass; it may also predict bleeding and graft thrombosis complicating off-pump CABG21. However, TEG® has a lower predictive accuracy in patients having revisional cardiac surgery for recurrent disease55. Fig. 1 Open in new tabDownload slide A typical thromboelastography (TEG®) trace. It measures the time to fibrin formation within a clot (R time), determined by clotting factors and inhibitor balance; the clot formation rate, given by the angle between the initial slope of the TEG® and the horizontal (α angle), which represents speed of clot strengthening by fibrin cross-linkage; and the strength of the clot, determined by the maximum amplitude (MA), which is the greatest amplitude of the TEG® trace and depends on the integrity of platelet function and fibrinogen binding The advantages of haemostasis assessment with TEG® extend beyond cardiac surgery. TEG® is also used widely during liver surgery, particularly transplantation56,57. As noted above, severe coagulopathy can occur following a number of contributing factors, which can be challenging to manage. Transfusion algorithms guided by TEG® results are associated with decreased red cell transfusions58. A further application of TEG® is in monitoring rVIIa. Hendriks and colleagues59 suggested that rFVIIa affected the physical properties of clot and its speed of formation which, they argued, could not be reliably detected by routine coagulation tests, but were demonstrated by TEG®. In a similar vein, Rugeri and co-workers60 suggested that TEG® could rapidly detect in vivo coagulation changes after trauma and might help guide transfusion. They reported good correlation between TEG®, standard coagulation parameters and platelet counts. In a lethal haemorrhage animal model, Kheirabadi and colleagues61 demonstrated TEG® to be a better indicator of coagulopathic bleeding and mortality than prothrombin time measurement. However, further clinical studies are required to establish whether TEG® use alters patient outcome. One limitation of TEG® is that the measurement time can last from 30 to 45 min and requires technical expertise. Clearly in an emergency the haemostasis profile may have changed by the time these ‘snapshot’ results are available. Furthermore, a wide variability in sample results when processed in the first 30 min after collection has also been observed62. The main limitation of TEG® is that a baseline trace is desirable for comparison, as the range of ‘normal’ values is wide. In addition, although POC tests such as the prothrombin time and activated partial thromboplastin time have been validated for algorithm use, this is not the case with TEG®63. The Sonoclot® Analyzer (Sienco, Denver, Colorado, USA), another viscoelastic POC haemostasis test64, is less widely used than TEG® (Table 2). It has the advantage of being subject to less intraoperator variability than TEG® and is faster to perform. Tuman and co-workers65 have shown the Sonoclot® Analyzer to be better than standard coagulation tests at predicting the risk of postoperative bleeding after cardiopulmonary bypass, with a sensitivity of 74 per cent. Yamada and colleagues66 demonstrated that a modified Sonoclot® assay could predict postoperative haemorrhage by comparing results before and after bypass. Clot Signature Analyzer® The Clot Signature Analyzer® (CSA) assay (Xylum, Scarsdale, New York, USA) measures global haemostasis as well as platelet function (Table 2). Although this POC test may be more physiological by using conditions of flow, it is not in widespread clinical use. Despite this, Faraday and colleagues67 have shown that the CSA can detect a clinical coagulopathic state after cardiopulmonary bypass and that this is independently predictive of the need for transfusion. Hemostasis Analysis System The Hemostasis Analysis System (Hemodyne, Richmond, Virginia, USA) measures the platelet ‘contractile force’ and can be used to monitor bleeding tendency after cardiopulmonary bypass (Table 2). During bypass, platelet surface adhesive ligands such as glycoproteins Ib alpha (CD42b) and IIIa (CD61) are shed68,69. The consequent reduction in contractile force measured by this device has been demonstrated to be related to early postoperative blood loss70. Criticisms of this assay include a lack of flow conditions and its high concentration of heparin (1 unit/ml). In addition, heparinase I is added after blood sampling, as the test is sensitive to residual heparin. All these factors mean that it is uncertain whether the results obtained are a true reflection of the situation in vivo. ICHOR/Plateletworks® The American College of Cardiology and the American Heart Association recommend the cessation of aspirin 7–10 days before coronary artery bypass surgery to reduce the risk of blood loss71. More recently, the European Association for Cardiothoracic Surgery has suggested discontinuing aspirin for 2–10 days and clopidogrel for 5–7 days before surgery, if the patient's condition allows72. Craft and colleagues34 demonstrated a good correlation between platelet aggregometry and the ICHOR/Plateletworks® (PW) assay (Helena Laboratories, Beaumont, Texas, USA) to measure recovery from clopidogrel; subsequently, the PW assay was used after bypass to try to predict increased risk of postoperative bleeding from preoperative aspirin ingestion73. This study found that, compared with platelet aggregometry, the PW assay was less able to identify patients who had recently ingested aspirin and less able to predict postoperative mediastinal blood loss. The PW assay has also been shown to overestimate recovery from clopidogrel74. Tests to identify thrombotic tendency: high-risk patients Antiplatelet therapy Platelets play an important role in atherothrombosis and hence the pathogenesis of cardiovascular diseases75. The Antiplatelet Trialists Collaboration found that aspirin administration in high-risk patients reduced stroke, MI and vascular death by 22 per cent76. More recent evidence suggests that not all patients respond in a uniform manner to antiplatelet agents, with some being non-responders (resistant) and others hyper-responders77,78. The concept of aspirin or clopidogrel resistance is controversial but can be defined as antiplatelet medication ingestion with failure to achieve the expected level of platelet inhibition79,80. The mechanism for this is uncertain and much debated. Possible explanations include reduced drug bioavailability (for example, poor absorption), altered platelet function (for example, increased platelet sensitivity to agonists such as adenosine diphosphate (ADP) and collagen) and abnormal platelet interactions with other blood cells (for example, endothelium and leucocytes)75. The label of ‘resistance’ is debated; some authors consider differences between individuals merely to reflect natural biological variations in the response to aspirin or clopidogrel and platelet physiology together with multiple factors that influence atherothrombosis. Aspirin resistance is said to occur in between 5 and 40 per cent of patients depending on the population studied; a recent systematic review gives a figure of 1 in 481,82. Although the debate over the existence of drug resistance continues, there are without doubt patients who benefit less from antiplatelet monotherapy. Neither the optimal management of a non-responder, nor which platelet function test is best for demonstrating ‘resistance’, has yet been resolved82. Patients labelled ‘aspirin resistant’ have platelets that are unusually sensitive to ADP, and the use of antiplatelet agents that inhibit ADP has been suggested78. However, the indiscriminate use of combination antiplatelet medication cannot be advocated, as bleeding risk may be increased without any reduction in atherothrombotic events83,84. POC platelet function testing would revolutionize pharmacotherapy if it could accurately identify antiplatelet resistance and allow stratification of atherothrombotic risk. This would allow antiplatelet therapy to be tailored to the individual. Tests to identify thrombotic tendency: the assays Thromboelastography The Thromboelastograph® (TEG®) system (Haemoscope, Niles, Illinois, USA) has recently been modified to allow monitoring of antiplatelet therapy (mTEG®) (Table 3; Fig. 1)85. To achieve this, the maximum amplitude of the standard TEG® is compared with that from the mTEG®. Agarwal and colleagues86 have suggested a use for mTEG® in monitoring the individual response to antiplatelet agents. Bliden and co-workers87 used it to predict postprocedural ischaemic events in patients taking clopidogrel undergoing percutaneous coronary intervention (PCI). Hobson and colleagues88 have similarly used mTEG® to calculate the ‘percentage clotting inhibition’ (calculated as the area under the clotting response curve at 15 min) in an attempt to determine individual response to antiplatelet agents. Although this POC test is not in routine use, it has the potential to detect antiplatelet resistance and possibly to predict postoperative bleeding complications in hyper-responders. This test has also been used to detect hypercoagulability after liver transplantation89 and major abdominal surgery90 that could not have been detected by routine laboratory investigation. Table 3 Point-of-care tests to assess thrombotic tendency in surgical patients Test . Methodology . Parameter assessed . Surgical use . Modified thromboelastography (mTEG®) Within the TEG® assay, thrombin is the strongest platelet agonist, and determines the MA. If the effects of thrombin are ameliorated then other agonists, e.g. ADP or AA are ‘unmasked’ and their effects can be monitored Platelet function Not in routine use Heparin inhibits thrombin and addition of this to the blood, as well as reptilase plus factor XIIIa, forms a fibrin network and the platelets are able to interact independently of thrombin. With the addition of AA or ADP the MA increases to near normal, but not in patients with adequate aspirin or clopidogrel response65. To monitor antiplatelet therapy, the MA of the standard TEG® is compared with that from the mTEG® Potential to detect non-responders to aspirin and clopidogrel and tailor individual antiplatelet regimens Potential to detect hyper-responders to predict patients at increased risk of bleeding while on antiplatelets Platelet function analyser device PFA-100® Whole blood flows at high shear rate (5000-6000/sec) through an aperture (147 µm in diameter) within a capillary membrane. This membrane is coated with either CEPI or CADP. Platelets in the flowing blood form a plug at the aperture and the time taken for this to happen is the CT (Fig. 2) Platelet function and whole blood haemostasis International Society on Thrombosis and Haemostasis Subcommittee of the Scientific and Standardization Committee has recommended further study before routine clinical use Maximum CT is 300 sec; some controversy exists regarding reference ranges, but these are defined as 85–65 sec for CEPI and 64–114 sec with CADP.43,71 CT is affected by vWF levels, haematocrit, platelet count, blood group (blood group O is associated with lower vWF levels and longer CT)72 and leucocyte count Ultegra® Rapid Platelet Function Analyzer and VerifyNow® aspirin and clopidogrel assays The Ultegra® Rapid Platelet Function Analyzer detects activated platelets binding to fibrinogen. Whole blood is added to a lyophilized preparation of human fibrinogen-coated beads and a platelet agonist such as TRAP. With activation, platelet agglutination to the beads occurs with a consequent increase in light transmission, detected by a light absorbance meter. The change in light absorbance over time is expressed as platelet aggregation units Platelet function. With test modifications, this can detect individual response to aspirin or clopidogrel Algorithms exist for use during vascular/coronary interventional radiology; however, these tests are not in routine clinical practice VerifyNow® P2Y12: ADP in whole blood stimulates platelet aggregation via both P2Y1 and P2Y12 receptors. Prostaglandin E1 is added to the assay to suppress free intracellular calcium to reduce the activation caused by ADP binding to the P2Y1 receptor. A second channel in the assay uses the agonist isoTRAP to give a baseline platelet function (isoTRAP-induced aggregation occurs independently of P2Y12 receptors). Changes in light transmission with platelet agglutination are measured. The results of this assay are expressed in P2Y12 reaction units VerifyNow® aspirin: the platelet agonist used is AA and results are expressed as aspirin reaction units Test . Methodology . Parameter assessed . Surgical use . Modified thromboelastography (mTEG®) Within the TEG® assay, thrombin is the strongest platelet agonist, and determines the MA. If the effects of thrombin are ameliorated then other agonists, e.g. ADP or AA are ‘unmasked’ and their effects can be monitored Platelet function Not in routine use Heparin inhibits thrombin and addition of this to the blood, as well as reptilase plus factor XIIIa, forms a fibrin network and the platelets are able to interact independently of thrombin. With the addition of AA or ADP the MA increases to near normal, but not in patients with adequate aspirin or clopidogrel response65. To monitor antiplatelet therapy, the MA of the standard TEG® is compared with that from the mTEG® Potential to detect non-responders to aspirin and clopidogrel and tailor individual antiplatelet regimens Potential to detect hyper-responders to predict patients at increased risk of bleeding while on antiplatelets Platelet function analyser device PFA-100® Whole blood flows at high shear rate (5000-6000/sec) through an aperture (147 µm in diameter) within a capillary membrane. This membrane is coated with either CEPI or CADP. Platelets in the flowing blood form a plug at the aperture and the time taken for this to happen is the CT (Fig. 2) Platelet function and whole blood haemostasis International Society on Thrombosis and Haemostasis Subcommittee of the Scientific and Standardization Committee has recommended further study before routine clinical use Maximum CT is 300 sec; some controversy exists regarding reference ranges, but these are defined as 85–65 sec for CEPI and 64–114 sec with CADP.43,71 CT is affected by vWF levels, haematocrit, platelet count, blood group (blood group O is associated with lower vWF levels and longer CT)72 and leucocyte count Ultegra® Rapid Platelet Function Analyzer and VerifyNow® aspirin and clopidogrel assays The Ultegra® Rapid Platelet Function Analyzer detects activated platelets binding to fibrinogen. Whole blood is added to a lyophilized preparation of human fibrinogen-coated beads and a platelet agonist such as TRAP. With activation, platelet agglutination to the beads occurs with a consequent increase in light transmission, detected by a light absorbance meter. The change in light absorbance over time is expressed as platelet aggregation units Platelet function. With test modifications, this can detect individual response to aspirin or clopidogrel Algorithms exist for use during vascular/coronary interventional radiology; however, these tests are not in routine clinical practice VerifyNow® P2Y12: ADP in whole blood stimulates platelet aggregation via both P2Y1 and P2Y12 receptors. Prostaglandin E1 is added to the assay to suppress free intracellular calcium to reduce the activation caused by ADP binding to the P2Y1 receptor. A second channel in the assay uses the agonist isoTRAP to give a baseline platelet function (isoTRAP-induced aggregation occurs independently of P2Y12 receptors). Changes in light transmission with platelet agglutination are measured. The results of this assay are expressed in P2Y12 reaction units VerifyNow® aspirin: the platelet agonist used is AA and results are expressed as aspirin reaction units ADP, adenosine diphosphate; MA, maximum amplitude; AA, arachidonic acid; CEPI, collagen and epinephrine; CADP, collagen and adenosine diphosphate; CT, closure time; vWF, von Willebrand factor; TRAP, thrombin receptor activating peptide. Open in new tab Table 3 Point-of-care tests to assess thrombotic tendency in surgical patients Test . Methodology . Parameter assessed . Surgical use . Modified thromboelastography (mTEG®) Within the TEG® assay, thrombin is the strongest platelet agonist, and determines the MA. If the effects of thrombin are ameliorated then other agonists, e.g. ADP or AA are ‘unmasked’ and their effects can be monitored Platelet function Not in routine use Heparin inhibits thrombin and addition of this to the blood, as well as reptilase plus factor XIIIa, forms a fibrin network and the platelets are able to interact independently of thrombin. With the addition of AA or ADP the MA increases to near normal, but not in patients with adequate aspirin or clopidogrel response65. To monitor antiplatelet therapy, the MA of the standard TEG® is compared with that from the mTEG® Potential to detect non-responders to aspirin and clopidogrel and tailor individual antiplatelet regimens Potential to detect hyper-responders to predict patients at increased risk of bleeding while on antiplatelets Platelet function analyser device PFA-100® Whole blood flows at high shear rate (5000-6000/sec) through an aperture (147 µm in diameter) within a capillary membrane. This membrane is coated with either CEPI or CADP. Platelets in the flowing blood form a plug at the aperture and the time taken for this to happen is the CT (Fig. 2) Platelet function and whole blood haemostasis International Society on Thrombosis and Haemostasis Subcommittee of the Scientific and Standardization Committee has recommended further study before routine clinical use Maximum CT is 300 sec; some controversy exists regarding reference ranges, but these are defined as 85–65 sec for CEPI and 64–114 sec with CADP.43,71 CT is affected by vWF levels, haematocrit, platelet count, blood group (blood group O is associated with lower vWF levels and longer CT)72 and leucocyte count Ultegra® Rapid Platelet Function Analyzer and VerifyNow® aspirin and clopidogrel assays The Ultegra® Rapid Platelet Function Analyzer detects activated platelets binding to fibrinogen. Whole blood is added to a lyophilized preparation of human fibrinogen-coated beads and a platelet agonist such as TRAP. With activation, platelet agglutination to the beads occurs with a consequent increase in light transmission, detected by a light absorbance meter. The change in light absorbance over time is expressed as platelet aggregation units Platelet function. With test modifications, this can detect individual response to aspirin or clopidogrel Algorithms exist for use during vascular/coronary interventional radiology; however, these tests are not in routine clinical practice VerifyNow® P2Y12: ADP in whole blood stimulates platelet aggregation via both P2Y1 and P2Y12 receptors. Prostaglandin E1 is added to the assay to suppress free intracellular calcium to reduce the activation caused by ADP binding to the P2Y1 receptor. A second channel in the assay uses the agonist isoTRAP to give a baseline platelet function (isoTRAP-induced aggregation occurs independently of P2Y12 receptors). Changes in light transmission with platelet agglutination are measured. The results of this assay are expressed in P2Y12 reaction units VerifyNow® aspirin: the platelet agonist used is AA and results are expressed as aspirin reaction units Test . Methodology . Parameter assessed . Surgical use . Modified thromboelastography (mTEG®) Within the TEG® assay, thrombin is the strongest platelet agonist, and determines the MA. If the effects of thrombin are ameliorated then other agonists, e.g. ADP or AA are ‘unmasked’ and their effects can be monitored Platelet function Not in routine use Heparin inhibits thrombin and addition of this to the blood, as well as reptilase plus factor XIIIa, forms a fibrin network and the platelets are able to interact independently of thrombin. With the addition of AA or ADP the MA increases to near normal, but not in patients with adequate aspirin or clopidogrel response65. To monitor antiplatelet therapy, the MA of the standard TEG® is compared with that from the mTEG® Potential to detect non-responders to aspirin and clopidogrel and tailor individual antiplatelet regimens Potential to detect hyper-responders to predict patients at increased risk of bleeding while on antiplatelets Platelet function analyser device PFA-100® Whole blood flows at high shear rate (5000-6000/sec) through an aperture (147 µm in diameter) within a capillary membrane. This membrane is coated with either CEPI or CADP. Platelets in the flowing blood form a plug at the aperture and the time taken for this to happen is the CT (Fig. 2) Platelet function and whole blood haemostasis International Society on Thrombosis and Haemostasis Subcommittee of the Scientific and Standardization Committee has recommended further study before routine clinical use Maximum CT is 300 sec; some controversy exists regarding reference ranges, but these are defined as 85–65 sec for CEPI and 64–114 sec with CADP.43,71 CT is affected by vWF levels, haematocrit, platelet count, blood group (blood group O is associated with lower vWF levels and longer CT)72 and leucocyte count Ultegra® Rapid Platelet Function Analyzer and VerifyNow® aspirin and clopidogrel assays The Ultegra® Rapid Platelet Function Analyzer detects activated platelets binding to fibrinogen. Whole blood is added to a lyophilized preparation of human fibrinogen-coated beads and a platelet agonist such as TRAP. With activation, platelet agglutination to the beads occurs with a consequent increase in light transmission, detected by a light absorbance meter. The change in light absorbance over time is expressed as platelet aggregation units Platelet function. With test modifications, this can detect individual response to aspirin or clopidogrel Algorithms exist for use during vascular/coronary interventional radiology; however, these tests are not in routine clinical practice VerifyNow® P2Y12: ADP in whole blood stimulates platelet aggregation via both P2Y1 and P2Y12 receptors. Prostaglandin E1 is added to the assay to suppress free intracellular calcium to reduce the activation caused by ADP binding to the P2Y1 receptor. A second channel in the assay uses the agonist isoTRAP to give a baseline platelet function (isoTRAP-induced aggregation occurs independently of P2Y12 receptors). Changes in light transmission with platelet agglutination are measured. The results of this assay are expressed in P2Y12 reaction units VerifyNow® aspirin: the platelet agonist used is AA and results are expressed as aspirin reaction units ADP, adenosine diphosphate; MA, maximum amplitude; AA, arachidonic acid; CEPI, collagen and epinephrine; CADP, collagen and adenosine diphosphate; CT, closure time; vWF, von Willebrand factor; TRAP, thrombin receptor activating peptide. Open in new tab Platelet function analysis The platelet function analyser device PFA-100® (Dade Behring, Marburg, Germany) was designed to be more physiological and to simulate primary haemostasis in vitro (Table 3; Fig. 2). The PFA-100® has been used to monitor the effects of aspirin therapy in patients with cardiovascular disease as aspirin selectively prolongs the collagen and epinephrine closure time (CEPI-CT). Roller and colleagues91 concluded that 40 per cent of patients with peripheral arterial disease had an inadequate response to aspirin using the PFA-100®, although they did not relate this to clinical outcome. Ziegler and co-workers92 showed a correlation between ‘non response’ to clopidogrel and increased restenosis in a small study of 98 patients undergoing angioplasty for peripheral arterial disease. In patients having cardiac surgery, shorter PFA-100® closure times were related to the development of postoperative myocardial ischaemia93. Fig. 2 Open in new tabDownload slide The platelet function analyser PFA-100® assay uses conditions of blood flow and a membrane impregnated with collagen and epinephrine or collagen and adenosine diphosphate (ADP) to stimulate vascular injury. The closure time is the time taken for a platelet plug to occlude the membrane For patients with coronary artery disease, there is better evidence correlating PFA-100® results with clinical outcome than for other vascular beds. In those having PCI, a CEPI-CT of less than 190 s measured within the first 24 h after surgery predicts death or MI94. In addition, a recent meta-analysis of eight prospective studies involving 1227 patients used PFA-100® closure time to predict cardiovascular events in aspirin-treated patients with cardiovascular disease95. Data from these studies demonstrated an odds ratio of 2·1 for recurrence of an ischaemic event in non-responders compared with responders. This suggests that POC testing can stratify patients based on their aspirin response status in relation to clinical outcome. However, these studies include patients with recent ischaemic events and their results cannot be extrapolated to those with stable coronary artery disease. Limitations of the PFA-100® include the potential overestimation of the prevalence of aspirin resistance96–100. Wide variability in duplicate testing has also been noted and may exceed 10 per cent101. Levels of von Willebrand factor antigen levels markedly affect closure times and are raised as part of the acute-phase response or following PCI99. This may render the PFA-100® less reliable after an acute thrombotic event or endovascular procedure. Mueller and co-workers96 have demonstrated that the PFA-100® is relatively insensitive to the variability of clopidogrel response or in determining clopidogrel resistance. This may be because of the high concentration of ADP in the assay24,102. Importantly, there is a paucity of literature defining at what closure times bleeding is likely to occur or the safe threshold for intervention24. Because of the inconsistencies in the literature, recommendations against routine use of the PFA-100® have been made103,104. Ultegra® and VerifyNow® systems The VerifyNow® aspirin and clopidogrel assays (Accumetrics, San Diego, California, USA), were initially developed as the Ultegra® Rapid Platelet Function Analyzer (RPFA) to monitor the effects of glycoprotein IIb IIIa inhibitors. The original assay is based on detection of activated platelets binding to fibrinogen with corresponding changes in light absorbance (Table 3, Fig. 3). However, the clinical application of this test has yielded conflicting results. In a large multicentre study, Wheeler and colleagues105 demonstrated that the accuracy and precision of the Ultegra® RPFA assay was equivalent to platelet aggregometry and to a receptor-binding assay with 125I-abciximab. Conversely, Hochholzer and co-workers106 showed that, in patients with stable coronary artery disease awaiting PCI, the correlation between Ultegra® RPFA results and aggregometry was poor. However, the GOLD study107 found that, in patients undergoing PCI and receiving glycoprotein IIb or IIIa receptors, the level of platelet inhibition as measured by the Ultegra® RPFA predicted future major adverse cardiac events but were not correlated with another platelet function assay. Fig. 3 Open in new tabDownload slide The VerifyNow® P2Y12 assay uses adenosine diphosphate (ADP) as an agonist to activate platelets via the P2Y12 and P2Y1 receptors. To ensure specificity for clopidogrel (a P2Y12 antagonist) prostaglandin E1 (PGE1) is used to attenuate P2Y1-mediated effects. As the platelets are activated and aggregate with the fibrinogen-coated beads, increased light absorbance is measured and expressed as P2Y12 reaction units. For example, patients responding to clopidogrel have lower light absorbance than those not taking the drug because of decreased platelet aggregate formation This assay has been modified to monitor the effects of aspirin and clopidogrel (VerifyNow® aspirin assay, VerifyNow® P2Y12 or clopidogrel assay; Table 3). The VerifyNow® P2Y12 assay results correlated well with ADP-induced platelet aggregation as assessed by aggregometry108 and have been used to determine clopidogrel response in clinical trials109. The potential role of these assays in surgical practice lies in the monitoring of a patient's response to antiplatelet medication. For example, therapy for outpatients or patients immediately after surgery could be tailored to the individual depending on the results. Chen and colleagues110 have recently used the VerifyNow® aspirin assay to predict risk of adverse clinical outcomes in patients with stable coronary artery disease. One limitation of the RPFA assay is the absence of flow conditions, which renders it less physiological than, for example, the PFA-100®. Harrison and co-workers111 have also shown a lack of consistency over time for this assay in the identification of aspirin-resistant individuals. Before these assays (especially the newer aspirin and clopidogrel assays) progress from being research tools to guiding clinical decisions, large prospective studies are required112,113. Discussion POC tests such as ACT and TEG® to assess haemostatic function in surgical patients are well established in clinical practice. There is good evidence that these tests can guide transfusion algorithms and also predict which patients are at increased risk of postoperative bleeding. Current shortage of blood products and risks associated with transfusion ensure that early identification of haemostatic abnormalities using these tests is desirable. POC tests have also been shown to guide rVIIa and desmopressin administration, although the evidence base is less impressive. For identifying surgical patients with cardiovascular disease who are at increased risk of thromboembolic complications, POC tests appear less robust. The PFA-100® arguably offers the most physiological POC assessment of platelet function and so antiplatelet response in vitro. However, it cannot be recommended for clinical use because of conflicting reports in the literature. More work is needed on the correlation of these tests with surgical outcomes, for instance restenosis following peripheral angioplasty or graft thrombosis in patients with peripheral arterial disease. Correlation between non-responders and increased risk of complications is also required. Studies in patients with coronary artery disease are numerous, and many show some correlation with outcome in stable disease. However, few have assessed patients with peripheral arterial disease. Caution is needed over the blind extrapolation of data from patients with coronary artery disease to those with peripheral vessel atheroma. References 1 Stief TW , Fareed J. Point of care: diagnostics in hemostasis—the wrong direction? Clin Appl Thromb Haemost 2003 ; 9 : 191 – 195 . Google Scholar Crossref Search ADS WorldCat 2 Harrison P . Progress in the assessment of platelet function . Br J Haematol 2000 ; 111 : 733 – 744 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 3 Born GVR . Aggregation of blood platelets by adenosine and its reversal . Nature 1962 ; 194 : 927 – 929 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Lodge JPA . Hemostasis in liver resection surgery . Semin Hematol 2004 ; 41 ( Suppl 1 ): 70 – 75 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Belghiti J , Noun R, Zante E, Ballet T, Sauvanet A. Portal triad clamping or hepatic vascular exclusion for major liver resection. A controlled study . Ann Surg 1996 ; 224 : 155 – 161 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Pringle JH . Notes on the arrest of hepatic hemorrhage due to trauma . Ann Surg 1908 ; 48 : 541 – 549 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Thompson JF , Mullee MA, Bell PR, Campbell WB, Chant AD, Darke SG et al. Intraoperative heparinisation, blood loss and myocardial infarction during aortic aneurysm surgery: a Joint Vascular Research Group study . Eur J Vasc Endovasc Surg 1996 ; 12 : 86 – 90 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Byun Y , Singh VK, Yang VC. Low molecular weight protamine: a potential non-toxic heparin antagonist . Thromb Res 1999 ; 94 : 53 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Coselli JS , Bozinovski J, LeMaire SA. Open surgical repair of 2286 thoracoabdominal aortic aneursysms . Ann Thorac Surg 2007 ; 83 : S862 – S864 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Conrad MF , Crawford RS, Davison JK, Cambria RP. Thoracoabdominal aneurysm repair: a 20-year perspective . Ann Thorac Surg 2007 ; 83 : S856 – S861 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Despotis GJ , Santoro SA, Spitznagel E, Kater KM, Cox JL, Barnes P et al. Prospective evaluation and clinical utility of on-site monitoring of coagulation in patients undergoing cardiac operation . J Thorac Cardiovasc Surg 1994 ; 107 : 271 – 279 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Khuri SF , Wolfe JA, Josa M, Axford TC, Szymanski I, Assousa S et al. Hematologic changes during and after cardiopulmonary bypass and their relationship to the bleeding time and nonsurgical blood loss . J Thorac Cardiovasc Surg 1992 ; 104 : 94 – 107 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Despotis GJ , Joist JH, Goodnough LT. Monitoring of hemostasis in cardiac surgical patients: impact of point-of-care testing on blood loss and transfusion outcomes . Clin Chem 1997 ; 43 : 1684 – 1696 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Hirsh J . Heparin . N Engl J Med 1991 ; 324 : 1565 – 1574 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 15 Khuri SF , Valeri CR, Loscalzo J, Weinstein MJ, Birjiniuk V, Healey NA et al. Heparin causes platelet dysfunction and induces fibrinolysis before cardiopulmonary bypass . Ann Thorac Surg 1995 ; 60 : 1008 – 1014 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Holloway DS , Summaria L, Sandesara J, Vagher JP, Alexander JC, Caprini JA. Decreased platelet number and function and increased fibrinolysis contributing to postoperative bleeding in cardiopulmonary bypass patients . Thromb Haemost 1988 ; 59 : 62 – 67 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 17 Weerasinghe A , Taylor KM. The platelet in cardiopulmonary bypass . Ann Thorac Surg 1998 ; 66 : 2145 – 2152 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Harker LA , Malpass TW, Branson HE, Hessel EA, Slichter SJ. Mechanism of abnormal bleeding in patients undergoing cardiopulmonary bypass: acquired transient platelet dysfunction associated with selective α-granule release . Blood 1980 ; 56 : 824 – 834 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 19 Chung JH , Gikakis N, Rao AK, Drake TA, Colman RW, Edmunds LH. Pericardial blood activates the extrinsic coagulation pathway during clinical cardiopulmonary bypass . Circulation 1996 ; 93 : 2014 – 2018 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Aldea GS , Soltow LO, Chandler WL, Triggs CM, Vocelka CR, Crockett GI et al. Limitation of thrombin generation, platelet activation and inflammation by elimination of cardiotomy suction in patients undergoing coronary artery bypass grafting treated with heparin-bonded circuits . J Thorac Cardiovasc Surg 2002 ; 123 : 742 – 755 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Hartmann M , Sucker C, Boehm O, Koch A, Loer S, Zacharowski K. Effects of cardiac surgery on hemostasis . Transfus Med Rev 2006 ; 20 : 230 – 241 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Hett DA , Walker D, Pilkington SN, Smith DC. Sonoclot analysis . Br J Anaesth 1995 ; 75 : 771 – 776 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Unsworth-White MJ , Herriot A, Valencia O, Poloniecki J, Smith EE, Murday AJ et al. Resternotomy for bleeding after cardiac operation: a marker for increased morbidity and mortality . Ann Thorac Surg 1995 ; 59 : 664 – 667 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Poston R , Gu J, Manchio J, Lee A, Brown J, Gammie J et al. Platelet function tests predict bleeding and thrombotic events after off-pump coronary bypass grafting . Eur J Cardiothorac Surg 2005 ; 27 : 584 – 591 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Agarwal S , Coakley M, Reddy K, Riddell A, Mallett S. Quantifying the effect of antiplatelet therapy: a comparison of the platelet function analyser (PFA-100®) and modified thromboelastography (mTEG®) with light transmission platelet aggregometry . Anesthesiology 2006 ; 105 : 676 – 683 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Tanaka KA , Szlam F, Kelly AB, Vega JD, Levy JH. Clopidogrel (Plavix) and cardiac surgical patients: implications for platelet function monitoring and postoperative bleeding . Platelets 2004 ; 15 : 325 – 332 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Brilakis ES , Banerjee S, Berger PB. Perioperative management of patients with coronary stents . J Am Coll Cardiol 2007 ; 49 : 2145 – 2150 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Stainsby D , MacLennan S, Thomas D, Isaac J, Hamilton PJ. British Committee for Standards in Haematology. Guidelines on the management of massive blood loss . Br J Haematol 2006 ; 135 : 634 – 641 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 29 Erber WN . Massive blood transfusion in the elective surgical setting . Transfus Apher Sci 2002 ; 27 : 83 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Geeraedts LM Jr , Demiral H, Schaap N, Kamphuisen PW, Pompe JC, Frolke JP. ‘Blind’ transfusion of blood products in exsanguinating trauma patients . Resuscitation 2007 ; 73 : 382 – 388 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Despotis GJ , Summerfield AL, Joist JH, Goodnough LT, Santoro SA, Spitznagel E et al. Comparison of activated coagulation time and whole blood heparin measurements with laboratory plasma anti-Xa heparin concentration in patients having cardiac operations . J Thorac Cardiovasc Surg 1994 ; 108 : 1076 – 1082 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Shirota K , Watanabe T, Takagi Y, Ohara Y, Usui A, Yasuura K. Maintenance of blood heparin concentration rather than activated clotting time better preserves the coagulation system in hypothermic cardiopulmonary bypass . Artif Organs 2000 ; 24 : 49 – 56 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Spiess BD , Gillies BS, Chandler W, Verrier E. Changes in transfusion therapy and reexploration rate after institution of a blood management program in cardiac surgical patients . J Cardiothorac Vasc Anesth 1995 ; 9 : 168 – 173 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Craft RM , Chavez JJ, Snider CC, Muenchen RA, Carroll RC. Comparison of modified Thromboelastograph and Plateletworks whole blood assays to optical platelet aggregation for monitoring reversal of clopidogrel inhibition in elective surgical patients . J Lab Clin Med 2005 ; 145 : 309 – 315 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Slater NPG . The risk of infection transfer by blood transfusion . Curr Pract Surg 1993 ; 5 : 122 – 126 . Google Scholar OpenURL Placeholder Text WorldCat 36 Parrot NR . Blood transfusion: should we avoid it in our cancer patients? Curr Pract Surg 1993 ; 5 : 127 – 131 . Google Scholar OpenURL Placeholder Text WorldCat 37 Jensen LS , Anderson AJ, Christensen PM, Hokland P, Juhl CO, Madsen G et al. Postoperative infection and natural killer cell function following blood transfusion in patients undergoing elective colorectal surgery . Br J Surg 1992 ; 79 : 513 – 516 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Zucker ML , Jobes C, Siegel M, Jobes D, LaDuca FM. Activated clotting time (ACT) testing: analysis of reproductibility . J Extra Corpor Technol 1999 ; 31 : 130 – 134 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 39 Despotis GJ , Levine V, Filos K, Santoro SA, Joist JH, Spitznagel E et al. Evaluation of a new point-of-care test that measures PAF-mediated acceleration of coagulation in cardiac surgical patients . Anesthesiology 1996 ; 85 : 1311 – 1323 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Despotis GJ , Levine V, Saleem R, Spitznagel E, Joist JH. Use of point-of-care identification of patients who can benefit from desmopressin during cardiac surgery: a randomised controlled trial . Lancet 1999 ; 354 : 106 – 110 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Ereth MH , Nuttall GA, Santrach PJ, Klindworth JT, Oliver WC Jr, Schaff HV. The relation between the platelet-activated clotting test (HemoSTATUS) and blood loss after cardiopulmonary bypass . Anesthesiology 1998 ; 88 : 962 – 969 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Forestier F , Coiffic A, Mouton C, Ekouevi D, Chene G, Janvier G. Platelet function point-of-care tests in post-bypass cardiac surgery: are they relevant? Br J Anaesth 2002 ; 89 : 715 – 721 . Google Scholar Crossref Search ADS PubMed WorldCat 43 Isgro F , Rehn E, Kiessling AH, Kretz KU, Kilian W, Saggau W. Platelet function test HemoSTATUS 2: tool or toy for optimised management of hemostasis? Perfusion 2002 ; 17 : 27 – 31 . Google Scholar Crossref Search ADS PubMed WorldCat 44 Shanberge JN , Murato M, Quattrociocchi-Longe T, van Neste L. Heparin–protamine complexes in the production of heparin rebound and other complications of extracorporeal bypass procedures . Am J Clin Pathol 1987 ; 87 : 210 – 217 . Google Scholar Crossref Search ADS PubMed WorldCat 45 Raymond PD , Ray MJ, Callen SN, Marsh NA. Heparin monitoring during cardiac surgery. Part 2: Calculating the overestimation of heparin by activated clotting time . Perfusion 2003 ; 18 : 277 – 281 . Google Scholar Crossref Search ADS PubMed WorldCat 46 Raymond PD , Ray MJ, Callen SN, Marsh NA. Heparin monitoring during cardiac surgery. Part 1: Validation of whole-blood heparin concentration and activated clotting time . Perfusion 2003 ; 18 : 269 – 276 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Pappalardo F , Franco A, Crescenzi G, de Simone F, Torracca L, Zangrillo A. Anticoagulation management in patients undergoing open heart surgery by activated clotting time and whole blood heparin concentration . Perfusion 2006 ; 21 : 285 – 290 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 48 Koster A , Fischer T, Praus M, Haberzettl H, Kuebler WM, Hetzer R et al. Hemostatic activation and inflammatory response during cardiopulmonary bypass: impact of heparin management . Anesthesiology 2002 ; 97 : 837 – 841 . Google Scholar Crossref Search ADS PubMed WorldCat 49 Slight RD , Buell R, Nzewi OC, McClelland DB, Mankad PS. A comparison of activated coagulation time-based techniques for anticoagulation during cardiac surgery with cardiopulmonary bypass . J Cardiothorac Vasc Anesth 2008 ; 22 : 47 – 52 . Google Scholar Crossref Search ADS PubMed WorldCat 50 Mongan PD , Hosking MP. The role of desmopressin acetate in patients undergoing coronary artery bypass surgery. A controlled clinical trial with thromboelastographic risk stratification . Anesthesiology 1992 ; 77 : 38 – 46 . Google Scholar Crossref Search ADS PubMed WorldCat 51 Welsby IJ , Jiao K, Ortel TL, Brudney CS, Roche AM, Bennett-Guerrero E et al. The kaolin-activated Thromboelastograph predicts bleeding after cardiac surgery . J Cardiothorac Vasc Anesth 2006 ; 20 : 531 – 535 . Google Scholar Crossref Search ADS PubMed WorldCat 52 Shih RL , Cherng YG, Chao A, Chen JT, Tsai AL, Lin CC. Prediction of bleeding diathesis in patients undergoing cardiopulmonary bypass during cardiac surgery: viscoelastic measures versus routine coagulation testing . Acta Anaesthesiol Sin 1997 ; 35 : 133 – 139 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 53 Ostrowsky J , Foes J, Warchol M, Tsarovsky G, Blay J. Plateletworks platelet function test compared to the thromboelastograph for prediction of post-operative outcomes . J Extra Corpor Technol 2004 ; 36 : 149 – 152 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 54 Shore-Lesserson L , Manspeizer HE, DePerio M, Francis S, Vale-Cantos F, Ergin MA. Thromboelastography guided transfusion algorithm reduces transfusions in complex cardiac surgery . Anesth Analg 1999 ; 88 : 312 – 319 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 55 Cherng YG , Chao A, Shih RL, Lin CS, Chan WH, Huang CH et al. Preoperative evaluation and postoperative prediction of hemostatic function with thromboelastography in patients undergoing redo cardiac surgery . Acta Anaesthesiol Sin 1998 ; 36 : 179 – 186 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 56 McNicol PL , Liu G, Harley ID, McCall PR, Przybylowski GM, Bowkett J et al. Blood loss and transfusion requirements in liver transplantation: experience with the first 75 cases . Anaesth Intensive Care 1994 ; 22 : 666 – 671 . Google Scholar Crossref Search ADS PubMed WorldCat 57 Gillies BS . Thromboelastography and liver transplantation . Semin Thromb Hemost 1995 ; 21 ( Suppl 4 ): 45 – 49 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 58 Kang YG , Martin DJ, Marquez J, Lewis JH, Bontempo FA, Shaw BW et al. Intraoperative changes in blood coagulation and thromboelastographic monitoring in liver transplantation . Anesth Analg 1985 ; 64 : 888 – 896 . Google Scholar Crossref Search ADS PubMed WorldCat 59 Hendriks HG , Meijer K, de Wolf JT, Porte RJ, Klompmaker IG, Lip H et al. Effect of recombinant activated factor VII on coagulation measured by thromboelastography in liver transplantation . Blood Coagul Fibrinolysis 2002 ; 13 : 309 – 313 . Google Scholar Crossref Search ADS PubMed WorldCat 60 Rugeri L , Levrat A, David JS, Delecroix E, Floccard B, Gross A et al. Diagnosis of early coagulation abnormalities in trauma patients by rotation thromboelastography . J Thromb Haemost 2006 ; 5 : 289 – 295 . Google Scholar Crossref Search ADS PubMed WorldCat 61 Kheirabadi BS , Crissey JM, Deguzman R, Holcomb JB. In vivo bleeding time and in vitro thromboelastography measurements are better indicators of dilutional hypothermic coagulopathy than prothrombin time . J Trauma 2007 ; 62 : 1352 – 1359 . Google Scholar Crossref Search ADS PubMed WorldCat 62 Vig S , Chitolie A, Bevan DH, Halliday A, Dormandy J. Thromboelastography: a reliable test? Blood Coagul Fibrinolysis 2001 ; 12 : 555 – 561 . Google Scholar Crossref Search ADS PubMed WorldCat 63 Samama CM , Ozier Y. Near-patient testing of haemostasis in the operating theatre: an approach to appropriate use of blood in surgery . Vox Sang 2003 ; 84 : 251 – 255 . Google Scholar Crossref Search ADS PubMed WorldCat 64 von Kualla KN , Ostendorf P, von Kaulla E. The impedance machine: a new bedside coagulation recording device . J Med 1975 ; 6 : 73 – 88 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 65 Tuman KJ , Speiss BD, McCarthy RJ, Ivankovich AD. Comparison of viscoelastic methods of coagulation after cardiopulmonary bypass . Anaesth Analg 1989 ; 69 : 69 – 75 . Google Scholar Crossref Search ADS WorldCat 66 Yamada T , Katori N, Tanaka KA, Takeda J. Impact of Sonoclot hemostasis analysis after cardiopulmonary bypass on postoperative hemorrhage in cardiac surgery . J Anesth 2007 ; 21 : 148 – 152 . Google Scholar Crossref Search ADS PubMed WorldCat 67 Faraday N , Guallar E, Sera VA, Bolton ED, Scharpf RB, Cartarius AM et al. Utility of whole blood hemostatometry using the Clot Signature Analyzer for assessment of hemostasus in cardiac surgery . Anesthesiology 2002 ; 96 : 1115 – 1122 . Google Scholar Crossref Search ADS PubMed WorldCat 68 Maquelin KN , Berckmans RJ, Nieuwland R, Schaap MC, ten Have K, Eijsman L et al. Disappearance of glycoprotein Ib from the platelet surface in pericardial blood during cardiopulmonary bypass . J Thorac Cardiovasc Surg 1998 ; 115 : 1160 – 1165 . Google Scholar Crossref Search ADS PubMed WorldCat 69 de Haan J , van Oeveren W. Platelets and soluble fibrin promote plasminogen actvation causing downregulation of platelet glycoprotein Ib/IX complexes: protection by aprotinin . Thromb Res 1998 ; 92 : 171 – 179 . Google Scholar Crossref Search ADS PubMed WorldCat 70 Greilich PE , Brouse C, Beckham J, Jessen ME, Martin EJ, Carr ME. Reductions in platelet contractile force correlate with duration of cardiopulmonary bypass and blood loss in patients undergoing cardiac surgery . Thromb Res 2002 ; 105 : 523 – 529 . Google Scholar Crossref Search ADS PubMed WorldCat 71 Smith SC Jr , Feldman TE, Hirshfeld JW Jr, Jacobs AK, Kern MJ, King SB et al. ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines . Circulation 2006 ; 113 : e166 – e286 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 72 Dunning J , Versteegh M, Fabbri A, Pavie A, Kolh P, Lochowandt U et al. Guideline on antiplatelet and anticoagulation management in cardiac surgery . Eur J Cardiothoracic Surg 2008 ; 34 : 73 – 92 . Google Scholar Crossref Search ADS WorldCat 73 Lennon MJ , Gibbs NM, Weightman WM, McGuire D, Michalopolous M. A comparison of Plateletworks and platelet aggregometry for the assessment of aspirin-related platelet dysfunction in cardiac surgical patients . J Cardiothorac Vasc Anesth 2004 ; 18 : 136 – 140 . Google Scholar Crossref Search ADS PubMed WorldCat 74 von Beckerath N , Pogatsa Murray G, Wieczoek A, Sibbing D, Schomig A, Kastrati A. Correlation of a new point-of-care test with conventional optical aggregometry for the assessment of clopidogrel responsiveness . Thromb Haemost 2006 ; 95 : 910 – 911 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 75 Ruggeri ZM . Platelets in atherothrombosis . Nat Med 2002 ; 8 : 1227 – 1234 . Google Scholar Crossref Search ADS PubMed WorldCat 76 Antithrombotic Trialists' Collaboration . Collaborative meta-analysis of randomised controlled trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high risk patients . BMJ 2002 ; 324 : 71 – 86 . Crossref Search ADS PubMed WorldCat 77 Gum PA , Kottke-Marchant K, Poggio ED, Gurm H, Welsh PA, Brooks L et al. Profile and prevalence of aspirin resistance in patients with cardiovascular disease . Am J Cardiol 2001 ; 88 : 230 – 235 . Google Scholar Crossref Search ADS PubMed WorldCat 78 Sonsken JR , Kong KL, Holder R. Magnitude and time course of impaired primary haemostasis after stopping chronic low and medium dose aspirin in healthy volunteers . Br J Anaesth 1999 ; 82 : 360 – 365 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 79 Fiore LD , Brophy MT, Lopez A, Janson P, Deykin D. The bleeding time response to aspirin. Identifying the hyperresponder . Am J Clin Pathol 1990 ; 94 : 292 – 296 . Google Scholar Crossref Search ADS PubMed WorldCat 80 Helgason CM , Bolin KM, Hoff JA, Winkler SR, Mangat A, Tortorice KL et al. Development of aspirin resistance in persons with previous ischemic stroke . Stroke 1994 ; 25 : 2331 – 2336 . Google Scholar Crossref Search ADS PubMed WorldCat 81 Gonzalez-Conejero R , Rivera J, Corral J, Acuna C, Guerrero JA, Vicente V. Biological assessment of aspirin efficacy on healthy individuals: heterogeneous response or aspirin failure? Stroke 2005 ; 36 : 276 – 280 . Google Scholar Crossref Search ADS PubMed WorldCat 82 Hovens MM , Snoep JD, Eikenboom JC, van der Bom JG, Mertens BJ, Huisman MV. Prevalence of persistent platelet reactivity despite use of aspirin: a systematic review . Am Heart J 2007 ; 153 : 175 – 181 . Google Scholar Crossref Search ADS PubMed WorldCat 83 Wang TH , Bhatt DL, Fox KA, Steinhubl SR, Brennan DM, Hacke W et al. An analysis of mortality rates with dual-antiplatelet therapy in the primary prevention population of the CHARISMA trial . Eur Heart J 2007 ; 28 : 2200 – 2207 . Google Scholar Crossref Search ADS PubMed WorldCat 84 Diener HC , Bogousslavsky J, Brass LM, Cimminiello C, Csiba L, Kaste M et al. , MATCH Investigators . Aspirin and clopidogrel compared with clopidogrel alone after recurrent ischaemic stroke or transient ischaemic attack in high-risk patients (MATCH): randomised, double-blind, placebo-controlled trial . Lancet 2004 ; 364 : 331 – 337 . Google Scholar Crossref Search ADS PubMed WorldCat 85 Swallow RA , Agarwala RA, Dawkins KD, Curzen NP. Thromboelastography: potential bedside tool to assess the effects of antiplatelet therapy? Platelets 2006 ; 17 : 385 – 392 . Google Scholar Crossref Search ADS PubMed WorldCat 86 Agarwal S , Coakley M, Reddy K, Riddell A, Mallett S. Quantifying the effect of anti-platelet therapy: a comparison of the platelet function analyzer (PFA-100) and modified thromboelastography (mTEG) with light transmission platelet aggregometry . Anesthesiology 2006 ; 105 : 676 – 683 . Google Scholar Crossref Search ADS PubMed WorldCat 87 Bliden KP , DiChiara J, Tantry US, Bassi AK, Chaganti SK, Gurbel PA. Increased risk in patients with high platelet aggregation receiving chronic clopidogrel therapy undergoing percutaneous coronary intervention: is the current antiplatelet therapy adequate? J Am Coll Cardiol 2007 ; 49 : 657 – 666 . Google Scholar Crossref Search ADS PubMed WorldCat 88 Hobson AR , Petley GW, Dawkins KD, Curzen N. A novel fifteen minute test for assessment of individual time-dependent clotting responses to aspirin and clopidogrel using modified thromboelastography . Platelets 2007 ; 18 : 497 – 505 . Google Scholar Crossref Search ADS PubMed WorldCat 89 Cerutti E , Stratta C, Romagnoli R, Schellino MM, Skurzak S, Rizzetto M et al. Thromboelastogram monitoring in the perioperative period of hepatectomy for adult living liver donation . Liver Transpl 2004 ; 10 : 289 – 294 . Google Scholar Crossref Search ADS PubMed WorldCat 90 Mahla E , Lang T, Vicenzi MN, Werkgartner G, Maier R, Probst C et al. Thromboelastography for monitoring prolonged hypercoagulability after major abdominal surgery . Anesth Analg 2001 ; 92 : 572 – 577 . Google Scholar Crossref Search ADS PubMed WorldCat 91 Roller RE , Dorr A, Ulrich S, Pilger E. Effect of aspirin treatment in patients with peripheral arterial disease monitored with the platelet function analyzer PFA-100 . Blood Coagul Fibrinolysis 2002 ; 13 : 277 – 281 . Google Scholar Crossref Search ADS PubMed WorldCat 92 Ziegler S , Maca T, Alt E, Speiser W, Schneider B, Minar E. Monitoring of anti-platelet therapy with the PFA-100 in peripheral angioplasty patients . Platelets 2002 ; 13 : 493 – 497 . Google Scholar Crossref Search ADS PubMed WorldCat 93 Hertfelder HJ , Bos M, Weber D, Winkler K, Hanfland P, Preusse CJ. Perioperative monitoring of primary and secondary hemostasis in coronary artery bypass grafting . Semin Thromb Hemost 2005 ; 31 : 426 – 440 . Google Scholar Crossref Search ADS PubMed WorldCat 94 Jacopo G , Elisabetta V, Silverio S, Massimiliano M, Sergio B, Grazia AM et al. Identification of platelet hyper-reactivity measured with a portable device immediately after percutaneous coronary intervention predicts in stent thrombosis . Thromb Res 2007 ; 121 : 407 – 412 . Google Scholar Crossref Search ADS PubMed WorldCat 95 Reny JL , Moerloose DE, Dauzat M, Fontana P. Use of the PFA-100 closure time to predict cardiovascular events in aspirin-treated cardiovascular patients: a systematic review and meta-analysis . J Thromb Haemost 2008 ; 6 : 444 – 450 . Google Scholar Crossref Search ADS PubMed WorldCat 96 Mueller T , Haltmayer M, Poeiz W, Haidinger D. Monitoring aspirin 100 mg and clopidogrel 75 mg with the PFA-100 device in patients with peripheral arterial disease . Vasc Endovasc Surg 2003 ; 37 : 117 – 123 . Google Scholar Crossref Search ADS WorldCat 97 Anderson K , Hurlen M, Amesen H, Seljeflot I. Aspirin non-responsiveness as measured by PFA-100 in patients with coronary artery disease . Thromb Res 2003 ; 108 : 37 – 42 . Google Scholar Crossref Search ADS WorldCat 98 Wong S , Ward C, Appleberg M, Lewis DR. Point of care testing of aspirin resistance in patients with vascular disease . ANZ J Surg 2006 ; 76 : 873 – 877 . Google Scholar Crossref Search ADS PubMed WorldCat 99 Panniccia R , Antonucci E, Gori AM, Marcucci R, Poli S, Romano E et al. Comparison of different methods to evaluate the effect of aspirin on platelet function in high-risk patients with ischemic heart disease receiving dual antiplatelet treatment . Am J Clin Pathol 2007 ; 128 : 143 – 149 . Google Scholar Crossref Search ADS PubMed WorldCat 100 Coma-Canella I , Velasco A, Castano S. Prevalence of aspirin resistance measured by PFA-100 . Int J Cardiol 2005 ; 101 : 71 – 76 . Google Scholar Crossref Search ADS PubMed WorldCat 101 Haubelt H , Anders C, Vogt A, Hoerdt P, Seyfert UT, Hellstern P. Variables influencing Platelet Function Analyzer-100 closure times in healthy individuals . Br J Haematol 2005 ; 130 : 759 – 767 . Google Scholar Crossref Search ADS PubMed WorldCat 102 Kotzailias N , Elwischger K, Sycha T, Rinner W, Quehenberger P, Auff E et al. Clopidogrel-induced platelet inhibition cannot be detected by the platelet function analyzer-100 system in stroke patients . J Stroke Cerebrovasc Dis 2007 ; 16 : 199 – 202 . Google Scholar Crossref Search ADS PubMed WorldCat 103 Cattaneo M . Resistance to antiplatelet drugs: molecular mechanisms and laboratory direction . J Thromb Haemost 2007 ; 5 ( Suppl 1 ): 230 – 237 . Google Scholar Crossref Search ADS PubMed WorldCat 104 Hayward CP , Harrison P, Cattaneo M, Ortel TL, Rao AK. Platelet function analyzer (PFA)-100 closure time in the evaluation of platelet disorders and platelet function . J Thromb Haemost 2006 ; 4 : 312 – 319 . Google Scholar Crossref Search ADS PubMed WorldCat 105 Wheeler GL , Braden GA, Steinhubl SR, Kereiakes DJ, Kottke-Marchant K, Michelson AD et al. The Ultegra rapid platelet function assay: comparison to standard platelet function assays in patients undergoing percutaneous coronary intervention with abciximab therapy . Am Heart J 2002 ; 143 : 602 – 611 . Google Scholar Crossref Search ADS PubMed WorldCat 106 Hochholzer W , Trenk D, Frundi D, Neumann FJ. Whole blood aggregometry for the evaluation of the anti-platelet effects of clopidogrel . Thromb Res 2007 ; 119 : 285 – 291 . Google Scholar Crossref Search ADS PubMed WorldCat 107 Steinhubl SR , Talley JD, Braden GA, Tcheng JE, Casterella PJ, Moltierno DJ et al. Point-of-care measured platelet inhibition correlates with a reduced risk of an adverse cardiac event after percutaneous coronary intervention . Circulation 2001 ; 103 : 2572 – 2578 . Google Scholar Crossref Search ADS PubMed WorldCat 108 Paniccia R , Antonucci E, Gori AM, Marcucci R, Giglioli C, Antoniucci D et al. Different methodologies for evaluating the effect of clopidogrel on platelet function in high-risk coronary artery disease patients . J Thromb Haemost 2007 ; 5 : 1839 – 1847 . Google Scholar Crossref Search ADS PubMed WorldCat 109 Malinin A , Pokov A, Spergling M, Defranco A, Schwartz K, Schwartz D et al. Monitoring platelet inhibition after clopidogrel with the VerfiyNow-P2Y12® rapid analyzer: the VERIfy Thrombosis risk ASsesssment (VERITAS) study . Thromb Res 2007 ; 119 : 277 – 284 . Google Scholar Crossref Search ADS PubMed WorldCat 110 Chen WH , Cheng X, Lee PY, Ng W, Kwok JY, Tse HF et al. Aspirin resistance and adverse clinical events in patients with coronary artery disease . Am J Med 2007 ; 120 : 631 – 635 . Google Scholar Crossref Search ADS PubMed WorldCat 111 Harrison P , Segal H, Silver L, Syed A, Cutherbertson FC, Rothwell PM. Lack of reproducibility of assessment of aspirin responsiveness by optical aggregometry and two platelet function tests . Platelets 2008 ; 19 : 119 – 124 . Google Scholar Crossref Search ADS PubMed WorldCat 112 Christiaens L , Macchi L. Monitoring of the antiplatelet drugs effect in patients with coronary artery disease: what is the real clinical impact? Curr Vasc Pharmacol 2007 ; 5 : 293 – 301 . Google Scholar Crossref Search ADS PubMed WorldCat 113 Mahmud E , Ang L. Monitoring antiplatelet therapy during peripheral vascular and coronary interventions . Tech Vasc Interv Radiol 2006 ; 9 : 56 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

Marret, E; Rolin, M; Beaussier, M; Bonnet, F

doi: 10.1002/bjs.6375pmid: 18844267

Abstract Background Continuous intravenous administration of lidocaine may decrease the duration of ileus and pain after abdominal surgery. Methods Three databases (Medline, Embase and the Cochrane Controlled Trials Register) were searched to retrieve randomized controlled trials comparing continuous intravenous lidocaine infusion during and after abdominal surgery with placebo. Study design was scored using the Oxford Quality Score based on randomization, double-blinding and follow-up. Outcome measures were duration of ileus, length of hospital stay, postoperative pain, and incidence of nausea and vomiting. Results Eight trials were selected. A total of 161 patients received intravenous lidocaine, with 159 controls. Intravenous lidocaine administration decreased the duration of ileus (weighted mean difference (WMD) − 8·36 h; P < 0·001), length of hospital stay (WMD − 0·84 days; P = 0·002), postoperative pain intensity at 24 h after operation on a 0–100-mm visual analogue scale (WMD − 5·93 mm; P = 0·002), and the incidence of nausea and vomiting (odds ratio 0·39; P = 0·006). Conclusion Continuous intravenous administration of lidocaine during and after abdominal surgery improves patient rehabilitation and shortens hospital stay. Introduction Postoperative ileus is a common reason for an extended hospital stay after major abdominal surgery, even when the surgical procedure is uncomplicated. Its pathophysiology is multifactorial. Anaesthetic agents such as opioids are thought to be among the causal factors. To reduce the stress response, the use of epidural anaesthesia with local anaesthetics has been advocated1. This provides better postoperative pain control than parenteral opioid analgesia and decreases surgical stress. Furthermore, it shortens the duration of postoperative ileus and, according to some studies, reduces the incidence of complications1–3. However, whether dependent on opioid or not, it increases the risk of urinary retention and arterial hypotension, induces partial muscle blockade thereby impairing mobilization and, in extreme but rare circumstances, causes serious complications such as epidural haematoma or abscess3,4. A meta-analysis has failed to demonstrate that epidural analgesia decreases length of hospital stay after colorectal surgery3. Intravenous lidocaine has analgesic, anti-inflammatory and antihyperalgesic properties5. As far back as in 1954, an intravenous infusion of lidocaine during general anaesthesia was shown to provide postoperative analgesia for more than 10 h, with a low incidence of postoperative nausea and vomiting6. Recent randomized controlled trials (RCTs) have suggested that continuous intravenous lidocaine administration may have beneficial effects on outcomes after colorectal surgery and may shorten hospital stay7,8. What follows is a systematic review of the literature aiming to assess the effect of intravenous lidocaine administration on recovery after abdominal surgery. Special emphasis has been given to the following endpoints: length of hospital stay, pain intensity, duration of ileus, and incidence of complications and side-effects. Methods This review was conducted according to the recommendations published in the Quality of Reporting of Meta-analyses (QUORUM) statement9. Literature review and identification of studies Three electronic databases were searched for studies published up to December 2007: Pubmed (Medline/Index Medicus), the Cochrane Controlled Trials Register and Embase. The medical subject heading (MeSH) terms used for the search were ‘lidocaine’ and ‘ileus’. Additional articles were retrieved through hyperlinks and by manually searching reference lists in original published articles, review articles and correspondence. There was no language restriction. Authors were contacted for additional information on methods and results when required. Study selection criteria Study selection criteria were abdominal surgery, a randomized double-blind design, and an Oxford Quality Score of at least 3 (see below). Exclusion criteria were inclusion of children, an Oxford Quality Score of below 3, no control group, comparisons of intravenous lidocaine infusion with epidural analgesia only, lidocaine administered by bolus with no continuous infusion (for example lidocaine co-administered with morphine by a patient-controlled analgesia device), and no perioperative lidocaine administration. Quality assessment of studies Study design quality was assessed by two investigators (M.R., E.M.) who were not blinded to the study authors or results. Disagreements were resolved by discussion with co-authors. Each article was scored using a five-point scale for randomization, blinding and patient follow-up10. A study was allocated one point if the design was randomized and an additional point if the randomization method was described and appropriate (such as a computer-generated table of random numbers). However, a point was subtracted if the randomization method was described but inappropriate (for example alternate allocation or allocation by date of birth). A study was also allocated one point if it was double-blind and two points if the double-blinding method was described and appropriate (for example identical placebo, active placebo, double-dummy). However, studies in which the double-blinding method was described but inappropriate received no points. Finally, one point was allocated to studies specifying numbers of, and reasons for, withdrawals and dropouts. The highest possible score was thus 5. Endpoints The primary endpoint was time to recover bowel function, that is the duration of postoperative ileus (time to first flatus, faeces or bowel movement). Secondary endpoints were length of hospital stay, 24- and 48-h postoperative pain scores measured on a visual analogue scale (VAS), opioid consumption, incidence of opioid side-effects, such as nausea and vomiting and sedation, and systemic lidocaine toxicity. The incidence of complications was recorded for each study as reported. To analyse continuous data, numerical data were extracted from the text of the article. If data were missing, the authors were contacted. In the absence of a reply, the data were extrapolated from figures. When nausea and vomiting were reported as separate outcomes, the authors were contacted to discover how many patients had nausea and vomiting. If they did not reply, the greater of the two numbers (number for either nausea or vomiting) was recorded. Statistical analysis If not reported in the article, an intention-to-treat analysis of the original data was carried out. All analyses were performed using Review Manager software (version 4.2, Cochrane Collaboration, The Nordic Cochrane Centre, Copenhagen, Denmark). For dichotomous data, the odds ratio and 95 per cent confidence interval (c.i.) were calculated using a fixed-effect model. When the test for heterogeneity (Cochran Q test) was significant (P < 0·100), a random-effects model analysis was carried out. For continuous data (length of hospital stay, VAS scores), weighted mean differences (WMDs) were calculated, taking into account study size and s.d. as reported in the individual trials. When mean(s.d.) values were not given, they were estimated from the median, range and size of the samples or from the interquartile range. Forest plots were used to show the information from the individual studies that was used in the meta-analysis. Sensitivity analysis was performed to explore the effect of lidocaine in different situations, namely cholecystectomy versus colonic resection and laparoscopic versus open surgery. The number needed to treat (NNT) was calculated as the reciprocal of the risk difference of postoperative nausea and vomiting between the lidocaine and control groups. The c.i. of the NNT were constructed by inverting and exchanging the limits of the 95 per cent c.i. for the risk difference. All tests were two sided, and P < 0·050 was considered statistically significant. Results Studies selected Of 65 articles retrieved by electronic and hand searching, 57 were excluded for the following reasons: 25 were letters or literature reviews, 12 were animal studies (ten concerned lidocaine use for ileus, colic or colonic surgery in horses), 11 were not relevant, six were studies on epidural lidocaine, two were orthopaedic studies and one was a critical care study (Fig. 1). The eight selected double-blind RCTs included patients scheduled for abdominal surgery only. Overall, 161 received intravenous lidocaine and 159 received placebo. Fig. 1 Open in new tabDownload slide Flow chart of systematic search Study details All eight RCTs were published between 1985 and 2007 and conducted in single centres (Table 1). All patients were American Association of Anesthesiologists (ASA) grade I–III. The procedure was open surgery in six trials and laparoscopy in two trials. The median Oxford Quality Score was 4 (range 3–5). Table 1 Randomized controlled trials comparing continuous intravenous lidocaine and placebo after abdominal surgery Reference . Oxford Quality Score . No. of patients . Type of surgery . Lidocaine administration . Additional measures . Endpoints . Lidocaine . Control . Cassuto et al.11 3 10 10 Open Bolus (100 mg), then Postop. pain (VAS) cholecystectomy 2 mg/min for 24 h Postop. opioid postop. PONV Groudine et al.12 5 20 20 Radical retropubic Bolus (1·5 mg/kg) NSAIDs Pain scores prostatectomy before induction, Postop. opioid then 3 mg/min Time to first flatulence or (BW > 70 kg) or bowel movement 2 mg/min Length of hospital stay (BW < 70 kg) for 1 h after skin closure Herroeder et al.8 5 31 29 Open colorectal Bolus (1·5 mg/kg) Fast-track protocol: Pain scores surgery before induction, normothermia, Gastrointestinal motility then 2 mg/min until PONV prophylaxis, Inflammatory mediators 4 h after skin closure paracetamol, Length of hospital stay NSAIDs, AOF Kaba et al.7 5 20 20 Laparoscopic Bolus (1·5 mg/kg) at Normothermia, Pain scores colectomy induction, then paracetamol, Postop. opioid 2 mg/kg/h intraop. NSAIDs, no postop. Fatigue scores and 1·33 mg/kg/h nasogastric tube, Length of hospital stay for 24 h postop. AOF, active Time to first flatus and mobilization defaecation Endocrine and metabolic responses Hospital discharge Koppert et al.13 5 20 20 Major abdominal Bolus (1·5 mg/kg in Length of hospital stay surgery 10 min), then Pain scores 1·5 mg/kg/h (started Postop. ileus 30 min before skin Postop. opioid incision) and PONV continued for 1 h after skin closure Kuo et al.14 4 20 20 Open colectomy Bolus (2 mg/kg), then Normothermia, Length of hospital stay for cancer 3 mg/kg/h started epidural analgesia Pain relief Postop. 30 min before surgery epidural consumption throughout surgery PONV Time to flatus Cytokine surge Rimback et al.15 3 15 15 Open Bolus (100 mg) before Return to motility cholecystectomy anaesthesia, then (radio-opaque markers, 3 mg/min for 24 h abnormal radiographs) postop. Duration of ileus Wu et al.16 5 25 25 Laparoscopic Lidocaine 3 mg/kg/h Normothermia Pain scores (VAS) cholecystectomy started 30 min Postop. opioid before and Time to first flatus continued throughout surgery Reference . Oxford Quality Score . No. of patients . Type of surgery . Lidocaine administration . Additional measures . Endpoints . Lidocaine . Control . Cassuto et al.11 3 10 10 Open Bolus (100 mg), then Postop. pain (VAS) cholecystectomy 2 mg/min for 24 h Postop. opioid postop. PONV Groudine et al.12 5 20 20 Radical retropubic Bolus (1·5 mg/kg) NSAIDs Pain scores prostatectomy before induction, Postop. opioid then 3 mg/min Time to first flatulence or (BW > 70 kg) or bowel movement 2 mg/min Length of hospital stay (BW < 70 kg) for 1 h after skin closure Herroeder et al.8 5 31 29 Open colorectal Bolus (1·5 mg/kg) Fast-track protocol: Pain scores surgery before induction, normothermia, Gastrointestinal motility then 2 mg/min until PONV prophylaxis, Inflammatory mediators 4 h after skin closure paracetamol, Length of hospital stay NSAIDs, AOF Kaba et al.7 5 20 20 Laparoscopic Bolus (1·5 mg/kg) at Normothermia, Pain scores colectomy induction, then paracetamol, Postop. opioid 2 mg/kg/h intraop. NSAIDs, no postop. Fatigue scores and 1·33 mg/kg/h nasogastric tube, Length of hospital stay for 24 h postop. AOF, active Time to first flatus and mobilization defaecation Endocrine and metabolic responses Hospital discharge Koppert et al.13 5 20 20 Major abdominal Bolus (1·5 mg/kg in Length of hospital stay surgery 10 min), then Pain scores 1·5 mg/kg/h (started Postop. ileus 30 min before skin Postop. opioid incision) and PONV continued for 1 h after skin closure Kuo et al.14 4 20 20 Open colectomy Bolus (2 mg/kg), then Normothermia, Length of hospital stay for cancer 3 mg/kg/h started epidural analgesia Pain relief Postop. 30 min before surgery epidural consumption throughout surgery PONV Time to flatus Cytokine surge Rimback et al.15 3 15 15 Open Bolus (100 mg) before Return to motility cholecystectomy anaesthesia, then (radio-opaque markers, 3 mg/min for 24 h abnormal radiographs) postop. Duration of ileus Wu et al.16 5 25 25 Laparoscopic Lidocaine 3 mg/kg/h Normothermia Pain scores (VAS) cholecystectomy started 30 min Postop. opioid before and Time to first flatus continued throughout surgery VAS, visual analogue scale; PONV, postoperative nausea and vomiting; BW, bodyweight; NSAID, non-steroidal anti-inflammatory drug; AOF, active oral feeding. Open in new tab Table 1 Randomized controlled trials comparing continuous intravenous lidocaine and placebo after abdominal surgery Reference . Oxford Quality Score . No. of patients . Type of surgery . Lidocaine administration . Additional measures . Endpoints . Lidocaine . Control . Cassuto et al.11 3 10 10 Open Bolus (100 mg), then Postop. pain (VAS) cholecystectomy 2 mg/min for 24 h Postop. opioid postop. PONV Groudine et al.12 5 20 20 Radical retropubic Bolus (1·5 mg/kg) NSAIDs Pain scores prostatectomy before induction, Postop. opioid then 3 mg/min Time to first flatulence or (BW > 70 kg) or bowel movement 2 mg/min Length of hospital stay (BW < 70 kg) for 1 h after skin closure Herroeder et al.8 5 31 29 Open colorectal Bolus (1·5 mg/kg) Fast-track protocol: Pain scores surgery before induction, normothermia, Gastrointestinal motility then 2 mg/min until PONV prophylaxis, Inflammatory mediators 4 h after skin closure paracetamol, Length of hospital stay NSAIDs, AOF Kaba et al.7 5 20 20 Laparoscopic Bolus (1·5 mg/kg) at Normothermia, Pain scores colectomy induction, then paracetamol, Postop. opioid 2 mg/kg/h intraop. NSAIDs, no postop. Fatigue scores and 1·33 mg/kg/h nasogastric tube, Length of hospital stay for 24 h postop. AOF, active Time to first flatus and mobilization defaecation Endocrine and metabolic responses Hospital discharge Koppert et al.13 5 20 20 Major abdominal Bolus (1·5 mg/kg in Length of hospital stay surgery 10 min), then Pain scores 1·5 mg/kg/h (started Postop. ileus 30 min before skin Postop. opioid incision) and PONV continued for 1 h after skin closure Kuo et al.14 4 20 20 Open colectomy Bolus (2 mg/kg), then Normothermia, Length of hospital stay for cancer 3 mg/kg/h started epidural analgesia Pain relief Postop. 30 min before surgery epidural consumption throughout surgery PONV Time to flatus Cytokine surge Rimback et al.15 3 15 15 Open Bolus (100 mg) before Return to motility cholecystectomy anaesthesia, then (radio-opaque markers, 3 mg/min for 24 h abnormal radiographs) postop. Duration of ileus Wu et al.16 5 25 25 Laparoscopic Lidocaine 3 mg/kg/h Normothermia Pain scores (VAS) cholecystectomy started 30 min Postop. opioid before and Time to first flatus continued throughout surgery Reference . Oxford Quality Score . No. of patients . Type of surgery . Lidocaine administration . Additional measures . Endpoints . Lidocaine . Control . Cassuto et al.11 3 10 10 Open Bolus (100 mg), then Postop. pain (VAS) cholecystectomy 2 mg/min for 24 h Postop. opioid postop. PONV Groudine et al.12 5 20 20 Radical retropubic Bolus (1·5 mg/kg) NSAIDs Pain scores prostatectomy before induction, Postop. opioid then 3 mg/min Time to first flatulence or (BW > 70 kg) or bowel movement 2 mg/min Length of hospital stay (BW < 70 kg) for 1 h after skin closure Herroeder et al.8 5 31 29 Open colorectal Bolus (1·5 mg/kg) Fast-track protocol: Pain scores surgery before induction, normothermia, Gastrointestinal motility then 2 mg/min until PONV prophylaxis, Inflammatory mediators 4 h after skin closure paracetamol, Length of hospital stay NSAIDs, AOF Kaba et al.7 5 20 20 Laparoscopic Bolus (1·5 mg/kg) at Normothermia, Pain scores colectomy induction, then paracetamol, Postop. opioid 2 mg/kg/h intraop. NSAIDs, no postop. Fatigue scores and 1·33 mg/kg/h nasogastric tube, Length of hospital stay for 24 h postop. AOF, active Time to first flatus and mobilization defaecation Endocrine and metabolic responses Hospital discharge Koppert et al.13 5 20 20 Major abdominal Bolus (1·5 mg/kg in Length of hospital stay surgery 10 min), then Pain scores 1·5 mg/kg/h (started Postop. ileus 30 min before skin Postop. opioid incision) and PONV continued for 1 h after skin closure Kuo et al.14 4 20 20 Open colectomy Bolus (2 mg/kg), then Normothermia, Length of hospital stay for cancer 3 mg/kg/h started epidural analgesia Pain relief Postop. 30 min before surgery epidural consumption throughout surgery PONV Time to flatus Cytokine surge Rimback et al.15 3 15 15 Open Bolus (100 mg) before Return to motility cholecystectomy anaesthesia, then (radio-opaque markers, 3 mg/min for 24 h abnormal radiographs) postop. Duration of ileus Wu et al.16 5 25 25 Laparoscopic Lidocaine 3 mg/kg/h Normothermia Pain scores (VAS) cholecystectomy started 30 min Postop. opioid before and Time to first flatus continued throughout surgery VAS, visual analogue scale; PONV, postoperative nausea and vomiting; BW, bodyweight; NSAID, non-steroidal anti-inflammatory drug; AOF, active oral feeding. Open in new tab In seven of the eight RCTs, a lidocaine bolus (1·5–2 mg/kg) was given before surgical incision followed by a continuous infusion during and after operation. In the RCT with no lidocaine bolus, infusion was started 30 min before skin incision16. Three trials reported, in the patients receiving lidocaine, that the end-tidal concentration of halogenated anaesthetic was reduced on average by 25 per cent7,14,16. Lidocaine or placebo was administered for a period ranging from throughout surgery to 24 h after operation. Control patients received intravenous infusion of isotonic saline in all RCTs. Exclusion criteria were chronic use of analgesics or steroids (six trials)7,8,12–14,16, use of laxatives (one)15, impaired liver function (three)7,8,11, a psychiatric disorder (one)7, ASA grade above II (two)12,14, renal disease (one)11, and cardiovascular disease, severe cardiac arrhythmia or treatment with antiarrhythmic drugs (three)8,11,13. Postoperative recovery programmes for fast-track surgery were implemented in four RCTs7,8,14,16. These programmes included multimodal analgesia with non-steroidal anti-inflammatory drugs (NSAIDs), normothermia during surgery, removal of the nasogastric tube at the end of the surgical procedure, active mobilization, active oral feeding and/or prophylaxis for nausea and vomiting7,8,14,16. The opioids used for postoperative pain relief were morphine, mepiridine or piritramide. All except one trial8 showed a significant 30–50 per cent reduction in opioid consumption in the postoperative period with lidocaine infusion. Outcomes Recovery of bowel function was evaluated in all except one RCT11. Duration of postoperative ileus was significantly diminished by a continuous intravenous infusion of lidocaine (WMD − 8·36 (95 per cent c.i. − 13·24 to − 3·47) h; P < 0·001) (Fig. 2). A subgroup analysis was conducted to explore the effects of intravenous lidocaine on postoperative gut dysfunction in different situations. Lidocaine decreased the duration of ileus significantly in the cholecystectomy subgroup (WMD − 1·23 (95 per cent c.i. − 2·12 to − 0·34) h; P = 0·007)15,16. Similarly, lidocaine was associated with a decrease in duration of ileus after colonic resection (WMD − 12·00 (95 per cent c.i. − 14·86 to − 9·13) h; P < 0·001)7,8,14. Lidocaine also decreased postoperative ileus in patients in whom a laparoscopy was performed (WMD − 1·06 (95 per cent c.i. − 2·00 to − 0·13); P = 0·030)7,16 or not (WMD − 7·90 (95 per cent c.i. − 9·88 to − 5·91) h; P < 0·001)8,12–15. Length of hospital stay was reported in five trials and was significantly shorter in patients receiving lidocaine than in controls (WMD − 0·84 (95 per cent c.i. − 1·38 to − 0·31) days; P = 0·002) (Fig. 3). Fig. 2 Open in new tabDownload slide Effect of intravenous lidocaine versus placebo on duration of postoperative ileus. *Values are mean(s.d.). Weighted mean differences (WMDs) are shown with 95 per cent confidence intervals Fig. 3 Open in new tabDownload slide Effect of intravenous lidocaine versus placebo on length of hospital stay. *Values are mean(s.d.). Weighted mean differences (WMDs) are shown with 95 per cent confidence intervals Postoperative pain intensity was measured with a 0–100-mm VAS in six RCTs including 250 patients. Pain scores at 24 h were significantly lower in patients receiving lidocaine than in controls (WMD − 5·93 (95 per cent c.i. − 9·63 to − 2·23); P = 0·002) (Fig. 4). Fig. 4 Open in new tabDownload slide Effect of intravenous lidocaine versus placebo on postoperative pain at 24 h after surgery. *Values are mean(s.d.). Weighted mean differences (WMDs) are shown with 95 per cent confidence intervals Nausea and vomiting as a single entity was reported in five trials (170 patients). Its incidence was 32 per cent in the lidocaine group and 52 per cent in the control group (odds ratio 0·39 (95 per cent c.i. 0·20 to 0·76); P = 0·006) (Fig. 5). The NNT to avoid one instance of nausea and vomiting was 5 (95 per cent c.i. 3 to 17). Fig. 5 Open in new tabDownload slide Effect of intravenous lidocaine versus placebo on postoperative nausea and vomiting. Odds ratios (ORs) are shown with 95 per cent confidence intervals Three trials11,13,15 evaluated sedation during the postoperative period. Six patients were sedated in the lidocaine groups versus seven in the control groups. Other side-effects of opioids were not always reported, precluding a pooled analysis. Cardiac arrhythmia with stable vital signs was reported in one patient who received lidocaine16. Discussion This meta-analysis of eight RCTs favours continuous perioperative intravenous lidocaine administration in that it reduces the duration of postoperative ileus, pain, nausea and vomiting, and the length of hospital stay. These endpoints are addressed in turn below. Gastrointestinal dysfunction after abdominal surgery has many causes, including autonomic nervous system dysfunction, inflammatory response, administration of anaesthetics and opioids, and gastrointestinal hormone disruption17. Intravenous lidocaine may shorten the duration of ileus by reducing opioid consumption (as noted in seven of eight trials), by preventing inflammatory processes and by decreasing sympathetic tone. In the two trials that addressed inflammatory response, systemic lidocaine significantly blunted any postoperative rise in plasma concentration of proinflammatory cytokines and complement, as well as any increase in integrin (CD11b) and selectin (CD62L and CD62P) expression at the surface of leucocyte and platelet–leucocyte aggregates8,14. However, the effects of intravenous lidocaine on sympathetic tone are less certain. At least two groups did not detect any difference in plasma or urinary catecholamine concentrations in patients who did, and who did not, receive intravenous lidocaine during and after surgery7,18. Nevertheless, direct blockade of a tonic inhibitory action in the mesenteric nervous plexus leading to greater responsiveness to contractile stimulation cannot be ruled out. In experimental studies, lidocaine produces action potentials and increases the amplitude of contractions in intestinal muscle, probably by suppressing intrinsic nervous inhibition19,20. In addition, it can stimulate responses in intact intestinal wall and ganglion-free muscle preparations21. These features suggest that postoperative ileus may be shorter as a result of a direct effect on smooth muscle cells. Local anaesthetics to relieve pain after abdominal surgery are usually administered continuously through an epidural catheter. However, epidural analgesia is inadequate in almost 30 per cent of patients because the catheter is either removed prematurely or malpositioned22,23. In addition, epidural analgesia conveys a risk of side-effects, such as hypotension, urinary retention and motor blockade. Intravenous lidocaine is, on the other hand, devoid of side-effects. The meta-analysis has established that intravenous lidocaine significantly decreases the intensity of postoperative pain and reduces opioid consumption. Consequently, lidocaine appears to be an appropriate option for pain relief when epidural analgesia is not possible, or when it is inappropriately invasive for the surgical procedure in question (for example video laparoscopy). The mechanism of pain relief by lidocaine may be inhibition of the ectopic impulse discharge generated at sites of experimental nerve injury and in axotomized dorsal root ganglion24. In experimental studies, lidocaine treatment before trauma suppressed secondary hyperalgesia mainly by peripheral mechanisms25,26. The safety of continuous intravenous lidocaine has not been evaluated in a large cohort of surgical patients but no local toxicity other than a single episode of transient arrhythmia16 was observed in any of the trials in this meta-analysis. The meta-analysis has also shown that continuous intravenous lidocaine administration reduces the length of hospital stay. Treatment of postoperative ileus by gastrointestinal opioid receptor antagonists, or by multimodal analgesia with NSAIDs, may also shorten stay after abdominal surgery by accelerating recovery of bowel function27,28. However, epidural analgesia after colonic surgery, despite decreasing postoperative ileus, does not reduce hospital stay3. Possible reasons include side-effects or technical failures, or because reducing hospital stay was simply not an objective of these trials3. In addition, active rehabilitation implemented in the more recent trials of epidural analgesia might mask an effect of epidural analgesia on hospital stay3. A direct comparison between epidural analgesia and intravenous lidocaine administration in patients within an active rehabilitation protocol might well be worthwhile. By lengthening hospital stay after elective surgery, ileus has an impact on hospital costs29. In a retrospective review of 83 patients, 25 per cent of hemicolectomies and 18 per cent of hysterectomies were associated with postoperative ileus and prolonged hospital stay; this increased costs per patient by US $ 12 416 (€8481) and $ 4512 (€3082) per patient respectively30. The benefit in terms of postoperative ileus is pronounced patients having colonic resection or an open procedure. Intravenous lidocaine may therefore be of greater interest in patients scheduled for major abdominal surgery where postoperative ileus has an impact on hospital stay. For colonic resection, lidocaine decreases ileus by around 12 h. Interestingly, some authors have found that decreasing time to first flatus by even half a day accelerates time to hospital discharge28,31. Nausea and vomiting may also increase costs after surgery. They are reduced in patients receiving intravenous lidocaine, probably because of lower opioid consumption32. Prophylactic antiemetic therapy has been shown to be cost-effective33. In summary, as the present meta-analysis has shown a significant reduction of nausea and vomiting, duration of ileus and hospital stay, intravenous lidocaine may be considered as a cost-effective strategy after abdominal surgery. The meta-analysis has several limitations. First, data were included only from patients undergoing abdominal surgery. Intravenous lidocaine inhibits visceromotor and cardiovascular reflexes evoked by colorectal distension suggesting a clear benefit on visceral pain34. The conclusions cannot, however, be extrapolated to other settings, such as orthopaedic surgery. Second, the duration of postoperative ileus was the primary endpoint in only three RCTs7,12,15. Third, all outcomes were not reported in all trials (Table 1). Finally, only RCTs with small patient numbers were retrieved, and the discrepancies between meta-analyses and large RCTs are well known35. On the other hand, a distinct advantage of the present meta-analysis is that all eight included trials were double-blind RCTs with a high Oxford Quality Score36. The risk of concluding that there is a benefit when there is none is, therefore, reduced. In conclusion, continuous perioperative and postoperative intravenous lidocaine infusion after abdominal surgery decreases the duration of ileus, the incidence of nausea and vomiting, the severity of pain and the length of hospital stay. It offers a simple clinical solution that can be applied in conjunction with, or instead of, the many popular postoperative regimens currently in use. References 1 Liu S , Carpenter RL, Neal JM. Epidural anesthesia and analgesia. Their role in postoperative outcome . Anesthesiology 1995 ; 82 : 1474 – 1506 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Wu CL , Cohen SR, Richman JM, Rowlingson AJ, Courpas GE, Cheung K et al. Efficacy of postoperative patient-controlled and continuous infusion epidural analgesia versus intravenous patient-controlled analgesia with opioids: a meta-analysis . Anesthesiology 2005 ; 103 : 1079 – 1088 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Marret E , Remy C, Bonnet F. Meta-analysis of epidural analgesia versus parenteral opioid analgesia after colorectal surgery . Br J Surg 2007 ; 94 : 665 – 673 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Giebler RM , Scherer RU, Peters J. Incidence of neurologic complications related to thoracic epidural catheterization . Anesthesiology 1997 ; 86 : 55 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Hollmann MW , Durieux ME. Local anesthetics and the inflammatory response: a new therapeutic indication? Anesthesiology 2000 ; 93 : 858 – 875 . Google Scholar Crossref Search ADS PubMed WorldCat 6 de Clive-Lowe SG , Gray PWS, North J. Succinyldicholine and lignocaine by continuous intravenous drip: report of 1000 administrations . Anaesthesia 1954 ; 9 : 96 – 104 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Kaba A , Laurent SR, Detroz BJ, Sessler DI, Durieux ME, Lamy ML et al. Intravenous lidocaine infusion facilitates acute rehabilitation after laparoscopic colectomy . Anesthesiology 2007 ; 106 : 11 – 18 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Herroeder S , Pecher S, Schonherr ME, Kaulitz G, Hahnenkamp K, Friess H et al. Systemic lidocaine shortens length of hospital stay after colorectal surgery: a double-blinded, randomized, placebo-controlled trial . Ann Surg 2007 ; 246 : 192 – 200 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Moher D , Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of Reporting of Meta-analyses . Lancet 1999 ; 354 : 1896 – 1900 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Jadad AR , Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996 ; 17 : 1 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Cassuto J , Wallin G, Hogstrom S, Faxen A, Rimback G. Inhibition of postoperative pain by continuous low-dose intravenous infusion of lidocaine . Anesth Analg 1985 ; 64 : 971 – 974 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Groudine SB , Fisher HA, Kaufman RP Jr, Patel MK, Wilkins LJ, Mehta SA et al. Intravenous lidocaine speeds the return of bowel function, decreases postoperative pain, and shortens hospital stay in patients undergoing radical retropubic prostatectomy . Anesth Analg 1998 ; 86 : 235 – 239 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 13 Koppert W , Weigand M, Neumann F, Sittl R, Schuettler J, Schmelz M et al. Perioperative intravenous lidocaine has preventive effects on postoperative pain and morphine consumption after major abdominal surgery . Anesth Analg 2004 ; 98 : 1050 – 1055 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 14 Kuo CP , Jao SW, Chen KM, Wong CS, Yeh CC, Sheen MJ et al. Comparison of the effects of thoracic epidural analgesia and i.v. infusion with lidocaine on cytokine response, postoperative pain and bowel function in patients undergoing colonic surgery . Br J Anaesth 2006 ; 97 : 640 – 646 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Rimback G , Cassuto J, Tollesson PO. Treatment of postoperative paralytic ileus by intravenous lidocaine infusion . Anesth Analg 1990 ; 70 : 414 – 419 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 16 Wu CT , Borel CO, Lee MS, Yu JC, Liou HS, Yi HD et al. The interaction effect of perioperative cotreatment with dextromethorphan and intravenous lidocaine on pain relief and recovery of bowel function after laparoscopic cholecystectomy . Anesth Analg 2005 ; 100 : 448 – 453 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Holte K , Kehlet H. Postoperative ileus: a preventable event . Br J Surg 2000 ; 87 : 1480 – 1493 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Wallin G , Cassuto J, Hogstrom S, Linden I, Faxen A, Rimback G et al. Effects of lidocaine infusion on the sympathetic response to abdominal surgery . Anesth Analg 1987 ; 66 : 1008 – 1013 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Biber B , Fara J. Intestinal motility increased by tetrodotoxin, lidocaine, and procaine . Experientia 1973 ; 29 : 551 – 552 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Wood JD . Excitation of intestinal muscle by atropine, tetrodotoxin, and xylocaine . Am J Physiol 1972 ; 222 : 118 – 125 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Bortoff A , Muller R. Stimulation of intestinal smooth muscle by atropine, procaine, and tetrodotoxin . Am J Physiol 1975 ; 229 : 1609 – 1613 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Ready LB . Acute pain: lessons learned from 25 000 patients . Reg Anesth Pain Med 1999 ; 24 : 499 – 505 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 23 Rigg JR , Jamrozik K, Myles PS, Silbert BS, Peyton PJ, Parsons RW et al. Epidural anaesthesia and analgesia and outcome of major surgery: a randomised trial . Lancet 2002 ; 359 : 1276 – 1282 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Devor M , Wall PD, Catalan N. Systemic lidocaine silences ectopic neuroma and DRG discharge without blocking nerve conduction . Pain 1992 ; 48 : 261 – 268 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Kawamata M , Sugino S, Narimatsu E, Yamauchi M, Kiya T, Furuse S et al. Effects of systemic administration of lidocaine and QX-314 on hyperexcitability of spinal dorsal horn neurons after incision in the rat . Pain 2006 ; 122 : 68 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Kawamata M , Takahashi T, Kozuka Y, Nawa Y, Nishikawa K, Narimatsu E et al. Experimental incision-induced pain in human skin: effects of systemic lidocaine on flare formation and hyperalgesia . Pain 2002 ; 100 : 77 – 89 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Taguchi A , Sharma N, Saleem RM, Sessler DI, Carpenter RL, Seyedsadr M et al. Selective postoperative inhibition of gastrointestinal opioid receptors . N Engl J Med 2001 ; 345 : 935 – 940 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Sinatra RS , Boice JA, Loeys TL, Ko AT, Kashuba MM, Jahr JS et al. Evaluation of the effect of perioperative rofecoxib treatment on pain control and clinical outcomes in patients recovering from gynecologic abdominal surgery: a randomized, double-blind, placebo-controlled clinical study . Reg Anesth Pain Med 2006 ; 31 : 134 – 142 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 29 Bosio RM , Delaney CP, Senagore AJ. Economic impact of POI and prolonged length of stay . Semin Colon Rectal Surg 2005 ; 16 : 235 – 238 . Google Scholar Crossref Search ADS WorldCat 30 Salvador CG , Sikirica M, Evans A, Pizzi L, Goldfarb N. Clinical and economic outcomes of prolonged postoperative ileus in patients undergoing hysterectomy and hemicolectomy . Pharmacy and Therapeutics 2005 ; 30 : 590 – 595 . Google Scholar OpenURL Placeholder Text WorldCat 31 Senagore AJ , Bauer JJ, Du W, Techner L. Alvimopan accelerates gastrointestinal recovery after bowel resection regardless of age, gender, race, or concomitant medication use . Surgery 2007 ; 142 : 478 – 486 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Marret E , Kurdi O, Zufferey P, Bonnet F. Effects of nonsteroidal antiinflammatory drugs on patient-controlled analgesia morphine side effects: meta-analysis of randomized controlled trials . Anesthesiology 2005 ; 102 : 1249 – 1260 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Hill RP , Lubarsky DA, Phillips-Bute B, Fortney JT, Creed MR, Glass PS et al. Cost-effectiveness of prophylactic antiemetic therapy with ondansetron, droperidol, or placebo . Anesthesiology 2000 ; 92 : 958 – 967 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Ness TJ . Intravenous lidocaine inhibits visceral nociceptive reflexes and spinal neurons in the rat . Anesthesiology 2000 ; 92 : 1685 – 1691 . Google Scholar Crossref Search ADS PubMed WorldCat 35 LeLorier J , Gregoire G, Benhaddad A, Lapierre J, Derderian F. Discrepancies between meta-analyses and subsequent large randomized, controlled trials . N Engl J Med 1997 ; 337 : 536 – 542 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Moher D , Pham B, Jones A, Cook DJ, Jadad AR, Moher M et al. Does quality of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses? Lancet 1998 ; 352 : 609 – 613 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes Presented in part to the Annual Meeting of the American Society of Anesthesiologists, San Francisco, California, USA, October 2007, and to the Annual Meeting of the Société Francaise d'Anesthésie Réanimation, Paris, France, September 2007 Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

Patti, R; Arcara, M; Bonventre, S; Sammartano, S; Sparacello, M; Vitello, G; Di Vita, G

doi: 10.1002/bjs.6236pmid: 18844269

Abstract Background Thrombosed external haemorrhoids are one of the most frequent anorectal emergencies. They are associated with swelling and intense pain. Internal sphincter hypertonicity plays a role in the aetiology of the pain. This study evaluated the efficacy and safety of an intrasphincteric injection of botulinum toxin for pain relief in patients with thrombosed external haemorrhoids. Methods Thirty patients with thrombosed external haemorrhoids who refused surgical operation were randomized into two groups. Patients received an intrasphincteric injection of either 0·6 ml saline or 0·6 ml of a solution containing 30 units botulinum toxin. Anorectal manometry was performed before treatment and 5 days afterwards. Results After 5 days of treatment, the maximum resting pressure fell in both groups, but was significantly lower in the botulinum toxin group (P = 0·004). Pain intensity was significantly reduced within 24 h of botulinum toxin treatment (P < 0·001), but only after 1 week in the placebo group (P = 0·019). Conclusion A single injection of botulinum toxin into the anal sphincter seems to be effective in rapidly controlling the pain associated with thrombosed external haemorrhoids, and could represent an effective conservative treatment for this condition. Registration number: NCT00717782 (http://www.clinicaltrials.gov). Introduction Thrombosed external haemorrhoids (TEH) are one of the most frequently diagnosed anorectal emergencies1. They commonly occur in young adult patients of both sexes2. Clinically, TEH present with a lump and intense pain. Pain is disproportionately severe given the size of these lesions, the thrombus forming within the lumen of the external haemorrhoidal vessels. External haemorrhoidal thrombosis can be associated with constipation or diarrhoea, but often there is no history of straining or exertion3,4. TEH are debilitating, and have an economic impact because of time off work, cost of medication and reduction in quality of life. The pain and lump may take several days or weeks to resolve and symptoms may be incapacitating. Internal anal sphincter hypertonicity has a role in the pain associated with TEH and is responsible for the maximum resting anal pressure (MRP), but few studies have measured the MRP in patients with TEH. Shafik demonstrated increased electromyographic activity in the internal anal sphincter associated with TEH, and pain relief afforded by a warm Sitzbath was associated with relaxation of the internal anal sphincter5. Some drugs that act by reducing internal anal sphincter tone have proved effective in the conservative treatment of TEH. In 1995, Gorfine reported complete or nearly complete pain relief within 2–3 min after the administration of 500–1000 mg 0·5 per cent nitroglycerin ointment to patients with TEH3. Pain relief was temporally associated with a decline in anal pressure3. In a more recent trial, topical application of 0·2 per cent glyceryl trinitrate (GTN) ointment was compared with incision and excision. Pain relief achieved by excision was significantly greater than that achieved by incision or GTN treatment6. In a randomized open comparison of topical therapy with lidocaine alone versus lidocaine plus 0·3 per cent nifedipine ointment, addition of nifedipine was associated with significant pain relief after 7 days of therapy7. Clinical studies have demonstrated that injection of botulinum toxin into the anal sphincter can temporarily reduce MRP in patients with chronic anal fissures8 and after haemorrhoidectomy9,10. This aim of this randomized study was to evaluate the efficacy and safety of intrasphincteric injection of botulinum toxin for pain relief in patients with TEH. Methods Thirty consecutive patients with TEH were included in this study (Fig. 1). After careful explanation of all available surgical options and their benefit–risk balance, including excision, incision and evacuation of the thrombus, and anal dilatation, they refused surgery for fear of procedure-related pain. Fig. 1 Open in new tabDownload slide Recruitment for randomized trial according to the CONSORT diagram Exclusion criteria were a history of previous anal surgery, presence of third- and fourth-degree haemorrhoids, anal fissure or fistula, and presumed or confirmed pregnancy. No patient included had taken any medication that could have modified anal sphincter tone during the preceding 3 months. None received oral antithrombotic drugs or had decompensated diabetes. Patients gave written informed consent, and the local ethics committee and departmental internal review board approved the study. In all patients, anorectal manometry was performed before and 5 days after treatment. Baseline normal values were derived from measurements obtained previously for 12 healthy subjects of mean(s.d.) age 42·3(13·4) years with no previous or concomitant history of anorectal disease. These comprised six men and six women, of whom four had previously had a caesarean delivery and two had not had children. Anorectal manometry was performed using a manometric sensor, with an external diameter of 2·1 mm, four circular orifices and a latex microballoon at its extremity (Marquat C87; Marquat Génie Biomédical, Boissy-Saint-Léger, France), connected to a polygraph (Byo-System MMS 200; Narco, Houston, Texas, USA). A station pull-through method was used, with perfusion of normal saline with the patient lying in the right lateral position. MRP was defined as the maximum pressure detected at rest. Upon recruitment (Fig. 1), patients were allocated to either a placebo or botulinum toxin group using a random number table. Both groups received therapy based on a high-fibre diet, bulk laxative, Sitzbath for 10 min at least three times a day, and oral analgesics (100-mg nimesulide tablets as requested). Before treatment, all patients were reassured about the procedure by a nurse. All patients were treated in the office setting during a day clinic. Patients in the placebo group received a 0·6-ml injection of saline, whereas those in botulinum group were injected with 0·6 ml of a solution containing 30 units botulinum toxin A (Botox®; Allergan, Westport, Ireland). In both groups a 27-G needle was used to give two injections of equal volume (0·3 ml) into the internal anal sphincter, one on each side of the anterior midline of the sphincter. Injections were performed after anal sphincter palpation without use of local anaesthesia. The dose of botulinum toxin and the site of injection were similar to those used for the treatment of chronic anal fissure11. Symptom onset, intensity of pain, analgesic consumption, time to healing, time to return to work, complications and side-effects of injection of botulinum toxin or normal saline were recorded on a pro forma. Pain was assessed daily for 2 weeks using a linear analogue scale ranging from 0 (no pain) to 10 (worse conceivable pain)12. Analgesic consumption over 14 days was noted. Anal incontinence both before and after treatments was evaluated using the Pescatori grading system13. All clinical variables were assessed by the same clinician (R.P.), who was blinded to the patient groups. Statistical analysis A sample size calculation estimated that 15 patients would be required in each group to show a difference of one standard deviation between the two groups with regard to pain level, with a power of 0·8 and α level of 0·05. Continuous variables were expressed as mean(s.d.) and qualitative data as absolute frequencies. Student's t test was used for statistical analysis of continuous variables and χ2 test to identify differences in proportions between the two groups. P < 0·050 was considered statistically significant. Pain scores in the two groups at each registration point were compared by means of a Student's t test for paired data. Results Fifteen patients were randomized to each group and included in the analysis. There were no significant differences between groups in age, sex, number of thrombosed piles, previous episodes of THE, and time between the onset of TEH and treatment (Table 1). Nor were there any significant differences with respect to the presence of internal haemorrhoids; 20 patients had first- and second-degree internal haemorrhoids, whereas the remaining ten were normal (Table 2). No patient complained of acute haemorrhoidal disease at the time of enrolment. No haematoma or infection developed at the injection site. No patient had surgery during the study period. Table 1 Patient characteristics . Botulinum toxin (n = 15) . Placebo (n = 15) . P† . Age (years)* 39·4(15·8) 40·6(14·8) 0·416 Sex ratio (M : F) 8 : 7 6 : 9 0·464‡ No. of thrombosed piles* 1·3(0·6) 1·4(0·6) 0·283 No. of previous TEH* 1·7(0·4) 1·5(0·6) 0·167 Time between onset of TEH and treatment (h)* 45·4(15·6) 46·2(14·1) 0·442 . Botulinum toxin (n = 15) . Placebo (n = 15) . P† . Age (years)* 39·4(15·8) 40·6(14·8) 0·416 Sex ratio (M : F) 8 : 7 6 : 9 0·464‡ No. of thrombosed piles* 1·3(0·6) 1·4(0·6) 0·283 No. of previous TEH* 1·7(0·4) 1·5(0·6) 0·167 Time between onset of TEH and treatment (h)* 45·4(15·6) 46·2(14·1) 0·442 * Values are mean(s.d.). TEH, thrombosed external haemorrhoids. † Student's t test unless indicated otherwise; ‡ χ2 test. Open in new tab Table 1 Patient characteristics . Botulinum toxin (n = 15) . Placebo (n = 15) . P† . Age (years)* 39·4(15·8) 40·6(14·8) 0·416 Sex ratio (M : F) 8 : 7 6 : 9 0·464‡ No. of thrombosed piles* 1·3(0·6) 1·4(0·6) 0·283 No. of previous TEH* 1·7(0·4) 1·5(0·6) 0·167 Time between onset of TEH and treatment (h)* 45·4(15·6) 46·2(14·1) 0·442 . Botulinum toxin (n = 15) . Placebo (n = 15) . P† . Age (years)* 39·4(15·8) 40·6(14·8) 0·416 Sex ratio (M : F) 8 : 7 6 : 9 0·464‡ No. of thrombosed piles* 1·3(0·6) 1·4(0·6) 0·283 No. of previous TEH* 1·7(0·4) 1·5(0·6) 0·167 Time between onset of TEH and treatment (h)* 45·4(15·6) 46·2(14·1) 0·442 * Values are mean(s.d.). TEH, thrombosed external haemorrhoids. † Student's t test unless indicated otherwise; ‡ χ2 test. Open in new tab Table 2 Degree of internal haemorrhoids in patients with thrombosed external haemorrhoids Degree of haemorrhoids . Botulinum toxin (n = 15) . Placebo (n = 15) . P* . None 5 5 1·000 I 5 4 0·690 II 5 6 0·705 Degree of haemorrhoids . Botulinum toxin (n = 15) . Placebo (n = 15) . P* . None 5 5 1·000 I 5 4 0·690 II 5 6 0·705 * Pearson χ2 test. Open in new tab Table 2 Degree of internal haemorrhoids in patients with thrombosed external haemorrhoids Degree of haemorrhoids . Botulinum toxin (n = 15) . Placebo (n = 15) . P* . None 5 5 1·000 I 5 4 0·690 II 5 6 0·705 Degree of haemorrhoids . Botulinum toxin (n = 15) . Placebo (n = 15) . P* . None 5 5 1·000 I 5 4 0·690 II 5 6 0·705 * Pearson χ2 test. Open in new tab Manometry The reference MRP value for the 12 normal subjects was 77(12) mmHg. MRP was 98(14) mmHg in the placebo group and 96(13) mmHg in the botulinum group, significantly higher than in the normal subjects (both P < 0·001). After 5 days the MRP in patients injected with botulinum toxin was 80(8) mmHg, compared with 92(12) mmHg in the placebo group (P = 0·004). The value after 5 days in the botulinum group was significantly lower than the pretreatment pressure (P < 0·001) and similar to that in normal controls (P = 0·410). The MRP in the placebo group was lower than the pretreatment value (P = 0·228), but higher than that in normal subjects (P = 0·002). Pain and analgesic consumption There was no difference between groups in pain intensity before treatment (pain score 5·9(1·8) in botulinum group versus 6·2(1·7) in placebo group; P = 0·324). Pain scores were lower in the botulinum group at the first nine assessments following treatment, after which the pain scores were similar (Table 3). The pain intensity was significantly reduced in the botulinum group within 24 h of injection (P < 0·001), whereas in the placebo group a significant reduction was noted from day 7 onwards (P = 0·019). Table 3 Pain scores evaluated by visual analogue scale Time after treatment (days) . Botulinum toxin (n = 15) . Placebo (n = 15) . P* . Before treatment 5·9(1·8) 6·2(1·7) 0·324 1 3·9(1·3) 6·0(1·5) < 0·001 2 3·8(1·4) 5·9(1·5) < 0·001 3 3·7(1·6) 5·8(1·6) < 0·001 4 3·7(1·4) 5·6(1·7) 0·002 5 3·5(1·6) 5·6(1·5) < 0·001 6 3·3(1·2) 5·0(1·7) 0·004 7 3·0(1·2) 4·6(1·7) 0·006 8 2·8(1·4) 4·5(1·5) 0·003 9 2·6(0·9) 4·0(1·7) 0·009 10 2·3(0·6) 3·0(1·4) 0·060 11 2·2(0·5) 2·8(1·5) 0·153 12 1·9(0·5) 2·4(1·0) 0·094 13 1·7(0·4) 2·2(0·9) 0·059 14 1·3(0·5) 1·7(0·8) 0·112 Time after treatment (days) . Botulinum toxin (n = 15) . Placebo (n = 15) . P* . Before treatment 5·9(1·8) 6·2(1·7) 0·324 1 3·9(1·3) 6·0(1·5) < 0·001 2 3·8(1·4) 5·9(1·5) < 0·001 3 3·7(1·6) 5·8(1·6) < 0·001 4 3·7(1·4) 5·6(1·7) 0·002 5 3·5(1·6) 5·6(1·5) < 0·001 6 3·3(1·2) 5·0(1·7) 0·004 7 3·0(1·2) 4·6(1·7) 0·006 8 2·8(1·4) 4·5(1·5) 0·003 9 2·6(0·9) 4·0(1·7) 0·009 10 2·3(0·6) 3·0(1·4) 0·060 11 2·2(0·5) 2·8(1·5) 0·153 12 1·9(0·5) 2·4(1·0) 0·094 13 1·7(0·4) 2·2(0·9) 0·059 14 1·3(0·5) 1·7(0·8) 0·112 Values are mean(s.d.). * Student's t test. Open in new tab Table 3 Pain scores evaluated by visual analogue scale Time after treatment (days) . Botulinum toxin (n = 15) . Placebo (n = 15) . P* . Before treatment 5·9(1·8) 6·2(1·7) 0·324 1 3·9(1·3) 6·0(1·5) < 0·001 2 3·8(1·4) 5·9(1·5) < 0·001 3 3·7(1·6) 5·8(1·6) < 0·001 4 3·7(1·4) 5·6(1·7) 0·002 5 3·5(1·6) 5·6(1·5) < 0·001 6 3·3(1·2) 5·0(1·7) 0·004 7 3·0(1·2) 4·6(1·7) 0·006 8 2·8(1·4) 4·5(1·5) 0·003 9 2·6(0·9) 4·0(1·7) 0·009 10 2·3(0·6) 3·0(1·4) 0·060 11 2·2(0·5) 2·8(1·5) 0·153 12 1·9(0·5) 2·4(1·0) 0·094 13 1·7(0·4) 2·2(0·9) 0·059 14 1·3(0·5) 1·7(0·8) 0·112 Time after treatment (days) . Botulinum toxin (n = 15) . Placebo (n = 15) . P* . Before treatment 5·9(1·8) 6·2(1·7) 0·324 1 3·9(1·3) 6·0(1·5) < 0·001 2 3·8(1·4) 5·9(1·5) < 0·001 3 3·7(1·6) 5·8(1·6) < 0·001 4 3·7(1·4) 5·6(1·7) 0·002 5 3·5(1·6) 5·6(1·5) < 0·001 6 3·3(1·2) 5·0(1·7) 0·004 7 3·0(1·2) 4·6(1·7) 0·006 8 2·8(1·4) 4·5(1·5) 0·003 9 2·6(0·9) 4·0(1·7) 0·009 10 2·3(0·6) 3·0(1·4) 0·060 11 2·2(0·5) 2·8(1·5) 0·153 12 1·9(0·5) 2·4(1·0) 0·094 13 1·7(0·4) 2·2(0·9) 0·059 14 1·3(0·5) 1·7(0·8) 0·112 Values are mean(s.d.). * Student's t test. Open in new tab Patients in the placebo group consumed a significantly larger number of analgesic tablets daily than those treated with botulinum toxin (2·3(0·8) versus 1·6(0·6) tablets; P = 0·008). Side-effects and follow-up Botulinum toxin treatment had no systemic or local side-effects. No anal incontinence was recorded in either group before or after treatment. Healing, determined by dissolution of the clot without ulceration through the anoderm, was achieved in all patients. There was no significant difference in time taken for clot resorption between the groups (14·1(3·7) days for botulinum group versus 15·2(4·5) days for placebo; P = 0·228). The time to return to work or usual activity was significantly less in the patients treated with botulinum toxin (6·4(2·1) versus 10·3(3·8) days; P < 0·001). Long-term follow-up was undertaken within 12 months of treatment. Three patients in the botulinum group and four in the placebo group developed recurrent TEH (P = 0·666). The incidence of skin tags at the site of TEH was similar in both groups. Discussion This study has shown that injection of botulinum toxin into the anal sphincter can produce rapid relief of the pain associated with TEH, with no side-effects. The aetiology of TEH is obscure. It is generally accepted that the blood coagulation mechanism is not disturbed14. Denis and colleagues15 reported that in recent perianal thrombosis the erythrocyte aggregation index was higher with a concomitant increase in blood fibrinogen. In another study all the main coagulo pathy indexes, such as antiprothrombin, activated partial thromboplastin, fibrinogen, antithrombin III, cross-linked degradation products, lupus anticoagulant and activated protein C resistance, were normal16. Levels of products of prothrombin degradation were significantly higher than in the control group, confirming that TEH are not the result of local or systemic blood dyscrasia16. The severity of symptoms and timing of presentation help determine the treatment for TEH. Currently accepted treatments include both conservative or surgical approaches, and the decision to operate reflects both patient and physician preference4. Surgical treatment of TEH is reserved for patients presenting within the first 48–72 h of the onset of acute pain, if ulceration or rupture has occurred, or if medical treatment has failed and chronic symptoms persist17. Surgical options range from excision to incision, and evacuation of the thrombus and anal dilatation4. Excision is the therapeutic approach of choice. It can be performed safely, with low recurrence and complication rates, while offering a high level of patient acceptance and satisfaction1. A variety of conservative treatment options have been proposed. Generally, conservative therapy includes a combination of local hygiene measures, Sitzbaths, high-fibre diets or fibre supplements, stool softeners, oral and topical analgesics4, and oral flavonoids18. Hypertonicity of the internal anal sphincter plays an important role in the genesis of pain in anal conditions. In this study patients with TEH had significantly increased MRP compared with healthy subjects. It is not known whether such hypertonicity is an initiating factor or a response to the pain of thrombosis. The use of drugs that reduce sphincter hypertonicity has been shown to be effective in reducing pain7. Injection of botulinum toxin into the internal anal sphincter resulted in pain reduction, reduced analgesic consumption and reduced incapacity. Pain was decreased for up to 9 days after treatment. No local complications or side-effects of botulinum toxin injection were noted. On day 5 there was a significant reduction in sphincter tone in those treated with botulinum toxin. The observed MRP values were similar to those in healthy subjects. The choice of botulinum toxin injection was based on previous evidence of its ability to reduce pain in anal diseases characterized by hypertonicity; it is also associated with a lower incidence of complications and greater compliance than GTN19,20. The exact mechanism of action of botulinum toxin is unclear. Some reports suggest that the toxin binds rapidly and prevents the release of acetylcholine from presynaptic nerve endings21, and of noradrenaline from sympathetic nerve endings22. It produces a relative weakness in the muscle mediated by transitory block of neuromuscular transmission but not completely abolishing voluntary control21. Weakness occurs within a few hours and the transmission of neuromuscular impulses resumes after the growth of new neuromuscular junctions within 3–4 months21. The lower perception of pain in patients treated with botulinum toxin could be the result of MRP reduction23, as well as other mechanisms24. It is also believed that botulinum toxin has an inhibitory effect on nociceptive neurones25–27, and that it reduces the release of mediators associated with pain25. In conclusion, injection of botulinum toxin into the anal sphincter is effective in rapidly controlling pain, which represents the most debilitating symptom of TEH. The absence of side-effects and the simplicity of application makes this an effective alternative option in the conservative treatment of TEH. References 1 Jongen J , Bach S, Stubinger SH, Bock JU. Excision of thrombosed external hemorrhoid under local anesthesia: a retrospective evaluation of 340 patients . Dis Colon Rectum 2003 ; 46 : 1226 – 1231 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Oh C . Acute thrombosed external haemorrhoids . Mt Sinai J Med 1989 ; 56 : 30 – 32 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 3 Gorfine SR . Treatment of benign anal disease with topical nitroglycerin . Dis Colon Rectum 1995 ; 38 : 453 – 456 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Greenspon J , Williams SB, Young HA, Orkin BA. Thrombosed external haemorrhoids: outcome after conservative or surgical management . Dis Colon Rectum 2004 ; 47 : 1493 – 1498 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Shafik A . The pathogenesis of haemorrhoids and their treatment by anorectal bandotomy . J Clin Gastroenterol 1984 ; 6 : 129 – 137 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Cavcic J , Turcic J, Martinac P, Mestrovic T, Mladina R, Pezerovic-Panijan R. Comparison of topically applied 0·2% glyceryl trinitrate ointment, incision and excision in the treatment of perianal thrombosis . Dig Liver Dis 2001 ; 33 : 335 – 340 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Perrotti P , Antropoli C, Molino D, De Stefano G, Antropoli M. Conservative treatment of acute thrombosed external haemorrhoids with topical nifedipine . Dis Colon Rectum 2001 ; 44 : 405 – 409 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Brisinda G , Maria G, Bentivoglio AR, Cassetta E, Gui D, Albanese A. A comparison of injections of botulinum toxin and topical nitroglycerin ointment for the treatment of chronic anal fissure . N Engl J Med 1999 ; 341 : 65 – 69 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Patti R , Almasio PL, Muggeo VM, Buscemi S, Arcara M, Matranga S et al. Improvement of wound healing after hemorrhoidectomy: a double-blind, randomised study of botulinum toxin injection . Dis Colon Rectum 2005 ; 48 : 2173 – 2179 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Patti R , Almasio PL, Arcara M, Sammartano S, Romano P, Fede C et al. Botulinum toxin vs. topical glyceryl trinitrate ointment for pain control in patients undergoing hemorrhoidectomy: a randomized trial . Dis Colon Rectum 2006 ; 49 : 1741 – 1748 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Brisinda G , Maria G, Sganga G, Bentivoglio AR, Albanese A, Castagneto M. Effectiveness of higher doses of botulinum toxin to induce healing in patients with chronic anal fissures . Surgery 2002 ; 131 : 179 – 184 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Huskisson EC . Measurement of pain . J Rheumatol 1982 ; 9 : 768 – 769 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 13 Pescatori M , Anastasio G, Bottini C, Mentasti A. New grading and scoring for anal incontinence. Evaluation of 335 patients . Dis Colon Rectum 1992 ; 35 : 482 – 487 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Ganchrow MI , Bowman HE, Clark JF. Thrombosed haemorrhoids: a clinicopathologic study . Dis Colon Rectum 1971 ; 14 : 331 – 340 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Denis J , Garrigues JM, Trochet JP, Gerentes I, Taccoen A. Rheological parameters in hemorrhoid pathology . Ann Gastroenterol Hepatol 1994 ; 30 : 181 – 184 . Google Scholar OpenURL Placeholder Text WorldCat 16 Gaj F , Trecca A, Suppa M, Sposato M, Coppola A, De Paola G et al. [Hemorrhoidal thrombosis. A clinical and therapeutical study on 22 consecutive patients.] Chir Ital 2006 ; 58 : 219 – 223 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 17 Janicke DM , Pundt MR. Anorectal disorders . Emerg Med Clin North Am 1996 ; 14 : 757 – 788 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Alonso-Coello P , Zhou Q, Martinez-Zapata MJ, Mills E, Heels-Ansdell D, Johanson JF et al. Meta-analysis of flavonoids for the treatment of haemorrhoids . Br J Surg 2006 ; 93 : 909 – 920 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Patti R , Angileri M, Migliore G, Sammartano S, Termine S, Crivello F et al. Effectiveness of contemporary injection of botulinum toxin and topical application of glyceryl trinitrate against postoperative pain after Milligan–Morgan haemorrhoidectomy . Ann It Chir 2006 ; 77 : 503 – 508 . Google Scholar OpenURL Placeholder Text WorldCat 20 Brisinda G , Cadeddu F, Brandara F, Marniga G, Maria G. Randomized clinical trial comparing botulinum toxin injections with 0·2 per cent nitroglycerin ointment for chronic anal fissure . Br J Surg 2007 ; 94 : 162 – 167 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Hallett M . One man's poison—clinical applications of botulinum toxin . N Engl J Med 1999 ; 341 : 118 – 120 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Jones OM , Moore JA, Brading AF, Mortenensen NJ. Botulinum toxin injection inhibits myogenic tone and sympathetic nerve function in the porcine anal sphincter . Colorectal Dis 2003 ; 5 : 552 – 557 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Jones OM . The antinociceptive effects of botulinum toxin therapy for anal fissure are unproven . Dis Colon Rectum 2007 ; 50 : 682 – 683 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Runfola M , Di Mugno M, Baletta A, Magalini SC, Gui D. Antinociceptive effect of botulinum toxin: an added value to chemical sphincterotomy in anal fissure? Dis Colon Rectum 2006 ; 49 : 1078 – 1083 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Cui M , Li Z, You S, Khanijou S, Aoki KR. Mechanism of the antinociceptive effects of subcutaneous Botox: inhibition of peripheral and central nociceptive processing . Naunyn Schmiedebergs Arch Pharmacol 2002 ; 365 : 17 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 26 Welch MJ , Purkiss JR, Foster KA. Sensitivity of embryonic rat dorsal root ganglia neurons to Clostridium botulinum neurotoxins . Toxicon 2000 ; 38 : 245 – 258 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Ishikawa H , Mitsui Y, Yoshitomi T, Mashimoto K, Aoki S, Mukuno K et al. [Presynaptic effects of botulinum toxin type A on the neuronally evoked response of albino and pigmented rabbit iris sphincter and dilator muscles.] Nippon Ganka Gakkai Zasshi 2001 ; 105 : 218 – 222 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Author notes The Editors are satisfied that all authors have contributed significantly to this publication Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

Gerjy, R; Lindhoff-Larson, A; Sjödahl, R; Nyström, P-O

doi: 10.1002/bjs.6379pmid: 18844245

Abstract Background The aim was to assess the feasibility of performing stapled haemorrhoidopexy under local anaesthesia. Methods Fifty-eight patients with haemorrhoid prolapse were randomized to receive local or general anaesthesia. The perianal block was applied immediately peripheral to the external sphincter. Submucosal block was added after applying the purse-string suture. Patients reported average and peak pain daily for 14 days using a visual analogue scale (VAS). They also completed anal symptom questionnaires before the operation and at follow-up. The surgeon assessed the restoration of the anal anatomy 3–6 months after surgery. Results The anal block was sufficient in all patients. The mean accumulated VAS score for average pain was 23·1 in the general anaesthesia group and 29·4 in the local anaesthesia group (P = 0·376); mean peak pain scores were 42·1 and 47·9 respectively (P = 0·537). Mean change in symptom load was also similar between the groups, with score differences of 7·0 in the general anaesthesia group and 6·1 in the local anaesthesia group. No patient had a recurrence of prolapse. Conclusion Perianal local block is easy to apply with a high degree of acceptability among patients. Postoperative pain, restoration of anatomy and symptom resolution were similar to that of stapled haemorrhoidopexy performed under general anaesthesia. Registration number: ISRCTN19930199 (http://www.controlled-trials.com). Introduction Preventing postoperative pain in haemorrhoid surgery is important. Conventional diathermy haemorrhoidectomy is often followed by excessive pain because of the wounds in the particularly sensitive anal canal. The introduction of the procedure for prolapse and haemorrhoids by Longo in 19981 reduced postoperative pain by making the excision wound in the lower rectal mucosa2–9. Traditionally, anal surgery is performed under general or spinal anaesthesia mainly as an inpatient procedure. More recently, several reports have described various forms of local anaesthetic infiltration for the procedure for prolapse and haemorrhoids10–15. Local anaesthetic infiltration for anal surgery aims to facilitate ambulatory surgery and to provide several hours of postoperative pain relief. The present authors have previously described a technique for local anaesthetic block of the ischiorectal fossae16,17 refined from that of Marti18. This method targets the terminal nerves to the anus and sphincters. To allow painless stapled haemorrhoidopexy, the perianal block was supplemented with a submucosal block of the lower rectum17. Although it is known that regional anaesthesia provides pain control on the day of surgery, the aim of this randomized trial was to elucidate the total pain experience in addition to the restoration of anal anatomy and symptom resolution with perianal block. Methods This was an open, prospective, single-centre randomized controlled trial. Of 88 patients invited to participate between April 2004 and November 2005, 30 declined participation owing to preference for either of the anaesthetic methods, and 58 were randomized to either local or general anaesthesia. The randomization was by means of permutated blocks of 20 operations generated from random numbers without any stratification, and equal allocation to the two anaesthetic methods. A research nurse managed the randomization. Randomization was done before the day of operation and patients were informed about their allocation before surgery. Each patient was informed verbally about the trial before consenting to participate. The ethics committee of Linköping University approved the study. Inclusion and exclusion criteria Patients with mucoanal prolapse needing manual reposition or with confirmed prolapse at examination were eligible. There was no age restriction, but patients had to be considered suitable for either general anaesthesia or a local block. Only patients who were considered unsuitable for a general anaesthetic for minor anal surgery were ineligible. Because the perianal block already was in practice for most anal surgery, the anaesthetist requested a local block in some randomized patients thought less fit for general anaesthesia at the time of surgery. Local anaesthetic technique No premedication was given. Patients were placed prone with a pillow under the hips and the legs held together. When they had found a comfortable position that could be endured for about 45 min, they were draped and the perianal block was applied using 40 ml ropivacaine (Narop®; AstraZeneca, Södertälje, Sweden) in a solution of 4·75 mg/ml. A 20-ml syringe was fitted with a 60-mm intramuscular needle and the solution was injected approximately 3 cm from the anal verge through the anococcygeal ligament into the ischiorectal fossae to the level of the levator muscle. A total of 5 ml of the anaesthetic solution was injected while withdrawing the needle. The needle was then directed at 45° anteriolaterally and 5 ml was injected on both sides in the perisphincteric space while the needle was withdrawn. This was repeated in the same fashion anteriorly in the perineum. The anaesthetic block was finished with 5 ml columns of anaesthetic solution on each side of the anus16,17. The onset of anaesthesia was approximately 5 min, with blockage of the branches to the anus from the anococcygeal and pudendal nerves: the inferior haemorrhoidal nerve portion and the anterior sphincteric nerve portion. Complete anaesthesia of the perianal skin and the anal canal ensued with relaxation of the sphincters rendered painless to dilatation. Before introducing the stapler, the submucosa beneath the purse-string suture was infiltrated with 10–15 ml of ropivacaine, 2 mg/ml. This ensured complete painlessness during closure and firing of the stapler. General anaesthetic technique Standard induction of anaesthesia with Propofol-Lipuro® (B. Braun, Stockholm, Sweden), remifentanil (Ultiva®; GlaxoSmithKline, Solna, Sweden) or both was followed by endotracheal intubation and anaesthesia maintained with sevoflurane. Patients were then turned prone for the procedure. Patients with general anaesthesia had the staple line infiltrated with 10–15 ml of ropivacaine (2 mg/ml) for postoperative pain relief after firing and removal of the stapler. Operative technique A PPH03 kit (Ethicon Endo-Surgery, Cincinnati, Ohio, USA) was used in all operations. The circular anal dilator was inserted and secured to the anus with sutures. A 2/0 polypropylene purse-string suture was applied between 3 cm and 4 cm above the dentate line. The stapler was opened to its maximum position and inserted so that the distal anvil was passed beyond the purse-string suture line. The purse-string suture was tied with a single knot. The ends of the thread were pulled through the lateral channels of the stapler head and clamped together externally. The stapler was closed while moderate traction was applied to the purse-string suture. In women, the posterior vaginal wall was ascertained to be free. The stapler was fired and removed. Any bleeding point was secured with suture or diathermy. The height of the final staple line above the dentate line was measured to the nearest half cm. The presence of skin tags before surgery and their excision during surgery were recorded. The operation time included the time taken to apply the perianal block. Postoperative pain and recovery Postoperative pain was the main outcome variable. Before leaving the hospital, patients were given a diary covering the first 14 postoperative days. The patients were asked verbally and in writing to enter a figure that best described the pain experienced during most of the day (daily average pain) and another for the maximum pain experienced at any time of the day (peak pain) using a 10-point visual analogue scale (VAS; 0, no pain; 10, worst imaginable pain). The use of pain medication was recorded in the diary as a dichotomous variable. Patients received 1 g paracetamol four times a day together with 50–150 mg diclofenac per day as needed. No morphine analgesics were provided. Time to recovery was defined as the postoperative day on which patients rated themselves as ‘normal’ or returned to work according to the diary. Patients were seen in the outpatient clinic between 3 and 6 months after surgery to record the final outcome. Self-reported symptoms of haemorrhoids Patients completed a questionnaire of self-reported symptoms associated with the haemorrhoids before the operation and again at the follow-up visit 3–6 months after surgery. Patients rated the frequency of five cardinal symptoms of haemorrhoids: anal pain, anal irritation or pruritus, bleeding, soiling and prolapse of haemorrhoids needing manual reposition at defaecation. Each of the five symptoms was scored according to its frequency: less than once a month (0 points), every month but not every week (1 point), 1–6 days a week (2 points) and every day (3 points). The points for each of the five symptoms were added together for a maximum of 15 points. For each patient, the difference between the preoperative and follow-up scores was taken to represent the treatment effect. Surgeon ratings The surgeon rated the severity of the anatomy before surgery on a 7-point scale (1, minor changes; 7, worst) relative to other patients suitable for surgery. On completion of the operation, the surgeon rated the technical complexity of the operation on a 7-point scale (1, simple; 7, very difficult) relative to experience of other operations for haemorrhoids. Finally, the surgeon rated the success of the operation in restoring normal anal anatomy at follow-up 3–6 months after surgery on a 7-point scale (1, no change; 7, complete success). The operations were performed by two surgeons with previous experience of the local anaesthetic block and the stapled haemorrhoidopexy procedure (P.-O.N. and R.G.). Statistical analysis The pain scores (VAS) were added for each patient from postoperative days 1 to 14 to obtain a total pain score for each patient. The daily average pain and peak pain scores were studied separately. The difference in total pain between groups was analysed using an exact permutation test. The difference in daily pain was analysed using the non-parametric Mann–Whitney U test. Categorical data were tested by Fisher's exact test. Other continuous data were tested with Student's t test. P < 0·050 was considered significant. The authors' previous study of perianal block showed an accumulated difference of 12 VAS points between postoperative days 1 and 14, which was not significant16. For this study, a difference of 21 VAS points equivalent to a difference of 1·5 VAS points per day was considered clinically relevant, and a sample size of 60 patients with equal allocation was planned. Because the first aim was to study postoperative pain with either anaesthetic method, a per-protocol analysis was used. Results Of 88 patients invited to participate in the study, 30 were excluded. Nine had a preference for local perianal block and nine were in favour of general anaesthesia. Ten declined without giving a reason. The anaesthetist decided against a general anaesthesia for the remaining two patients. A total of 58 patients were randomized. Seven were excluded after randomization, two for surgical reasons (one had conventional haemorrhoidectomy because of a narrow anus and one had an anal fissure at the time of surgery), three because the anaesthetist requested local block and two because the patients wanted general anaesthesia at the time of surgery. Therefore 25 patients in the general anaesthesia (GA) group and 26 in the local anaesthesia (LA) group had the allocated intervention (Fig. 1). Including these excluded patients in an intention-to-treat analysis did not alter the pain result. Fig. 1 Open in new tabDownload slide CONSORT flow chart of the trial Surgical results The characteristics of the two groups were comparable (Table 1). Nineteen patients in the GA group and 21 in the LA group stated that the mucoanal prolapse required manual reposition at defaecation, and the surgeons noted a prolapse in all. An external component (anodermal folds and tags) was present in 15 of the GA group and 21 of the LA group, with excision of one or more tags performed in 14 and 17 patients respectively. The mean height of the staple line above the dentate line differed between the groups: 16 mm in the LA group and 19 mm in the GA group (P = 0·019). Because the operation time included the time taken to give the local block, it was significantly longer under perianal block than under general anaesthetic: 30·1 versus 24·7 min respectively, P = 0·004. The local block was complete in all patients. Table 1 Characteristics of 51 patients who had the allocated anaesthesia . General anaesthesia (n = 25) . Perianal block (n = 26) . P† . Sex ratio (M : F) 11 : 14 13 : 13 Age (years)* 54 (29–75) 57 (32–81) Day surgery 22 23 Manual reposition of prolapse  Yes 19 21  No 6 5 Skin tags before surgery 15 21 Excision of skin tags 14 17 Height of staple line above dentate line (cm)* 1·9 (1·5–2·0) 1·6 (0·5–2·5) 0·019 Operation time (min)* 25 (16–36) 30 (14–42) 0·004 . General anaesthesia (n = 25) . Perianal block (n = 26) . P† . Sex ratio (M : F) 11 : 14 13 : 13 Age (years)* 54 (29–75) 57 (32–81) Day surgery 22 23 Manual reposition of prolapse  Yes 19 21  No 6 5 Skin tags before surgery 15 21 Excision of skin tags 14 17 Height of staple line above dentate line (cm)* 1·9 (1·5–2·0) 1·6 (0·5–2·5) 0·019 Operation time (min)* 25 (16–36) 30 (14–42) 0·004 * Values are mean (range). † Student's t test. Open in new tab Table 1 Characteristics of 51 patients who had the allocated anaesthesia . General anaesthesia (n = 25) . Perianal block (n = 26) . P† . Sex ratio (M : F) 11 : 14 13 : 13 Age (years)* 54 (29–75) 57 (32–81) Day surgery 22 23 Manual reposition of prolapse  Yes 19 21  No 6 5 Skin tags before surgery 15 21 Excision of skin tags 14 17 Height of staple line above dentate line (cm)* 1·9 (1·5–2·0) 1·6 (0·5–2·5) 0·019 Operation time (min)* 25 (16–36) 30 (14–42) 0·004 . General anaesthesia (n = 25) . Perianal block (n = 26) . P† . Sex ratio (M : F) 11 : 14 13 : 13 Age (years)* 54 (29–75) 57 (32–81) Day surgery 22 23 Manual reposition of prolapse  Yes 19 21  No 6 5 Skin tags before surgery 15 21 Excision of skin tags 14 17 Height of staple line above dentate line (cm)* 1·9 (1·5–2·0) 1·6 (0·5–2·5) 0·019 Operation time (min)* 25 (16–36) 30 (14–42) 0·004 * Values are mean (range). † Student's t test. Open in new tab Forty-five patients were discharged from hospital a few hours after surgery. Two patients in the GA group were admitted overnight for difficulty voiding and one because of postoperative nausea and vomiting. Two other patients in the GA group were treated for urinary retention after discharge but not admitted. In the LA group, one patient was admitted overnight for a reoperation because of postoperative bleeding. One stayed overnight because of postoperative nausea and vomiting and one because of excessive pain. Postoperative pain and recovery The patient diary was returned by 21 of 25 patients in the GA group and 23 of 26 patients in the LA group. The intensity of the postoperative pain was assessed from the diaries. The mean sum of VAS scores for daily average pain was 23·1 in the GA group and 29·4 in the LA group (P = 0·376). Mean scores for peak pain were 42·1 and 47·9 respectively (P = 0·537). The resolution of daily average pain and peak pain is shown in Fig. 2. The use of pain medication was similar between the groups (Fig. 3). Eight patients in the GA group but only one in the LA group returned to work or considered themselves ‘normal’ the first day after the operation (P = 0·004). However, there was no significant difference between the groups for postoperative days 2 to 14 in this respect (Fig. 4). Fig. 2 Open in new tabDownload slide Resolution of a average pain and b peak pain according to visual analogue scale (VAS) scores given by patients. There was no significant difference between the groups Fig. 3 Open in new tabDownload slide The proportion of patients still taking pain medication on each day after surgery. There was no significant difference between the groups Fig. 4 Open in new tabDownload slide The proportion of patients who rated themselves as ‘normal’ or returned to work on each day after surgery. On day 1, the difference was significant (P = 0·004); thereafter there were no significant differences Surgeon ratings Preoperative anatomy, operation complexity and postoperative success were rated similarly for the two groups (Table 2). Two patients declined follow-up examination 3–6 months after the operation. None of the 49 examined patients had a recurrence of prolapse, but nine in the GA group and 19 in the LA group had one or more residual skin tags. Table 2 Surgeon's anatomy scores and patients' self-reported symptom scores . General anaesthesia (n = 25) . Perianal block (n = 26) . Preoperative anatomical score* 4 (2–7) 4 (3–7) Operative complexity* 5 (2–7) 4 (1–7) Postoperative success rate* 6 (4–7) 6 (4–7) Preoperative symptom score† 8·8(3·0) 8·3(2·4) Postoperative symptom score† 1·8(2·5) 2·2(2·3) Preoperative prolapse according to the surgeon 25 26 . General anaesthesia (n = 25) . Perianal block (n = 26) . Preoperative anatomical score* 4 (2–7) 4 (3–7) Operative complexity* 5 (2–7) 4 (1–7) Postoperative success rate* 6 (4–7) 6 (4–7) Preoperative symptom score† 8·8(3·0) 8·3(2·4) Postoperative symptom score† 1·8(2·5) 2·2(2·3) Preoperative prolapse according to the surgeon 25 26 * Values are mean of surgeons' rating on a 7-point scale (range). † Values are mean(s.d.). There were no significant differences between the groups. Open in new tab Table 2 Surgeon's anatomy scores and patients' self-reported symptom scores . General anaesthesia (n = 25) . Perianal block (n = 26) . Preoperative anatomical score* 4 (2–7) 4 (3–7) Operative complexity* 5 (2–7) 4 (1–7) Postoperative success rate* 6 (4–7) 6 (4–7) Preoperative symptom score† 8·8(3·0) 8·3(2·4) Postoperative symptom score† 1·8(2·5) 2·2(2·3) Preoperative prolapse according to the surgeon 25 26 . General anaesthesia (n = 25) . Perianal block (n = 26) . Preoperative anatomical score* 4 (2–7) 4 (3–7) Operative complexity* 5 (2–7) 4 (1–7) Postoperative success rate* 6 (4–7) 6 (4–7) Preoperative symptom score† 8·8(3·0) 8·3(2·4) Postoperative symptom score† 1·8(2·5) 2·2(2·3) Preoperative prolapse according to the surgeon 25 26 * Values are mean of surgeons' rating on a 7-point scale (range). † Values are mean(s.d.). There were no significant differences between the groups. Open in new tab Symptom resolution In the GA group, the mean(s.d.) symptom score was 8·8(3·0) before and 1·8(2·5) after surgery, a mean change of 7·0 points. In the LA group, the scores were 8·3(2·5) before and 2·2(2·3) after surgery, a mean change of 6·1 points. Resolution of symptoms therefore did not differ significantly between the groups. Complications One patient in the LA group had a reoperation the same day for postoperative bleeding. Four patients in the GA group were treated for urinary retention. Two patients in the LA group and one in the GA group had excessive postoperative pain. Of these, one in the LA group developed acute anal fissure and the other in the LA group had anal spasm that resolved with diltiazem ointment. The pain of the patient in the GA group resolved spontaneously. At follow-up, three patients in the LA group and two in the GA group complained of mild faecal urgency. Discussion This randomized study comparing local block with general anaesthesia for haemorrhoidopexy showed statistically similar pain scores for patients in both groups, although numerically the scores were slightly lower after general anaesthesia. This compares with the authors' previous (non-randomized) study, in which pain in both groups was statistically similar but with a larger numerical difference in favour of perianal block. Pain after haemorrhoid surgery varies greatly; the coefficient of variation exceeds 50 per cent for VAS scores on any postoperative day as well as for the total pain experienced over 14 days. This wide variation may influence the result in small studies. It seemed reasonable that a clinically relevant difference would be an average daily difference of 1·5 VAS points or 21 VAS points accumulated over 14 days. A post hoc analysis of the results showed that the difference in pain scores between perianal local block and general anaesthesia would have been less with a power of 85 per cent and P = 0·050. Excision of skin tags19 and a lower staple line20 may account for increased pain, although excision of skin tags was not associated with increased pain in the authors' previous study21 or in the present study. However, there was a significant inverse correlation between the height of the staple line and total pain. In another study, patients with a staple line 22 mm above the dentate line required less pain management and returned to work earlier21. In the present study, the mean height of the staple line in both groups was less (16 mm and 19 mm in the LA and GA groups respectively). Although the aim for the height of the staple line should be 20 mm, the reason is not primarily pain control: a higher staple line inadequately controls haemorrhoidal prolapse and invites recurrence below the staple line. A high staple line was originally proposed by Longo to reduce pain22, but it is a possible explanation of the worse prolapse control after haemorrhoidopexy compared with conventional haemorrhoid excision23,24. The reason for the lower staple line in the LA group is not clear. It is possible that the submucosal deposition of local anaesthetic to block the site of the stapler closure may have changed conditions for the final staple line. However, in the authors' previous study of perianal block, the mean staple line height was 20 mm16. Stapled haemorrhoidopexy is predominantly performed under general anaesthesia. Gabrielli and colleagues13 described a posterior ischiorectal block but supplemented it with an anaesthetic injection into the intersphincteric space, which required conscious sedation. Ong and co-workers14 also described a technique for local anal infiltration in the intersphincteric space, injecting directly into the anal sphincters, levator ani muscle and surrounding perianal skin under intravenous sedation consisting of 3–5 mg midazolam and 70–100 µg of fentanyl. Esser and colleagues12 infiltrated the perianal skin and submucosa in all four quadrants with 0·5 per cent lidocaine with 1 : 200 000 epinephrine solution under conscious sedation with midazolam, fentanyl or both. Injections into the sphincters, the intersphincteric space and the anoderm are painful and best avoided. When injecting the anaesthetic solution into the perisphincteric space, the main discomfort is the sting in the perianal skin. Delikoukos and co-workers11 described a technique similar to Marti's posterior ischiorectal block, with deposition of the anaesthetic solution perisphincterically in the infralevator space. They proposed using the index finger inside the anus to guide the syringe. This method should block the inferior haemorrhoidal branches of the pudendal nerve but does not reach the anterior sphincteric branches of the same nerve. The present authors have achieved this relaxation by blocking the anterior sphincteric branches with a similar injection anteriorly. Imbelloni and colleagues25 used a nerve stimulator to administer a bilateral pudendal nerve block, with better pain relief during the first 24 h after surgery. None of the studies described above recorded pain beyond the first 24 h. The present study examined the potential benefit at later stages of recovery. An infralevator block alone is often insufficient because the rectal wall at the levator may have unblocked the visceral nerve supply from above. Adding a small submucosal block under the purse-string suture provided complete pain control for closure and firing of the stapler16,17. The anaesthetic injection of the submucosa should be performed after insertion of the purse-string suture to avoid narrowing the view. Perianal local block has several advantages: the surgeon can apply it as part of the procedure in approximately 5 min, patients operated on in the prone position can place themselves in a comfortable position on the table, more patients can be treated, surveillance after anaesthesia is easier, and patients are without pain for several hours and can expect to return home pain-free. However, this and the previous study showed that local block did not ameliorate pain on subsequent days or enhance recovery. Nevertheless, an alternative anaesthetic method allows patients a choice26. Stapled haemorrhoidopexy under perianal local anaesthetic block is effective, convenient and acceptable to most patients compared with the same procedure under general anaesthesia, providing a similar clinical result. Acknowledgements The authors thank Olle Eriksson at the Department of Mathematics, Linköping University, for assistance with the statistical analysis. References 1 Longo A . Treatment of hemorrhoidal disease by reduction of mucosa and hemorrhoidal prolapse with a circular suturing device: a new procedure . In Proceedings of the 6th World Congress of Endoscopic Surgery and 6th International Congress of the European Association for Endoscopic Surgery, Rome, 3–6 Jun 1998. Mundozzi Editore: Bologna, 1998 ; 777 – 784 . 2 Cheetham MJ , Cohen CR, Kamm MA, Phillips RK. A randomized, controlled trial of diathermy hemorrhoidectomy vs. stapled hemorrhoidectomy in an intended day-care setting with longer-term follow-up . Dis Colon Rectum 2003 ; 46 : 491 – 497 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Correa-Rovelo J , Tellez O, Obregon L, Miranda-Gomez A, Moran S. Stapled rectal mucosectomy vs. closed hemorrhoidectomy: a randomized, clinical trial . Dis Colon Rectum 2002 ; 45 : 1367 – 1374 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Ganio E , Altomare DF, Gabrielli F, Milito G, Canuti S. Prospective randomized multicentre trial comparing stapled with open haemorrhoidectomy . Br J Surg 2001 ; 88 : 669 – 674 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Kairaluoma M , Nuorva K, Kellokumpu I. Day-case stapled (circular) vs. diathermy hemorrhoidectomy: a randomized, controlled trial evaluating surgical and functional outcome . Dis Colon Rectum 2003 ; 46 : 93 – 99 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Mehigan BJ , Monson JR, Hartley JE. Stapling procedure for haemorrhoids versus Milligan-Morgan haemorrhoidectomy: randomised controlled trial . Lancet 2000 ; 355 : 782 – 785 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Ortiz H , Marzo J, Armendariz P. Randomized clinical trial of stapled haemorrhoidopexy versus conventional diathermy haemorrhoidectomy . Br J Surg 2002 ; 89 : 1376 – 1381 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Rowsell M , Bello M, Hemingway DM. Circumferential mucosectomy (stapled haemorrhoidectomy) versus conventional haemorrhoidectomy: randomised controlled trial . Lancet 2000 ; 355 : 779 – 781 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Senagore AJ , Singer M, Abcarian H, Fleshman J, Corman M, Wexner S et al. A prospective, randomized, controlled multicenter trial comparing stapled hemorrhoidopexy and Ferguson hemorrhoidectomy: perioperative and one-year results . Dis Colon Rectum 2004 ; 47 : 1824 – 1836 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Ho KS , Eu KW, Heah SM, Seow-Choen F, Chan YW. Randomized clinical trial of haemorrhoidectomy under a mixture of local anaesthesia versus general anaesthesia . Br J Surg 2000 ; 87 : 410 – 413 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Delikoukos S , Zacharoulis D, Hatzitheofilou C. Local posterior perianal block for proctologic surgery . Int Surg 2006 ; 91 : 348 – 351 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 12 Esser S , Khubchandani I, Rakhmanine M. Stapled hemorrhoidectomy with local anesthesia can be performed safely and cost-efficiently . Dis Colon Rectum 2004 ; 47 : 1164 – 1169 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Gabrielli F , Chiarelli M, Cioffi U, Guttadauro A, De Simone M, Di Mauro P et al. Day surgery for mucosal-hemorrhoidal prolapse using a circular stapler and modified regional anesthesia . Dis Colon Rectum 2001 ; 44 : 842 – 844 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Ong CH , Chee Boon Foo E, Keng V. Ambulatory circular stapled haemorrhoidectomy under local anaesthesia versus circular stapled haemorrhoidectomy under regional anaesthesia . ANZ J Surg 2005 ; 75 : 184 – 186 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Vinson-Bonnet B , Coltat JC, Fingerhut A, Bonnet F. Local infiltration with ropivacaine improves immediate postoperative pain control after hemorrhoidal surgery . Dis Colon Rectum 2002 ; 45 : 104 – 108 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Gerjy R , Derwinger K, Nyström PO. Perianal local block for stapled anopexy . Dis Colon Rectum 2006 ; 49 : 1914 – 1921 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Nyström PO , Derwinger K, Gerjy R. Local perianal block for anal surgery . Tech Coloproctol 2004 ; 8 : 23 – 26 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Marti MC . [Loco-regional anesthesia in proctological surgery] . Ann Chir 1993 ; 47 : 250 – 255 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 19 Raahave D , Jepsen LV, Pedersen IK. Primary and repeated stapled hemorrhoidopexy for prolapsing hemorrhoids : follow-up to five years . Dis Colon Rectum 2008 ; 51 : 334 – 341 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Plocek MD , Kondylis LA, Duhan-Floyd N, Reilly JC, Geisler DP, Kondylis PD. Hemorrhoidopexy staple line height predicts return to work . Dis Colon Rectum 2006 ; 49 : 1905 – 1909 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Gerjy R , Nyström PO. Excision of residual skin tags during stapled anopexy does not increase postoperative pain . Colorectal Dis 2007 ; 9 : 754 – 757 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Longo A . Stapled anopexy and stapled hemorrhoidectomy: two opposite concepts and procedures . Dis Colon Rectum 2002 ; 45 : 571 – 572 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 23 Jayaraman S , Colquhoun PH, Malthaner RA. Stapled versus conventional surgery for hemorrhoids . Cochrane Database Syst Rev 2006 ; ( 4 ): CD005393 . Google Scholar OpenURL Placeholder Text WorldCat 24 Shao WJ , Li GC, Zhang ZH, Yang BL, Sun GD, Chen YQ. Systematic review and meta-analysis of randomized controlled trials comparing stapled haemorrhoidopexy with conventional haemorrhoidectomy . Br J Surg 2008 ; 95 : 147 – 160 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Imbelloni LE , Vieira EM, Gouveia MA, Netinho JG, Spirandelli LD, Cordeiro JA. Pudendal block with bupivacaine for postoperative pain relief . Dis Colon Rectum 2007 ; 50 : 1656 – 1661 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Kushwaha R , Hutchings W, Davies C, Rao NG. Randomized clinical trial comparing day-care open haemorrhoidectomy under local versus general anaesthesia . Br J Surg 2008 ; 95 : 555 – 563 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

van Wely, B J; Smidt, M L; de Kievit, I M; Wauters, C A P; Strobbe, L J A

doi: 10.1002/bjs.6348pmid: 18844270

Abstract Background The clinical significance of false-negative axillary sentinel lymph node (SLN) biopsy has yet to be established. The aim of this study was to assess the axillary recurrence rate and false-negative rate, to identify prognostic factors and to monitor survival. Methods A prospective registry of sentinel lymph node biopsy (SLNB) procedures performed between 1998 and 2004 was analysed. All nodes retrieved were examined by haematoxylin and eosin and immunohistochemical staining. Further surgical treatment was performed only for positive SLNs. Adjuvant treatment was given according to Dutch guidelines. Results Of 592 patients, 392 had a negative SLNB. After a median follow-up of 65 months, 11 patients developed axillary recurrence (2·8 per cent). Ten of these patients were primarily treated by simple mastectomy and therefore had no external-beam radiation therapy; no further prognostic factors could be identified. The false-negative rate was 6·9 per cent. The median time from SLNB to detection of the axillary recurrence was 27 months. The median follow-up after recurrence was 35 months, and one patient developed systemic disease. Pathology revision revealed two previously undetected micrometastases. Conclusion Axillary recurrence and false-negative rates after SLNB increase with longer follow-up. Introduction Sentinel lymph node biopsy (SLNB) is widely accepted as a minimally invasive technique to stage the axilla in node-negative breast cancer. The sensitivity and specificity of SLNB has been reported, and two meta-analyses have been conducted1,2. The false-negative rates (negative sentinel lymph nodes (SLN) but tumour-positive non-SLN in the completion axillary lymph node dissection (ALND) specimen) range from 5·1 to 7·3 per cent. Many authors have conducted follow-up studies after negative SLNB in patients who did not have ALND to determine the significance of these false-negative test results3–23. No prognostic factors for developing axillary recurrences after false-negative SLNB have been identified owing to the small numbers of patients involved. Nor has the impact of axillary recurrence on survival been investigated yet. The aim of this study was to explore these concerns in a prospectively followed population of patients with negative SLNB. Methods A prospective registry of all SLNB procedures for breast cancer done between January 1998 and December 2004 was analysed. Sentinel lymph node biopsy As recommended by the Dutch guidelines, 60 MBq technetium-99m nanocolloid was used as radioactive tracer. Peritumoral lymphatic mapping was performed using 2 ml of patent blue dye (Patent Blue V; Guerbet, Aulnay-sous-Bois, France). Lymphoscintigraphy was performed the day before SLNB, and a hand-held gamma probe was used to identify hot nodes during the procedure. The blue dye was injected around the tumour before the procedure (after previous lumpectomy it was injected around the scar); after a brief massage in the direction of the axilla, SLNB was performed. All axillary lymph nodes that were hot or blue were harvested. When suspect nodes were found during the procedure that were not blue or hot, they were excised and examined as non-SLNs. Extra-axillary lymph nodes were not harvested. Pathological examination All lymph nodes were serially sectioned at 500 µm intervals, stained with haematoxylin and eosin and immunohistochemically with CAM 5.2 at all levels. Further surgical treatment After an initial validation phase including 20 procedures, ALND was performed only when SLN contained tumour. These included micrometastatic disease (less than 2 mm) and isolated tumour cells (less than 0·2 mm). No further axillary surgery was done after a negative SLNB. Adjuvant treatment and follow-up Adjuvant treatment was given according to contemporary Dutch guidelines24. After initial treatment, patients visited the outpatient clinic 6-monthly. Mammography was done yearly, but ultrasonographic examination of the axilla was not routine; additional tests were only performed if indicated. For the purposes of this study, axillary recurrence was defined as the development of positive axillary nodes in a patient who had a negative SLNB. In SLNB (after the validation phase) a false-negative result also included patients in whom tumour-positive non-SLN were harvested in the absence of a tumour-positive SLN25. Results A total of 592 patients with clinically node-negative breast cancer were registered. Of this group, 392 had a negative SLNB. Four of these had a non-SLN harvested during the procedure that was tumour-positive in the absence of a positive SLN. Thus there were 388 patients without full axillary clearance for follow-up and analysis. The median follow-up was 65 (range 33–123) months after SLNB. In this interval 11 patients developed an axillary recurrence (2·8 per cent). The false-negative SLNB rate (including tumour-positive non-SLN) was 6·9 per cent. The median delay from SLNB to the detection of the axillary recurrence was 27 (range 4–63) months. None of the patients with axillary recurrence had systemic disease on restaging. All had ALND, and a median of 6 (range 1–17) tumour-positive lymph nodes were found. All patients had adjuvant treatment after ALND: ten had systemic treatment, and eight received radiation therapy. The median follow-up after recurrence was 35 (range 13–73) months. In this time one patient developed systemic disease, but none has died. Analysis of axillary recurrences Patient and tumour characteristics The median age of the 11 patients was 58 (range 35–75) years. All original tumours were invasive ductal carcinomas. Seven were T1 tumours; the remaining four were T2. Eight were located in the upper outer quadrant, one in the upper inner quadrant and two in the lower inner quadrant. Seven tumours were classified as Bloom–Richardson grade III, three were grade II and one was grade I. A total of nine tumours were oestrogen-receptor positive; three were progesterone-receptor positive as well. Sentinel lymph node biopsy procedure Review of the scintigraphy showed that all but one of the scintigraphically identified SLNs were harvested. This was a parasternal SLN that was not harvested or sought during the SLNB procedure; only the axillary nodes were harvested. A median of 1 (range 1–3) SLN was harvested. In ten of 11 patients the SLN was identified by both radioactive tracer and blue dye. Pathology revision at time of detection of the axillary recurrence revealed previously undetected micrometastases in two SLN; in one after further sectioning the SLN. Initial treatment (surgical and adjuvant) Ten of 11 SLN-negative patients were treated by simple mastectomy; one had breast-conserving surgery and had radiation therapy as part of the initial treatment. This is in contrast to the whole group studied, in which about two-thirds of patients had breast-conserving treatment. Three patients treated by simple mastectomy received adjuvant systemic treatment based on the characteristics of the primary tumour. Discussion The axillary recurrence rate in the present study was 2·8 per cent after a median follow-up of over 5 years. Axillary recurrence continued to occur up to 63 months, showing how important the length of follow-up is in investigating axillary recurrence. The axillary recurrence rate after negative SLNB varies widely. Published follow-up results have reported rates ranging from 0 to 2·7 per cent3–23 after follow-up varying from 16 to 79 months. Some authors have shown results similar to the present findings, with an axillary recurrence rate of 2·7 per cent after a median of 65 months6. A recent update of a randomized controlled trial, however, showed an axillary recurrence rate of 0·6 per cent after a median follow-up of 79 months23. The same trial reported a false-negative rate of 8·7 per cent in the validation phase when all SLNB were followed by completion ALND. In this respect the role of micrometastatic disease in SLN is important. A meta-analysis of non-SLN metastases associated with micrometastatic SLN showed that 10–15 per cent of patients with SLN micrometastasis had non-SLN involvement26. The exact importance of micrometastatic disease in the SLN remains uncertain. Trials conducted to investigate this aspect are required27. A retrospective study with 20 years of follow-up, however, stated that patients with micrometastasis had a significantly worse prognosis than those with negative nodes28. In the present study, one of the patients who developed an axillary recurrence actually had micrometastatic involvement of the SLN at pathological review. The false-negative rate is important in continued validation of SLNB25. Most authors do not mention the rate of tumour-positive non-SLN, thus making it impossible to determine false-negative rates. When this was calculated from the available results, the false-negative rates varied from 1·2 to 4·9 per cent7,8,15,16,20. In the present population the false-negative rate was 6·9 per cent. Another study objective was to identify prognostic factors predisposing to axillary recurrence in these patients. Apart from a high proportion of patients with grade III tumours, tumour and patient characteristics were similar in the recurrence group to those in all SLN-negative patients. However, the group was too small to calculate any statistical significance. For this reason, analysis of a larger group of axillary recurrences after negative SLNB is important. All but one of the patients in this study who developed an axillary recurrence had mastectomy. Therefore ten of the patients did not receive external-beam radiation therapy as part of their initial treatment. Radiotherapy to the axillary tail of the breast includes the caudal part of the axilla. Irradiation of this part of the axilla might be enough to affect the recurrence rate. This could explain the very low axillary recurrence rate found in the randomized controlled trial mentioned earlier, in which all patients had breast conservation treatment (with radiotherapy)23. Although the number of patients was small, the survival after axillary recurrence following node-negative SLNB was good after a median follow-up of 35 months. When comparing patients with recurrence after negative SLNB, and those with recurrence after ALND29, the outcome seemed to be better in the SLNB group: one of 11 patients after median of 35 months compared with 50 per cent after a median of 23 months29. References 1 Kim T , Giuliano AE, Lyman GH. Lymphatic mapping and sentinel lymph node biopsy in early-stage breast carcinoma: a metaanalysis . Cancer 2006 ; 106 : 4 – 16 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Miltenburg DM , Miller C, Karamlou TB, Brunicardi FC. Meta-analysis of sentinel lymph node biopsy in breast cancer . J Surg Res 1999 ; 84 : 138 – 142 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Badgwell BD , Povoski SP, Abdessalam SF, Young DC, Farrar WB, Walker MJ et al. Patterns of recurrence after sentinel lymph node biopsy for breast cancer . Ann Surg Oncol 2003 ; 10 : 376 – 380 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Heuts EM , van der Ent FW, Hulsewe KW, Heeren PA, Hoofwijk AG. Incidence of axillary recurrence in 344 sentinel node negative breast cancer patients after intermediate follow-up. A prospective study into the accuracy of sentinel node biopsy in breast cancer patients . Acta Chir Belg 2007 ; 107 : 279 – 283 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Chung MA , Steinhoff MM, Cady B. Clinical axillary recurrence in breast cancer patients after a negative sentinel node biopsy . Am J Surg 2002 ; 184 : 310 – 314 . Google Scholar Crossref Search ADS PubMed WorldCat 6 de Kanter AY , Menke-Pluymers MM, Wouters MW, Burgmans I, van Geel AN, Eggermont AM. 5-Year follow-up of sentinel node negative breast cancer patients . Eur J Surg Oncol 2006 ; 32 : 282 – 286 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Estourgie SH , Nieweg OE, Valdes Olmos RA, Rutgers EJ, Peterse JL, Kroon BB. Eight false negative sentinel node procedures in breast cancer: what went wrong? Eur J Surg Oncol 2003 ; 29 : 336 – 340 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Jeruss JS , Winchester DJ, Sener SF, Brinkmann EM, Bilimoria MM, Barrera E et al. Axillary recurrence after sentinel node biopsy . Ann Surg Oncol 2005 ; 12 : 34 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Kokke MC , Jannink I, Barneveld PC, van der Linden JC, Gelderman WA, Wissing JC et al. Incidence of axillary recurrence in 113 sentinel node negative breast cancer patients: a 3-year follow-up study . Eur J Surg Oncol 2005 ; 31 : 221 – 225 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Langer I , Marti WR, Guller U, Moch H, Harder F, Oertli D et al. Axillary recurrence rate in breast cancer patients with negative sentinel lymph node (SLN) or SLN micrometastases: prospective analysis of 150 patients after SLN biopsy . Ann Surg 2005 ; 241 : 152 – 158 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Palesty JA , Foster JM, Hurd TC, Watroba N, Rezaishiraz H, Edge SB. Axillary recurrence in women with a negative sentinel lymph node and no axillary dissection in breast cancer . J Surg Oncol 2006 ; 93 : 129 – 132 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Reitsamer R , Peintinger F, Prokop E, Rettenbacher L, Menzel C. 200 Sentinel lymph node biopsies without axillary lymph node dissection—no axillary recurrences after a 3-year follow-up . Br J Cancer 2004 ; 90 : 1551 – 1554 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Rosing DK , Dauphine CE, Vargas MP, Gonzalez K, Burla M, Kaufmann P et al. Axillary regional recurrence after sentinel lymph node biopsy for breast cancer . Am Surg 2006 ; 72 : 939 – 942 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Roumen RM , Kuijt GP, Liem IH, van Beek MW. Treatment of 100 patients with sentinel node-negative breast cancer without further axillary dissection . Br J Surg 2001 ; 88 : 1639 – 1643 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Sanjuan A , Vidal-Sicart S, Zanon G, Pahisa J, Velasco M, Fernandez PL et al. Clinical axillary recurrence after sentinel node biopsy in breast cancer: a follow-up study of 220 patients . Eur J Nucl Med Mol Imaging 2005 ; 32 : 932 – 936 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Schrenk P , Hatzl-Griesenhofer M, Shamiyeh A, Waynad W. Follow-up of sentinel node negative breast cancer patients without axillary lymph node dissection . J Surg Oncol 2001 ; 77 : 165 – 170 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Smidt ML , Janssen CM, Kuster DM, Bruggink ED, Strobbe LJ. Axillary recurrence after a negative sentinel node biopsy for breast cancer: incidence and clinical significance . Ann Surg Oncol 2005 ; 12 : 29 – 33 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Swenson KK , Mahipal A, Nissen MJ, Tuttle TM, Heaton K, Lally RM et al. Axillary disease recurrence after sentinel lymph node dissection for breast carcinoma . Cancer 2005 ; 104 : 1834 – 1839 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Torrenga H , Fabry H, van der Slip JR, van Diest PJ, Pijpers R, Meijer S. Omitting axillary lymph node dissection in sentinel node negative breast cancer patients is safe: a long term follow-up analysis . J Surg Oncol 2004 ; 88 : 4 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 20 van der Vegt B , Doting MH, Jager PL, Wesseling J, de Vries J. Axillary recurrence after sentinel lymph node biopsy . Eur J Surg Oncol 2004 ; 30 : 715 – 720 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Veronesi U , Galimberti V, Mariani L, Gatti G, Paganelli G, Viale G et al. Sentinel node biopsy in breast cancer: early results in 953 patients with negative sentinel node biopsy and no axillary dissection . Eur J Cancer 2005 ; 41 : 231 – 237 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Zavagno G , Carcoforo P, Franchini Z, Renier M, Barutta L, De Salvo GL et al. Axillary recurrence after negative sentinel lymph node biopsy without axillary dissection: a study on 479 breast cancer patients . Eur J Surg Oncol 2005 ; 31 : 715 – 720 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Veronesi U , Paganelli G, Viale G, Luini A, Zurrida S, Galimberti V et al. Sentinel-lymph-node biopsy as a staging procedure in breast cancer: update of a randomised controlled study . Lancet Oncol 2006 ; 7 : 983 – 990 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Oncoline Oncological Guidelines. www.oncoline.nl (accessed 1 September 2008). 25 Estourgie SH , Nieweg OE, Rutgers EJ, Kroon BB. What is a false-negative result for sentinel node procedures in breast cancer? J Surg Oncol 2003 ; 82 : 141 – 142 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Cserni G , Gregori D, Merletti F, Sapino A, Mano MP, Ponti A et al. Meta-analysis of non-sentinel node metastases associated with micrometastatic sentinel nodes in breast cancer . Br J Surg 2004 ; 91 : 1245 – 1252 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Rutgers EJ . Sentinel node micrometastasis in breast cancer . Br J Surg 2004 ; 91 : 1241 – 1242 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Tan LK , Giri D, Hummer AJ, Panageas KS, Brogi E, Norton L et al. Occult axillary node metastases in breast cancer are prognostically significant: results in 368 node-negative patients with 20-year follow-up . J Clin Oncol 2008 ; 26 : 1803 – 1809 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Newman LA , Hunt KK, Buchholz T, Kuerer HM, Vlastos G, Mirza N et al. Presentation, management and outcome of axillary recurrence from breast cancer . Am J Surg 2000 ; 180 : 252 – 256 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

Rusby, J E; Brachtel, E F; Othus, M; Michaelson, J S; Koerner, F C; Smith, B L

doi: 10.1002/bjs.6349pmid: 18844271

Abstract Background This prospective study aimed to build a predictive model using preoperative information to aid selection for nipple-sparing mastectomy. Methods Two hundred consecutive skin-sparing mastectomy specimens without overt nipple involvement were evaluated. Demographic, preoperative pathology and imaging information was collected. Nipple specimens (2 × 2 × 2 cm) were sectioned at 3-mm intervals. Haematoxylin and eosin-stained slides were examined by a breast pathologist for involvement by tumour. Logistic regression analyses of 65 therapeutic procedures identified factors associated with occult involvement and created a predictive model. This was tested on specimens from a further 65 therapeutic procedures. Results Occult nipple involvement was noted in 32 (24·6 per cent) of 130 mastectomy specimens. In the training set, imaging diameter of the lesion and its distance from the nipple predicted nipple involvement on univariable analysis (P = 0·011 and P = 0·014 respectively). The multivariable logistic regression model was validated in the test set. The areas under the receiver–operating characteristic curve were 0·824 and 0·709 for the training and test sets respectively. Conclusion Three-quarters of women undergoing mastectomy did not have occult nipple involvement. A clinical tool including tumour size and distance from the nipple has been developed to improve patient selection for nipple-sparing mastectomy. Introduction More than half of all women with breast carcinoma undergo breast-conserving surgery1,2. The remainder require or request mastectomy. Increasing numbers of women are undergoing testing for genetic predisposition to breast cancer and are opting for prophylactic mastectomy to reduce their risk3. Surgical techniques for mastectomy are evolving continually and many women are able to undergo less disfiguring surgery than modified radical mastectomy, which was the standard procedure for many years. Improved techniques for immediate breast reconstruction have led to the development of skin-sparing mastectomy in which only the nipple and areola are excised, preserving the remainder of the native skin envelope4. The oncological safety of skin-sparing mastectomy is now generally accepted in women undergoing mastectomy for breast cancer5–7. Almost 90 per cent of women choose nipple reconstruction after mastectomy and breast reconstruction8. This highlights the importance of the nipple–areola complex in cosmetic outcome, and suggests that nipple-sparing mastectomy may offer a further improvement in patient satisfaction9,10. However, nipple-sparing mastectomy has been adopted in only a few centres11–14. The two main concerns cited by surgeons are the risk of occult involvement of the nipple by tumour and the potential complication of loss of viability of the nipple due to damage to its blood supply during mastectomy15,16. Several studies have examined the incidence of nipple involvement in patients with breast cancer. Most series were retrospective and, owing to heterogeneous study populations and different methods of assessment, reported a wide range of nipple involvement from 0 to 58 per cent. Furthermore, many of these studies were carried out in the prescreening era of larger invasive breast cancers and less in situ disease. Despite this uncertainty, some women opt for nipple-sparing mastectomy and a preoperative prediction of the risk of occult nipple involvement would be of value to them. This prospective study aimed to build a predictive model using preoperative information to aid selection for nipple-sparing mastectomy. Methods Consecutive women undergoing skin-sparing, nipple-sacrificing mastectomy between 1 January and 31 December 2006 were studied. Institutional review board approval was obtained to collect patient information and discarded tissue. Patient information Demographic information was gathered prospectively from the electronic medical record at the time of mastectomy. This included age, menopausal status and hormone use, number of first-degree relatives with breast or ovarian cancer, and past breast cancer history. Patients with inflammatory breast cancer, those in whom the nipple was clinically involved by invasive carcinoma and those with Paget's disease of the nipple were excluded. Radiological information Tumour size and distance from the base of the nipple were measured on digital images by an observer who was blinded to nipple involvement. Lesion diameter was defined as the maximum extent of architectural distortion or malignant microcalcification on mammography, enhancement on magnetic resonance imaging (MRI), or hypoechogenicity on ultrasonography. Distance from the nipple was defined as the shortest distance from the base of the nipple to the closest edge of the tumour. If more than one imaging modality had been used to assess the tumour, the largest measured tumour diameter and the shortest distance from the nipple were recorded. In 39·2 per cent of cases both lesion diameter and distance from the nipple were derived from mammography, 20·8 per cent from MRI, 0·8 per cent from ultrasonography and 13·1 per cent from a combination of two methods (predominantly mammography and MRI). It was not possible to acquire data on both lesion diameter and distance from the nipple in 26·2 per cent of patients for two reasons. First, as it is a tertiary referral centre, many patients who undergo mastectomy in this hospital have had investigations elsewhere. Second, although the lesion diameter was measured at ultrasonography, it was unusual for the distance from the nipple to be recorded. Histopathological information As the aim of this study was to develop predictors of occult involvement by tumour, only pathological information available before mastectomy was analysed; this included core and open biopsies. Grade, multifocality, presence of lymphovascular invasion and hormone receptor status were recorded, when available. In some cases, the tumour grade was not reported from core biopsies. These slides were retrieved if possible, reviewed, and graded if sufficient material was available. Occult nipple involvement The nipple is not routinely processed for histopathology in this hospital. For this study a 2 × 2 × 2-cm nipple specimen was removed from the fresh mastectomy specimen without affecting clinical pathology processes. Nipple specimens were fixed in formalin and subsequently cut in the coronal plane (perpendicular to the axis of the nipple) into blocks approximately 3 mm thick. One haematoxylin and eosin-stained section was prepared from each block and reviewed by a breast pathologist. All sections containing pathological abnormalities were reviewed by a second breast pathologist. Additional sections and immunohistochemical studies were performed as required for diagnosis. For skin-sparing mastectomy, the surgical plane between the skin and the breast lies approximately 2–5 mm beneath the skin surface17. The retroareolar margin was defined as the section 3 mm below the nipple–areola junction (Fig. 1). Occult nipple involvement was defined as the finding of invasive ductal or lobular carcinoma, lymphovascular invasion or ductal carcinoma in situ (DCIS) in the retroareolar margin section or in any section within the nipple. The presence of lobular carcinoma in situ (LCIS) was not counted as involvement. Fig. 1 Open in new tabDownload slide Sagittal section through a nipple to illustrate the level of the retroareolar margin and the sections taken Statistical analysis The first half of the consecutive series was designated as the training data set and the second half as the test data set. These two data sets were assessed to ensure that the populations were comparable, using Wilcoxon rank sum test for continuous variables and Fisher's exact test for categorical variables. Associations between occult nipple involvement by tumour and potential clinical predictive factors were evaluated in the training set by a series of univariable analyses: logistic regression was performed on continuous variables and Fisher's exact test for categorical variables. All P values reported were two sided and P < 0·050 was considered statistically significant. Using the variables found to be significant on univariable analysis, a multivariable logistic regression model was created using the training data. To assess the diagnostic accuracy of the multivariable model, receiver–operator characteristic curves were constructed and the area under the curve (AUC) calculated. The coefficients from the multivariable logistic regression model were used to derive a nipple involvement probability equation that could be used to determine probability for an individual. The nipple involvement probability equation was applied to the test data set to validate the predictive ability of the model. Statistical analyses were performed using Stata version 918 and R19. Results Two hundred nipple specimens were collected from consecutive skin-sparing mastectomy specimens in a 1-year period. One hundred and thirty mastectomies were performed for breast carcinoma and 70 for prophylaxis. None of the prophylactic specimens was found to have occult nipple involvement, and they were therefore excluded from any further analysis. Thirty-two (24·6 per cent) of the 130 therapeutic mastectomies were found to have occult nipple involvement: 23 were involved by DCIS, six by invasive ductal carcinoma, three by invasive lobular carcinoma and eight by lymphovascular invasion. Seven nipple specimens contained more than one pathological form of involvement. Sixteen contained LCIS but, as mentioned, this was not classified as tumour involvement. Therapeutic mastectomies were divided consecutively, with the first half used as the training set (65) and the second half as the test set (65). The two sets were well matched; none of the variables was statistically significantly different (Tables 1 and 2). Table 1 Comparison of continuous data between training and test sets . Range (n = 130) . Training set (n = 65)* . Test set (n = 65)* . P† . Age (years) 32–86 50·7(8·7) 52·0(11·6) 0·704 Tumour diameter (cm) 0·4–14·8 3·26(2·81) 3·93(3·05) 0·134 Distance of tumour from nipple (cm) 0·5–13·5 4·54(2·93) 4·05(2·08) 0·653 . Range (n = 130) . Training set (n = 65)* . Test set (n = 65)* . P† . Age (years) 32–86 50·7(8·7) 52·0(11·6) 0·704 Tumour diameter (cm) 0·4–14·8 3·26(2·81) 3·93(3·05) 0·134 Distance of tumour from nipple (cm) 0·5–13·5 4·54(2·93) 4·05(2·08) 0·653 * Values are mean(s.d.). † Wilcoxon rank sum test. Open in new tab Table 1 Comparison of continuous data between training and test sets . Range (n = 130) . Training set (n = 65)* . Test set (n = 65)* . P† . Age (years) 32–86 50·7(8·7) 52·0(11·6) 0·704 Tumour diameter (cm) 0·4–14·8 3·26(2·81) 3·93(3·05) 0·134 Distance of tumour from nipple (cm) 0·5–13·5 4·54(2·93) 4·05(2·08) 0·653 . Range (n = 130) . Training set (n = 65)* . Test set (n = 65)* . P† . Age (years) 32–86 50·7(8·7) 52·0(11·6) 0·704 Tumour diameter (cm) 0·4–14·8 3·26(2·81) 3·93(3·05) 0·134 Distance of tumour from nipple (cm) 0·5–13·5 4·54(2·93) 4·05(2·08) 0·653 * Values are mean(s.d.). † Wilcoxon rank sum test. Open in new tab Table 2 Comparison of categorical data between training and test sets . Training set . Test set . P* . Menopausal status 0·647  Premenopausal 29 (45) 25 (38)  Perimenopausal 1 (2) 9 (14)  Postmenopausal 35 (54) 31 (48) Previous ipsilateral breast cancer 0·458  No 61 (94) 56 (86)  Yes 4 (6) 9 (14) Side 0·324  Left 29 (45) 31 (48)  Right 36 (55) 34 (52) Neoadjuvant chemotherapy 0·294  No 58 (89) 62 (95)  Yes 7 (11) 3 (5) Pathology 0·139  Invasive 14 (22) 16 (26)  In situ 22 (34) 19 (31)  Both 29 (45) 27 (44) Invasive type 0·142  Other 1 (3) 0 (0)  Ductal 32 (80) 37 (86)  Lobular 7 (18) 6 (14) Invasive grade 0·900  1 6 (17) 3 (8)  2 18 (50) 17 (44)  3 12 (33) 19 (49) Grade of DCIS 0·298  1 5 (12) 11 (27)  2 17 (41) 9 (22)  3 19 (46) 21 (51) Nipple involvement 0·317  No 48 (74) 50 (77)  Yes 17 (26) 15 (23) . Training set . Test set . P* . Menopausal status 0·647  Premenopausal 29 (45) 25 (38)  Perimenopausal 1 (2) 9 (14)  Postmenopausal 35 (54) 31 (48) Previous ipsilateral breast cancer 0·458  No 61 (94) 56 (86)  Yes 4 (6) 9 (14) Side 0·324  Left 29 (45) 31 (48)  Right 36 (55) 34 (52) Neoadjuvant chemotherapy 0·294  No 58 (89) 62 (95)  Yes 7 (11) 3 (5) Pathology 0·139  Invasive 14 (22) 16 (26)  In situ 22 (34) 19 (31)  Both 29 (45) 27 (44) Invasive type 0·142  Other 1 (3) 0 (0)  Ductal 32 (80) 37 (86)  Lobular 7 (18) 6 (14) Invasive grade 0·900  1 6 (17) 3 (8)  2 18 (50) 17 (44)  3 12 (33) 19 (49) Grade of DCIS 0·298  1 5 (12) 11 (27)  2 17 (41) 9 (22)  3 19 (46) 21 (51) Nipple involvement 0·317  No 48 (74) 50 (77)  Yes 17 (26) 15 (23) Values in parentheses are percentages. DCIS, ductal carcinoma in situ. * Fisher's exact test. Open in new tab Table 2 Comparison of categorical data between training and test sets . Training set . Test set . P* . Menopausal status 0·647  Premenopausal 29 (45) 25 (38)  Perimenopausal 1 (2) 9 (14)  Postmenopausal 35 (54) 31 (48) Previous ipsilateral breast cancer 0·458  No 61 (94) 56 (86)  Yes 4 (6) 9 (14) Side 0·324  Left 29 (45) 31 (48)  Right 36 (55) 34 (52) Neoadjuvant chemotherapy 0·294  No 58 (89) 62 (95)  Yes 7 (11) 3 (5) Pathology 0·139  Invasive 14 (22) 16 (26)  In situ 22 (34) 19 (31)  Both 29 (45) 27 (44) Invasive type 0·142  Other 1 (3) 0 (0)  Ductal 32 (80) 37 (86)  Lobular 7 (18) 6 (14) Invasive grade 0·900  1 6 (17) 3 (8)  2 18 (50) 17 (44)  3 12 (33) 19 (49) Grade of DCIS 0·298  1 5 (12) 11 (27)  2 17 (41) 9 (22)  3 19 (46) 21 (51) Nipple involvement 0·317  No 48 (74) 50 (77)  Yes 17 (26) 15 (23) . Training set . Test set . P* . Menopausal status 0·647  Premenopausal 29 (45) 25 (38)  Perimenopausal 1 (2) 9 (14)  Postmenopausal 35 (54) 31 (48) Previous ipsilateral breast cancer 0·458  No 61 (94) 56 (86)  Yes 4 (6) 9 (14) Side 0·324  Left 29 (45) 31 (48)  Right 36 (55) 34 (52) Neoadjuvant chemotherapy 0·294  No 58 (89) 62 (95)  Yes 7 (11) 3 (5) Pathology 0·139  Invasive 14 (22) 16 (26)  In situ 22 (34) 19 (31)  Both 29 (45) 27 (44) Invasive type 0·142  Other 1 (3) 0 (0)  Ductal 32 (80) 37 (86)  Lobular 7 (18) 6 (14) Invasive grade 0·900  1 6 (17) 3 (8)  2 18 (50) 17 (44)  3 12 (33) 19 (49) Grade of DCIS 0·298  1 5 (12) 11 (27)  2 17 (41) 9 (22)  3 19 (46) 21 (51) Nipple involvement 0·317  No 48 (74) 50 (77)  Yes 17 (26) 15 (23) Values in parentheses are percentages. DCIS, ductal carcinoma in situ. * Fisher's exact test. Open in new tab In the training set, the diameter of the abnormality on imaging and the distance from the nipple by imaging were significantly associated with nipple involvement on univariable analysis (Table 3). An increase in diameter of the primary tumour was associated with an increased risk of nipple involvement, whereas an increase in the distance of the primary tumour from the nipple was associated with a decreased risk. These two significant variables were used to construct the multivariable model (Table 4). As information on distance of the lesion from the nipple was not available for some women, the final model was based on 51 patients in the training set (of a possible 65). The AUC for this model was 0·824 (95 per cent confidence interval (c.i.) 0·664 to 0·984) in the training set. Although the association with distance from the nipple was not statistically significant (P = 0·106), it was included in this final model to help improve prediction because previous studies have established that the removal of variables that are not statistically significant can decrease the predictive ability of a model20,21. When the model was applied to the test data set, the calculated AUC was 0·709 (95 per cent c.i. 0·500 to 0·950). Table 3 Results of univariable analysis of the training set . P* . Age (years) 0·388† Menopausal status 0·167 Past history of ipsilateral breast cancer 0·669 Diameter of lesion on imaging (cm) 0·011† Distance from nipple (cm) 0·014† Invasive, in situ or both on preoperative pathology 0·106 Invasive type 1·000 Invasive grade if known 0·565 Grade of DCIS 0·230 . P* . Age (years) 0·388† Menopausal status 0·167 Past history of ipsilateral breast cancer 0·669 Diameter of lesion on imaging (cm) 0·011† Distance from nipple (cm) 0·014† Invasive, in situ or both on preoperative pathology 0·106 Invasive type 1·000 Invasive grade if known 0·565 Grade of DCIS 0·230 DCIS, ductal carcinoma in situ. * Fisher's exact test unless indicated otherwise; † univariable logistic regression. Open in new tab Table 3 Results of univariable analysis of the training set . P* . Age (years) 0·388† Menopausal status 0·167 Past history of ipsilateral breast cancer 0·669 Diameter of lesion on imaging (cm) 0·011† Distance from nipple (cm) 0·014† Invasive, in situ or both on preoperative pathology 0·106 Invasive type 1·000 Invasive grade if known 0·565 Grade of DCIS 0·230 . P* . Age (years) 0·388† Menopausal status 0·167 Past history of ipsilateral breast cancer 0·669 Diameter of lesion on imaging (cm) 0·011† Distance from nipple (cm) 0·014† Invasive, in situ or both on preoperative pathology 0·106 Invasive type 1·000 Invasive grade if known 0·565 Grade of DCIS 0·230 DCIS, ductal carcinoma in situ. * Fisher's exact test unless indicated otherwise; † univariable logistic regression. Open in new tab Table 4 Results of multivariable analysis (two co-variable model) . Coefficient . P . Intercept − 0·747 0·465 Diameter 0·250 0·075 Distance from nipple − 0·322 0·106 . Coefficient . P . Intercept − 0·747 0·465 Diameter 0·250 0·075 Distance from nipple − 0·322 0·106 Open in new tab Table 4 Results of multivariable analysis (two co-variable model) . Coefficient . P . Intercept − 0·747 0·465 Diameter 0·250 0·075 Distance from nipple − 0·322 0·106 . Coefficient . P . Intercept − 0·747 0·465 Diameter 0·250 0·075 Distance from nipple − 0·322 0·106 Open in new tab The equation for risk of occult nipple involvement (p) was as follows: ln(p/(1−p))=−0.747+(0.250×diameter)−(0.322×distance from nipple)p=e(−0.747+(0.250×diameter)−(0.322×distance from nipple))1+e(−0.747+(0.250×diameter)−(0.322×distance from nipple)) For example, for a patient with a lesion diameter of 1·3 cm, at a distance of 5·4 cm from the nipple, the probability of involvement is 0·10 (10 per cent). For a patient with a cancer of diameter 7·1 cm, at distance of 1·4 cm from the nipple, the probability of involvement is 0·64 (64 per cent). Fig. 2 can be used to estimate the risk of nipple involvement in a patient in whom the diameter of the cancer on imaging and the distance from the nipple on imaging are known. The diagonal lines join points of equal risk and are placed at probability intervals of 0·1. A web-based calculator for this predictive model is available at http://www.cancermath.net under the breast cancer link. Fig. 2 Open in new tabDownload slide Isorisk lines in graph of tumour diameter on imaging versus distance from the nipple. The final model was used to predict risk of occult nipple involvement at various values of diameter and distance from the nipple. Isorisk lines join points of equal risk and allow the reader to estimate risk by reading off the probability of nipple involvement of the lines between which the patient's values fall. Intervals of probability of 0·1 are shown Discussion Three-quarters of women undergoing mastectomy in this study did not have occult nipple involvement. There are other studies of nipple involvement in breast cancer, reporting rates from 0 to 58 per cent, but they involved heterogeneous patient populations, a variety of pathological techniques and differing definitions of tumour involvement. Previous studies often report the combined incidence of clinical and occult nipple involvement22–25. However, patients with clinically obvious nipple involvement would not be candidates for nipple-sparing mastectomy. As the boundaries of safe breast-conserving surgery are expanded by better imaging, localization wires and oncoplastic techniques, an increasing proportion of women undergoing mastectomy have disease that is not amenable to breast conservation. This may partly explain why the rate of occult nipple involvement in mastectomy specimens was higher in this study and other recent series26,27 than in some of the older reports22,28,29. In addition, the advent of breast screening has led to an increase in the frequency of non-palpable, in situ disease which accounted for only 8 per cent of the cases in one older series22 and was excluded in another23. The rates of occult nipple involvement in these two studies were 6·5 and 1 per cent respectively. Because occult involvement of the nipple is often in the form of DCIS (23 of 32 involved nipples in the present series), these cases are of particular significance. Standard texts recommend a single sagittal section through the nipple for histopathology30. A major limitation is that each sagittal section includes only a few ducts. When present in the nipple, DCIS commonly affects a single duct which may be located at the periphery of the duct bundle31. Studies using detailed pathological analysis, such as serial sagittal sections23 or multiple coronal sections25,31, report a high incidence of nipple involvement. Similarly, the depth of the base of the nipple specimen is relevant25. Because skin flaps are typically 2–5 mm thick, the present study examined the level of the retroareolar margin, the histology block beneath the nipple–areola junction. This represented the surgical plane of dissection in nipple-sparing mastectomy11,14,32. Some studies have reported an association between nipple involvement and pathological size of invasive tumour24,29,33, but others did not26,31. As many patients with invasive disease have already undergone breast surgery or neoadjuvant chemotherapy before mastectomy, an accurate measure of invasive tumour size may have been difficult to obtain in some patients. Furthermore, DCIS is the most common form of nipple involvement, and pathology reports rarely describe the size of an area of DCIS. The extent of disease seen on imaging was used here to avoid problems with pathological measurement, and to obtain information that could be used to predict nipple involvement before surgery. Pathological tumour size, positive axillary lymph node status and presence of lymphovascular invasion have been found to be associated with occult nipple involvement26,27,29, but are only known after mastectomy, not relevant in cases of DCIS and are therefore not helpful for prediction. Schecter and colleagues33 reported a prediction model previously, but this was based on only 31 mastectomies over a 4-year period. The present model needs validation in other series but may be valuable in managing women with breast cancer. The possibility of predicting the risk of nipple involvement leads to the question of what constitutes a safe level of risk. An individualized probability facilitates a meaningful discussion between patient and surgical team, and aids decision making about the appropriateness of nipple-sparing mastectomy. Final assessment of the safety of retaining the nipple is only possible after complete histopathological analysis of the retroareolar margin and nipple core tissue of the mastectomy specimen. The ability to detect and manage an unexpected positive margin by local excision of the nipple minimizes the adverse prognostic consequences of false-negative prediction. Case series of nipple-sparing mastectomy with follow-up data suggest that recurrence in the nipple is a rare problem in well selected patients9,12,14. Acknowledgements The authors thank Patricia Della Pelle and Nicole Brousaides for preparing sections, and Leon Chen for development of the accompanying web calculator. This study was supported financially by a philanthropic grant to B.L.S. References 1 Morrow M , Bucci C, Rademaker A. Medical contraindications are not a major factor in the underutilization of breast conserving therapy . J Am Coll Surg 1998 ; 186 : 269 – 274 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Baum M , Budzar AU, Cuzick J, Forbes J, Houghton JH, Klijn JG et al. Anastrozole alone or in combination with tamoxifen versus tamoxifen alone for adjuvant treatment of postmenopausal women with early breast cancer: first results of the ATAC randomised trial . Lancet 2002 ; 359 : 2131 – 2139 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 3 Rebbeck TR , Friebel T, Lynch HT, Neuhausen SL, van't Veer L, Garber JE et al. Bilateral prophylactic mastectomy reduces breast cancer risk in BRCA1 and BRCA2 mutation carriers: the PROSE Study Group . J Clin Oncol 2004 ; 22 : 1055 – 1062 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Toth BA , Lappert P. Modified skin incisions for mastectomy: the need for plastic surgical input in preoperative planning . Plast Reconstr Surg 1991 ; 87 : 1048 – 1053 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Carlson GW , Bostwick J, Styblo TM, Moore B, Bried JT, Murray DR et al. Skin-sparing mastectomy. Oncologic and reconstructive considerations . Ann Surg 1997 ; 225 : 570 – 575 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Cunnick GH , Mokbel K. Skin-sparing mastectomy . Am J Surg 2004 ; 188 : 78 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Greenway RM , Schlossberg L, Dooley WC. Fifteen-year series of skin-sparing mastectomy for stage 0 to 2 breast cancer . Am J Surg 2005 ; 190 : 918 – 922 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Jabor MA , Shayani P, Collins DR, Karas T, Cohen BE. Nipple–areola reconstruction: satisfaction and clinical determinants . Plast Reconstr Surg 2002 ; 110 : 457 – 463 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Gerber B , Krause A, Reimer T, Muller H, Kuchenmeister I, Makovitzky J et al. Skin-sparing mastectomy with conservation of the nipple–areola complex and autologous reconstruction is an oncologically safe procedure . Ann Surg 2003 ; 238 : 120 – 127 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 10 Nahabedian MY , Tsangaris TN. Breast reconstruction following subcutaneous mastectomy for cancer: a critical appraisal of the nipple–areola complex . Plast Reconstr Surg 2006 ; 117 : 1083 – 1090 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Crowe JP JR , Kim JA, Yetman R, Banbury J, Patrick RJ, Baynes D. Nipple-sparing mastectomy: technique and results of 54 procedures . Arch Surg 2004 ; 139 : 148 – 150 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Caruso F , Ferrara M, Castiglione G, Trombetta G, De Meo L, Catanuto G et al. Nipple sparing subcutaneous mastectomy: sixty-six months follow-up . Eur J Surg Oncol 2006 ; 32 : 937 – 940 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Petit JY , Veronesi U, Orecchia R, Luini A, Rey P, Intra M et al. Nipple-sparing mastectomy in association with intra operative radiotherapy (ELIOT): a new type of mastectomy for breast cancer treatment . Breast Cancer Res Treat 2006 ; 96 : 47 – 51 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Sacchini V , Pinotti JA, Barros AC, Luini A, Pluchinotta A, Pinotti M et al. Nipple-sparing mastectomy for breast cancer and risk reduction: oncologic or technical problem? J Am Coll Surg 2006 ; 203 : 704 – 714 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Going JJ , Moffat DF. Escaping from Flatland: clinical and biological aspects of human mammary duct anatomy in three dimensions . J Pathol 2004 ; 203 : 538 – 544 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Rusby JE , Brachtel EF, Michaelson JS, Koerner FC, Smith BL. Breast duct anatomy in the human nipple: three-dimensional patterns and clinical implications . Breast Cancer Res Treat 2007 ; 106 : 171 – 179 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Carlson GW . Skin sparing mastectomy: anatomic and technical considerations . Am Surg 1996 ; 62 : 151 – 155 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 18 StataCorp . Stata Statistical Software (9th edn). StataCorp : College Station , 2005 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 19 R: a Language and Environment for Statistical Computing . R Foundation for Statistical Computing : Vienna , 2005 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 20 Harrell FE , Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors . Stat Med 1996 ; 15 : 361 – 387 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Van Zee KJ , Manasseh DM, Bevilacqua JL, Boolbol SK, Fey JV, Tan LK et al. A nomogram for predicting the likelihood of additional nodal metastases in breast cancer patients with a positive sentinel node biopsy . Ann Surg Oncol 2003 ; 10 : 1140 – 1151 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Smith J , Payne WS, Carney JA. Involvement of the nipple and areola in carcinoma of the breast . Surg Gynecol Obstet 1976 ; 143 : 546 – 548 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 23 Parry RG , Cochran TC, Wolfort FG. When is there nipple involvement in carcinoma of the breast? Plast Reconstr Surg 1977 ; 59 : 535 – 537 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Lagios MD , Gates EA, Westdahl PR, Richards V, Alpert BS. A guide to the frequency of nipple involvement in breast cancer. A study of 149 consecutive mastectomies using a serial subgross and correlated radiographic technique . Am J Surg 1979 ; 138 : 135 – 142 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Andersen JA , Gram JB, Pallesen RM. Involvement of the nipple and areola in breast cancer. Value of clinical findings . Scand J Plast Reconstr Surg 1981 ; 15 : 39 – 42 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Laronga C , Kemp B, Johnston D, Robb GL, Singletary SE. The incidence of occult nipple–areola complex involvement in breast cancer patients receiving a skin-sparing mastectomy . Ann Surg Oncol 1999 ; 6 : 609 – 613 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Vlajcic Z , Zic R, Stanec S, Lambasa S, Petrovecki M, Stanec Z. Nipple–areola complex preservation: predictive factors of neoplastic nipple–areola complex invasion . Ann Plast Surg 2005 ; 55 : 240 – 244 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Fisher ER , Gregorio RM, Fisher B, Redmond C, Vellios F, Sommers SC. The pathology of invasive breast cancer. A syllabus derived from findings of the National Surgical Adjuvant Breast Project (protocol no. 4) . Cancer 1975 ; 36 : 1 – 85 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Kissin MW , Kark AE. Nipple preservation during mastectomy . Br J Surg 1987 ; 74 : 58 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Lester S . Manual of Surgical Pathology (2nd edn). Elsevier Churchill Livingstone : Philadelphia , 2006 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 31 Menon RS , van Geel AN. Cancer of the breast with nipple involvement . Br J Cancer 1989 ; 59 : 81 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Garcia-Etienne CA , Borgen PI. Update on the indications for nipple-sparing mastectomy . J Support Oncol 2006 ; 4 : 225 – 230 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 33 Schecter AK , Freeman MB, Giri D, Sabo E, Weinzweig J. Applicability of the nipple–areola complex-sparing mastectomy: a prediction model using mammography to estimate risk of nipple–areola complex involvement in breast cancer patients . Ann Plast Surg 2006 ; 56 : 498 – 504 . Google Scholar Crossref Search ADS PubMed WorldCat Author notes Presented to a meeting of the Society of Surgical Oncology, Washington, DC, USA, April 2007 Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

Martin, L; Jia, C; Rouvelas, I; Lagergren, P

doi: 10.1002/bjs.6374pmid: 18844265

Abstract Background Oesophageal cancer surgery is often followed by malnutrition, but the factors causing weight loss are unknown. The aim of this population-based study was to identify such risk factors. Methods Data were collected from a nationwide Swedish organization for research on surgery for oesophageal cancer. A total of 340 patients (75·9 per cent of those eligible) responded to a study-specific questionnaire concerning height and weight, just before and 6 months after surgery. Factors influencing malnutrition, defined as loss of body mass index of at least 15 per cent 6 months after operation, were identified by logistic regression. Results Neoadjuvant therapy (received by 10·6 per cent of all patients) and female sex were associated with at least a twofold increased risk of weight loss (odds ratio (OR) 2·41 (95 per cent confidence interval 1·01 to 5·77) and 2·14 (1·07 to 4·28) respectively), whereas preoperative weight loss was associated with a decreased risk (OR 0·13 (0·03 to 0·65)). Age, tumour stage and location, type of oesophageal substitute, suture technique and postoperative complications did not influence the risk. Conclusion Neoadjuvant therapy and female sex appear to be associated with an increased risk of malnutrition after oesophageal cancer surgery. Introduction Resection of the oesophagus can offer cure for oesophageal cancer, but at the risk of high postoperative morbidity1,2. Weight loss is common among patients with oesophageal cancer, as a result of cancer cachexia and dysphagia with poor food intake3,4. However, the reduction in weight seems to persist or accelerate after surgery. In a recent population-based study, two-thirds of patients had lost more than 10 per cent of their preoperative body mass within 6 months of surgery, and one in five had lost more than 20 per cent5. Although there is no generally established definition, an unintentional decrease in weight of 10–15 per cent within 3–6 months is often considered as malnutrition6–8. Duration and severity of malnutrition affect the outcome of surgery in terms of mortality, morbidity and quality of life9,10. Risk factors underlying malnutrition after oesophageal cancer surgery need to be explored, as they might be preventable. The contradictory findings from randomized clinical trials addressing survival as the main outcome have resulted in disagreement about the optimal curative treatment for oesophageal cancer, mainly regarding the surgical approach and whether or not to use neoadjuvant therapy11,12. The selection of patients for surgical and neoadjuvant therapy is at least partly subjective, and depends on various patient and tumour characteristics. The influence of preoperative weight loss on postoperative malnutrition is unknown. Early severe surgical complications are linked with longstanding morbidity13, but their influence on weight change is uncertain. The aim of this study was to identify risk factors for malnutrition 6 months after oesophageal cancer surgery in a population-based nationwide study in Sweden. Methods Potential risk factors for malnutrition 6 months after oesophageal cancer surgery were investigated in a prospective, population-based cohort study. All patients with oesophageal or gastric cardia cancer who underwent both macroscopically and microscopically radical (R0) tumour resection in Sweden between 2 April 2001 and 31 December 2005 were eligible to participate. Patients were followed up until June 2006. The ethics committee at Karolinska University Hospital, Karolinska Institute, Stockholm, Sweden, approved the study. Informed consent was obtained from each patient before inclusion. Data collection Comprehensive data were collected from a national Swedish organization for research on patients having surgery for oesophageal cancer including tumours of the gastro-oesophageal junction. The organization14 and data collection process1 have been described in detail elsewhere. Collaboration with relevant hospital departments enabled complete prospective data collection for 90 per cent of all patients treated surgically in Sweden during the study period. Data on patient and tumour characteristics, surgery and neoadjuvant therapy were collected and evaluated uniformly according to the study protocol. Six months after surgery, patients responded to a written study-specific questionnaire concerning body height and weight, just before and at 6 months after surgery. In a previous study, the preoperative self-reported weight as recalled 6 months after surgery was compared with the objectively measured weight recorded in the operation charts in a sample of 104 patients5. The correlation coefficient was 0·77, indicating good agreement. Preoperative weight loss was calculated as the difference between the average adult body mass index (BMI) and that measured immediately before operation, and was categorized as less than 10, 10–14·9 per cent, or 15 per cent or more. Patient characteristics comprised sex and age, categorized into less than 60, 60–69 or at least 70 years. Tumour characteristics comprised tumour stage, categorized into 0–I, II, III or IV; tumour location, categorized into lower oesophagus and cardia or upper and middle oesophagus; and histological type, categorized as adenocarcinoma (including high-grade dysplasia) or squamous cell carcinoma. Tumour stage was defined according to the tumour node metastasis classification of the International Union Against Cancer15. The Siewert definitions were used to classify the location of adenocarcinomas adjacent to the gastro-oesophageal junction16. Neoadjuvant therapy consisted of chemotherapy, radiotherapy or combined chemotherapy and radiotherapy before the surgical procedure. Patients were grouped into those who had neoadjuvant therapy and those who did not. Technical surgical factors included surgical approach, categorized into transhiatal (abdominal only) or transthoracic (abdominal incision and a right posterolateral thoracotomy), type of oesophageal substitute (stomach, jejunum or colon), and type of anastomosis (handsewn or stapled). Surgical complications that occurred within 30 days after surgery were predefined and included postoperative bleeding exceeding 2000 ml or requiring reoperation, anastomotic leakage, necrosis of the substitute, intra-abdominal abscess, intrathoracic abscess, septicaemia, wound infection, rupture of the wound, renal or hepatic failure, respiratory insufficiency, recurrent laryngeal nerve injury, lymphatic leakage into the thorax and perforation of the stomach. Patients were categorized according to the number of complications: none, one or two, or more than two. Technical surgical complications (postoperative bleeding, anastomotic leakage, necrosis of the substitute, recurrent laryngeal nerve injury, lymphatic leakage into the thorax, perforation of the stomach) were analysed as a separate variable. For this analysis, patients were grouped into those who had such complications and those who did not. The study outcome, postoperative malnutrition, was defined as a decrease in BMI of at least 15 per cent between the immediate preoperative value and the value 6 months after surgery. Statistical analysis For comparisons of patient characteristics and clinical variables, the χ2 test (when the sample size was at least five in each group) or Fisher's exact test (sample size less than five in any of the groups) was used to test for statistically significant difference at the 5 per cent level. Logistic regression analysis was used to estimate relative risks of malnutrition, expressed as odds ratios (ORs) with 95 per cent confidence intervals (c.i.). P values for trends were calculated by introducing the variable with continuous scales in the model. All analyses were first performed in a crude model without adjustment. Thereafter, adjustments for potential confounding factors were carried out in two models. The basic model included adjustment for age, sex and tumour stage. In the more extensive multivariable model further adjustments were made for histological type, tumour location, surgical approach, surgical complications, co-morbidity, neoadjuvant therapy, preoperative BMI and nutritional catheter jejunostomy. Adjustment was made for preoperative weight loss instead of preoperative BMI in the analysis of neoadjuvant therapy and risk of malnutrition. With the exception of co-morbidity (categorized into none, one or two, or more than two conditions) and use of a catheter jejunostomy for feeding (yes or no), all included co-factors were grouped as described above. The statistical software Stata® version 9.2 (Stata Corporation, College Station, Texas, USA) was used for all analyses. Results Of 616 patients operated on for oesophageal or cardia cancer in the original study cohort, 168 who died within 6 months after surgery or did not undergo microscopically or macroscopically radical surgery were not eligible for this study. A total of 108 of the 448 eligible patients were not included, because of delayed or missed registration (70), non-participation (21), or missing data regarding height or weight (17), leaving 340 patients for analysis (75·9 per cent of those eligible). Weight loss of 15 per cent or more was generally more pronounced after operation (36·8 per cent) than before (12·9 per cent). Patient characteristics, stratified by postoperative weight loss of at least 15 per cent or less than 15 per cent, are presented in Table 1. There were no differences in median age at the time of surgery, but the proportion of patients aged 70 years or more was lower in the group with malnutrition. The sex distributions were similar. Preoperative weight loss was less pronounced in those with postoperative malnutrition. Tumour type, stage and location were similar in the two groups. Eight patients (2·4 per cent) had high-grade dysplasia. Table 1 Characteristics and clinical details of 340 patients who had resection for oesophageal cancer stratified by postoperative weight loss . Postoperative weight loss (%) . P* . . < 15 . ≥ 15 . Whole group 215 (63·2) 125 (36·8) Age (years) 0·045  < 60 46 (21·4) 36 (28·8)  60–69 73 (34·0) 50 (40·0)  ≥ 70 96 (44·7) 39 (31·2) Sex 0·622  F 40 (18·6) 26 (20·8)  M 175 (81·4) 99 (79·2) Preoperative weight loss (%) < 0·001  < 10 136 (63·3) 102 (81·6)  10–14·9 24 (11·2) 6 (4·8)  ≥ 15 42 (19·5) 2 (1·6)  Not known 13 (6·0) 15 (12·0) Histology 0·074  Adenocarcinoma including dysplasia 159 (74·0) 103 (82·4)  Squamous cell carcinoma 56 (26·0) 22 (17·6) Tumour stage 0·462  0–I 47 (21·9) 31 (24·8)  II 72 (33·5) 41 (32·8)  III 86 (40·0) 45 (36·0)  IV 9 (4·2) 7 (5·6)  Not known 1 (0·5) 1 (0·8) Tumour location 0·310  Lower oesophagus and cardia 124 (57·7) 65 (52·0)  Upper and middle oesophagus 91 (42·3) 60 (48·0) Neoadjuvant therapy 0·082  No 197 (91·6) 107 (85·6)  Yes 18 (8·4) 18 (14·4) Surgical approach 0·059  Abdominal 28 (13·0) 26 (20·8)  Thoracoabdominal 187 (87·0) 99 (79·2) Type of substitute 0·939  Stomach 170 (79·1) 97 (77·6)  Jejunum 39 (18·1) 25 (20·0)  Colon 6 (2·8) 3 (2·4) Type of anastomosis 0·247  Handsewn 91 (42·3) 61 (48·8)  Stapled 124 (57·7) 64 (51·2) Site of anastomosis 0·193  Neck 16 (7·4) 9 (7·2)  Thorax 176 (81·9) 94 (75·2)  Abdomen 23 (10·7) 22 (17·6) Complications 0·409  0 124 (57·7) 63 (50·4)  1–2 70 (32·6) 49 (39·2)  > 2 21 (9·8) 13 (10·4) . Postoperative weight loss (%) . P* . . < 15 . ≥ 15 . Whole group 215 (63·2) 125 (36·8) Age (years) 0·045  < 60 46 (21·4) 36 (28·8)  60–69 73 (34·0) 50 (40·0)  ≥ 70 96 (44·7) 39 (31·2) Sex 0·622  F 40 (18·6) 26 (20·8)  M 175 (81·4) 99 (79·2) Preoperative weight loss (%) < 0·001  < 10 136 (63·3) 102 (81·6)  10–14·9 24 (11·2) 6 (4·8)  ≥ 15 42 (19·5) 2 (1·6)  Not known 13 (6·0) 15 (12·0) Histology 0·074  Adenocarcinoma including dysplasia 159 (74·0) 103 (82·4)  Squamous cell carcinoma 56 (26·0) 22 (17·6) Tumour stage 0·462  0–I 47 (21·9) 31 (24·8)  II 72 (33·5) 41 (32·8)  III 86 (40·0) 45 (36·0)  IV 9 (4·2) 7 (5·6)  Not known 1 (0·5) 1 (0·8) Tumour location 0·310  Lower oesophagus and cardia 124 (57·7) 65 (52·0)  Upper and middle oesophagus 91 (42·3) 60 (48·0) Neoadjuvant therapy 0·082  No 197 (91·6) 107 (85·6)  Yes 18 (8·4) 18 (14·4) Surgical approach 0·059  Abdominal 28 (13·0) 26 (20·8)  Thoracoabdominal 187 (87·0) 99 (79·2) Type of substitute 0·939  Stomach 170 (79·1) 97 (77·6)  Jejunum 39 (18·1) 25 (20·0)  Colon 6 (2·8) 3 (2·4) Type of anastomosis 0·247  Handsewn 91 (42·3) 61 (48·8)  Stapled 124 (57·7) 64 (51·2) Site of anastomosis 0·193  Neck 16 (7·4) 9 (7·2)  Thorax 176 (81·9) 94 (75·2)  Abdomen 23 (10·7) 22 (17·6) Complications 0·409  0 124 (57·7) 63 (50·4)  1–2 70 (32·6) 49 (39·2)  > 2 21 (9·8) 13 (10·4) Values in parentheses are percentages. * χ2 test (when sample size was at least five in each group) or Fisher's exact test (when sample size was less than five in any of the groups). Open in new tab Table 1 Characteristics and clinical details of 340 patients who had resection for oesophageal cancer stratified by postoperative weight loss . Postoperative weight loss (%) . P* . . < 15 . ≥ 15 . Whole group 215 (63·2) 125 (36·8) Age (years) 0·045  < 60 46 (21·4) 36 (28·8)  60–69 73 (34·0) 50 (40·0)  ≥ 70 96 (44·7) 39 (31·2) Sex 0·622  F 40 (18·6) 26 (20·8)  M 175 (81·4) 99 (79·2) Preoperative weight loss (%) < 0·001  < 10 136 (63·3) 102 (81·6)  10–14·9 24 (11·2) 6 (4·8)  ≥ 15 42 (19·5) 2 (1·6)  Not known 13 (6·0) 15 (12·0) Histology 0·074  Adenocarcinoma including dysplasia 159 (74·0) 103 (82·4)  Squamous cell carcinoma 56 (26·0) 22 (17·6) Tumour stage 0·462  0–I 47 (21·9) 31 (24·8)  II 72 (33·5) 41 (32·8)  III 86 (40·0) 45 (36·0)  IV 9 (4·2) 7 (5·6)  Not known 1 (0·5) 1 (0·8) Tumour location 0·310  Lower oesophagus and cardia 124 (57·7) 65 (52·0)  Upper and middle oesophagus 91 (42·3) 60 (48·0) Neoadjuvant therapy 0·082  No 197 (91·6) 107 (85·6)  Yes 18 (8·4) 18 (14·4) Surgical approach 0·059  Abdominal 28 (13·0) 26 (20·8)  Thoracoabdominal 187 (87·0) 99 (79·2) Type of substitute 0·939  Stomach 170 (79·1) 97 (77·6)  Jejunum 39 (18·1) 25 (20·0)  Colon 6 (2·8) 3 (2·4) Type of anastomosis 0·247  Handsewn 91 (42·3) 61 (48·8)  Stapled 124 (57·7) 64 (51·2) Site of anastomosis 0·193  Neck 16 (7·4) 9 (7·2)  Thorax 176 (81·9) 94 (75·2)  Abdomen 23 (10·7) 22 (17·6) Complications 0·409  0 124 (57·7) 63 (50·4)  1–2 70 (32·6) 49 (39·2)  > 2 21 (9·8) 13 (10·4) . Postoperative weight loss (%) . P* . . < 15 . ≥ 15 . Whole group 215 (63·2) 125 (36·8) Age (years) 0·045  < 60 46 (21·4) 36 (28·8)  60–69 73 (34·0) 50 (40·0)  ≥ 70 96 (44·7) 39 (31·2) Sex 0·622  F 40 (18·6) 26 (20·8)  M 175 (81·4) 99 (79·2) Preoperative weight loss (%) < 0·001  < 10 136 (63·3) 102 (81·6)  10–14·9 24 (11·2) 6 (4·8)  ≥ 15 42 (19·5) 2 (1·6)  Not known 13 (6·0) 15 (12·0) Histology 0·074  Adenocarcinoma including dysplasia 159 (74·0) 103 (82·4)  Squamous cell carcinoma 56 (26·0) 22 (17·6) Tumour stage 0·462  0–I 47 (21·9) 31 (24·8)  II 72 (33·5) 41 (32·8)  III 86 (40·0) 45 (36·0)  IV 9 (4·2) 7 (5·6)  Not known 1 (0·5) 1 (0·8) Tumour location 0·310  Lower oesophagus and cardia 124 (57·7) 65 (52·0)  Upper and middle oesophagus 91 (42·3) 60 (48·0) Neoadjuvant therapy 0·082  No 197 (91·6) 107 (85·6)  Yes 18 (8·4) 18 (14·4) Surgical approach 0·059  Abdominal 28 (13·0) 26 (20·8)  Thoracoabdominal 187 (87·0) 99 (79·2) Type of substitute 0·939  Stomach 170 (79·1) 97 (77·6)  Jejunum 39 (18·1) 25 (20·0)  Colon 6 (2·8) 3 (2·4) Type of anastomosis 0·247  Handsewn 91 (42·3) 61 (48·8)  Stapled 124 (57·7) 64 (51·2) Site of anastomosis 0·193  Neck 16 (7·4) 9 (7·2)  Thorax 176 (81·9) 94 (75·2)  Abdomen 23 (10·7) 22 (17·6) Complications 0·409  0 124 (57·7) 63 (50·4)  1–2 70 (32·6) 49 (39·2)  > 2 21 (9·8) 13 (10·4) Values in parentheses are percentages. * χ2 test (when sample size was at least five in each group) or Fisher's exact test (when sample size was less than five in any of the groups). Open in new tab Neoadjuvant therapy was received by 10·6 per cent of the patients and this treatment was more common in the malnutrition group. Surgical factors were similarly distributed between groups. In most patients a thoracoabdominal approach was used (84·1 per cent), with the stomach as the replacement conduit (78·5 per cent). Nearly half of the patients (45·0 per cent) had at least one complication occurring within 30 days of surgery, but there was no major difference between the groups with and without postoperative malnutrition (Table 1). Patients with more preoperative weight loss had a lower risk of weight loss after surgery than those with less preoperative weight loss, and this difference remained strong after adjustment for all evaluated confounders (Table 2). Among patients who had lost at least 15 per cent of BMI before surgery, the adjusted OR for postoperative weight loss was 0·13 (95 per cent c.i. 0·03 to 0·65) compared with those who had lost less than 10 per cent of BMI before surgery. Table 2 Patient and tumour risk factors for malnutrition (weight loss at least 15 per cent) after resection for oesophageal cancer among 340 patients . Basic model* . Multivariable model† . . Odds ratio . P‡ . P§ . Odds ratio . P‡ . P§ . Preoperative weight loss (%) < 0·001 0·006  < 10 (reference) 1·00 1·00  10–14·9 0·32 (0·13, 0·84) 0·020 0·51 (0·17, 1·57) 0·240  ≥ 15 0·06 (0·01, 0·27) < 0·001 0·13 (0·03, 0·65) 0·013 Age (years) 0·019 0·162  < 60 (reference) 1·00 1·00  60–70 0·88 (0·49, 1·57) 0·668 1·13 (0·59, 2·20) 0·709  ≥ 70 0·52 (0·29, 0·92) 0·026 0·66 (0·34, 1·29) 0·226 Sex  M (reference) 1·00 1·00  F 1·26 (0·71, 2·25) 0·430 2·14 (1·07, 4·28) 0·032 Tumour stage 0·875 0·633  0–I (reference) 1·00 1·00  II 0·84 (0·46, 1·55) 0·582 1·17 (0·59, 2·33) 0·645  III 0·86 (0·48, 1·55) 0·617 1·23 (0·64, 2·39) 0·532  IV 1·19 (0·39, 3·57) 0·762 1·03 (0·31, 3·40) 0·956 Tumour location  Upper and middle oesophagus (reference) 1·00 1·00  Lower oesophagus and cardia 0·80 (0·51, 1·26) 0·338 0·92 (0·50, 1·67) 0·775 Histological type  Adenocarcinoma (reference) 1·00 1·00  Squamous cell carcinoma 0·55 (0·31, 0·99) 0·045 0·98 (0·47, 2·06) 0·959 . Basic model* . Multivariable model† . . Odds ratio . P‡ . P§ . Odds ratio . P‡ . P§ . Preoperative weight loss (%) < 0·001 0·006  < 10 (reference) 1·00 1·00  10–14·9 0·32 (0·13, 0·84) 0·020 0·51 (0·17, 1·57) 0·240  ≥ 15 0·06 (0·01, 0·27) < 0·001 0·13 (0·03, 0·65) 0·013 Age (years) 0·019 0·162  < 60 (reference) 1·00 1·00  60–70 0·88 (0·49, 1·57) 0·668 1·13 (0·59, 2·20) 0·709  ≥ 70 0·52 (0·29, 0·92) 0·026 0·66 (0·34, 1·29) 0·226 Sex  M (reference) 1·00 1·00  F 1·26 (0·71, 2·25) 0·430 2·14 (1·07, 4·28) 0·032 Tumour stage 0·875 0·633  0–I (reference) 1·00 1·00  II 0·84 (0·46, 1·55) 0·582 1·17 (0·59, 2·33) 0·645  III 0·86 (0·48, 1·55) 0·617 1·23 (0·64, 2·39) 0·532  IV 1·19 (0·39, 3·57) 0·762 1·03 (0·31, 3·40) 0·956 Tumour location  Upper and middle oesophagus (reference) 1·00 1·00  Lower oesophagus and cardia 0·80 (0·51, 1·26) 0·338 0·92 (0·50, 1·67) 0·775 Histological type  Adenocarcinoma (reference) 1·00 1·00  Squamous cell carcinoma 0·55 (0·31, 0·99) 0·045 0·98 (0·47, 2·06) 0·959 Values in parentheses are 95 per cent confidence intervals. * Basic model includes age, sex and tumour stage; † multivariable model includes age, sex, tumour stage, histological type, tumour location, surgical approach, complications, co-morbidity, neoadjuvant therapy, preoperative body mass index and nutritional catheter jejunostomy. ‡ Logistic regression analysis; § test for trend. Open in new tab Table 2 Patient and tumour risk factors for malnutrition (weight loss at least 15 per cent) after resection for oesophageal cancer among 340 patients . Basic model* . Multivariable model† . . Odds ratio . P‡ . P§ . Odds ratio . P‡ . P§ . Preoperative weight loss (%) < 0·001 0·006  < 10 (reference) 1·00 1·00  10–14·9 0·32 (0·13, 0·84) 0·020 0·51 (0·17, 1·57) 0·240  ≥ 15 0·06 (0·01, 0·27) < 0·001 0·13 (0·03, 0·65) 0·013 Age (years) 0·019 0·162  < 60 (reference) 1·00 1·00  60–70 0·88 (0·49, 1·57) 0·668 1·13 (0·59, 2·20) 0·709  ≥ 70 0·52 (0·29, 0·92) 0·026 0·66 (0·34, 1·29) 0·226 Sex  M (reference) 1·00 1·00  F 1·26 (0·71, 2·25) 0·430 2·14 (1·07, 4·28) 0·032 Tumour stage 0·875 0·633  0–I (reference) 1·00 1·00  II 0·84 (0·46, 1·55) 0·582 1·17 (0·59, 2·33) 0·645  III 0·86 (0·48, 1·55) 0·617 1·23 (0·64, 2·39) 0·532  IV 1·19 (0·39, 3·57) 0·762 1·03 (0·31, 3·40) 0·956 Tumour location  Upper and middle oesophagus (reference) 1·00 1·00  Lower oesophagus and cardia 0·80 (0·51, 1·26) 0·338 0·92 (0·50, 1·67) 0·775 Histological type  Adenocarcinoma (reference) 1·00 1·00  Squamous cell carcinoma 0·55 (0·31, 0·99) 0·045 0·98 (0·47, 2·06) 0·959 . Basic model* . Multivariable model† . . Odds ratio . P‡ . P§ . Odds ratio . P‡ . P§ . Preoperative weight loss (%) < 0·001 0·006  < 10 (reference) 1·00 1·00  10–14·9 0·32 (0·13, 0·84) 0·020 0·51 (0·17, 1·57) 0·240  ≥ 15 0·06 (0·01, 0·27) < 0·001 0·13 (0·03, 0·65) 0·013 Age (years) 0·019 0·162  < 60 (reference) 1·00 1·00  60–70 0·88 (0·49, 1·57) 0·668 1·13 (0·59, 2·20) 0·709  ≥ 70 0·52 (0·29, 0·92) 0·026 0·66 (0·34, 1·29) 0·226 Sex  M (reference) 1·00 1·00  F 1·26 (0·71, 2·25) 0·430 2·14 (1·07, 4·28) 0·032 Tumour stage 0·875 0·633  0–I (reference) 1·00 1·00  II 0·84 (0·46, 1·55) 0·582 1·17 (0·59, 2·33) 0·645  III 0·86 (0·48, 1·55) 0·617 1·23 (0·64, 2·39) 0·532  IV 1·19 (0·39, 3·57) 0·762 1·03 (0·31, 3·40) 0·956 Tumour location  Upper and middle oesophagus (reference) 1·00 1·00  Lower oesophagus and cardia 0·80 (0·51, 1·26) 0·338 0·92 (0·50, 1·67) 0·775 Histological type  Adenocarcinoma (reference) 1·00 1·00  Squamous cell carcinoma 0·55 (0·31, 0·99) 0·045 0·98 (0·47, 2·06) 0·959 Values in parentheses are 95 per cent confidence intervals. * Basic model includes age, sex and tumour stage; † multivariable model includes age, sex, tumour stage, histological type, tumour location, surgical approach, complications, co-morbidity, neoadjuvant therapy, preoperative body mass index and nutritional catheter jejunostomy. ‡ Logistic regression analysis; § test for trend. Open in new tab According to the average adult BMI before surgery, 58·4 per cent of patients were overweight (BMI over 25 kg/m2) and 23·0 per cent were obese (BMI over 30 kg/m2). The corresponding values 6 months after the operation were 20·4 and 2·3 per cent respectively. Older patients tended to have a lower risk of postoperative weight loss than younger patients (Table 2). The difference between patients aged at least 70 years and those below 60 years did not, however, remain statistically significant in the full multivariable model (OR 0·66 (95 per cent c.i. 0·34 to 1·29)). Female sex was associated with a significantly increased risk of postoperative weight loss (OR 2·14 (95 per cent c.i. 1·07 to 4·28)). Tumour stage, location or histology did not significantly influence the risk of postoperative malnutrition according to the multivariable analysis (Table 2). Use of neoadjuvant therapy was associated with a greater than twofold increase in the risk of postoperative malnutrition in the fully adjusted model (OR 2·41 (95 per cent c.i. 1·01 to 5·77)) (Table 3). To take into account the delay in surgery due to neoadjuvant therapy, a further adjustment for preoperative weight loss was made when analysing this variable. Table 3 Surgical risk factors for malnutrition (weight loss at least 15 per cent) after resection for oesophageal cancer among 340 patients . Basic model† . Multivariable model‡ . . Odds ratio . P¶ . P# . Odds ratio . P¶ . P# . Neoadjuvant therapy  No (reference) 1·00 1·00  Yes 1·70 (0·83, 3·50) 0·149 2·41 (1·01, 5·77)§ 0·048 Approach  Thoracoabdominal (reference) 1·00 1·00  Abdominal 1·81 (0·98, 3·37) 0·059 2·04 (0·98, 4·23) 0·056 Substitute  Stomach (reference) 1·00 1·00  Jejunum 1·11 (0·62, 1·97) 0·729 0·65 (0·26, 1·61) 0·350  Colon 0·77 (0·18, 3·27) 0·725 1·17 (0·18, 7·55) 0·872 Anastomosis  Handsewn (reference) 1·00 1·00  Stapled 0·73 (0·46, 1·16) 0·188 0·84 (0·49, 1·45) 0·532 Complications 0·256 0·251  0 (reference) 1·00 1·00  1–2 1·33 (0·82, 2·16) 0·255 1·24 (0·72, 2·13) 0·430  > 2 1·34 (0·62, 2·91) 0·453 1·55 (0·65, 3·69) 0·325 Technical surgical complications  No (reference) 1·00 1·00  Yes* 1·17 (0·64, 2·17) 0·610 1·02 (0·46, 2·28) 0·953 . Basic model† . Multivariable model‡ . . Odds ratio . P¶ . P# . Odds ratio . P¶ . P# . Neoadjuvant therapy  No (reference) 1·00 1·00  Yes 1·70 (0·83, 3·50) 0·149 2·41 (1·01, 5·77)§ 0·048 Approach  Thoracoabdominal (reference) 1·00 1·00  Abdominal 1·81 (0·98, 3·37) 0·059 2·04 (0·98, 4·23) 0·056 Substitute  Stomach (reference) 1·00 1·00  Jejunum 1·11 (0·62, 1·97) 0·729 0·65 (0·26, 1·61) 0·350  Colon 0·77 (0·18, 3·27) 0·725 1·17 (0·18, 7·55) 0·872 Anastomosis  Handsewn (reference) 1·00 1·00  Stapled 0·73 (0·46, 1·16) 0·188 0·84 (0·49, 1·45) 0·532 Complications 0·256 0·251  0 (reference) 1·00 1·00  1–2 1·33 (0·82, 2·16) 0·255 1·24 (0·72, 2·13) 0·430  > 2 1·34 (0·62, 2·91) 0·453 1·55 (0·65, 3·69) 0·325 Technical surgical complications  No (reference) 1·00 1·00  Yes* 1·17 (0·64, 2·17) 0·610 1·02 (0·46, 2·28) 0·953 Values in parentheses are 95 per cent confidence intervals. * At least one complication including: postoperative bleeding (more than 2 litres or need for reoperation), anastomotic leakage, necrosis of the substitute, septicaemia, recurrent laryngeal nerve injury, lymphatic leakage in the thorax, perforation of the stomach. † Basic model includes age, sex and tumour stage; ‡ multivariable model includes age, sex, tumour stage, histological type, tumour location, surgical approach, complications, co-morbidity, neoadjuvant therapy, preoperative body mass index, nutritional catheter jejunostomy; § multivariable model except that adjustment was made for preoperative weight loss instead of preoperative body mass index. ¶ Logistic regression analysis; # test for trend. Open in new tab Table 3 Surgical risk factors for malnutrition (weight loss at least 15 per cent) after resection for oesophageal cancer among 340 patients . Basic model† . Multivariable model‡ . . Odds ratio . P¶ . P# . Odds ratio . P¶ . P# . Neoadjuvant therapy  No (reference) 1·00 1·00  Yes 1·70 (0·83, 3·50) 0·149 2·41 (1·01, 5·77)§ 0·048 Approach  Thoracoabdominal (reference) 1·00 1·00  Abdominal 1·81 (0·98, 3·37) 0·059 2·04 (0·98, 4·23) 0·056 Substitute  Stomach (reference) 1·00 1·00  Jejunum 1·11 (0·62, 1·97) 0·729 0·65 (0·26, 1·61) 0·350  Colon 0·77 (0·18, 3·27) 0·725 1·17 (0·18, 7·55) 0·872 Anastomosis  Handsewn (reference) 1·00 1·00  Stapled 0·73 (0·46, 1·16) 0·188 0·84 (0·49, 1·45) 0·532 Complications 0·256 0·251  0 (reference) 1·00 1·00  1–2 1·33 (0·82, 2·16) 0·255 1·24 (0·72, 2·13) 0·430  > 2 1·34 (0·62, 2·91) 0·453 1·55 (0·65, 3·69) 0·325 Technical surgical complications  No (reference) 1·00 1·00  Yes* 1·17 (0·64, 2·17) 0·610 1·02 (0·46, 2·28) 0·953 . Basic model† . Multivariable model‡ . . Odds ratio . P¶ . P# . Odds ratio . P¶ . P# . Neoadjuvant therapy  No (reference) 1·00 1·00  Yes 1·70 (0·83, 3·50) 0·149 2·41 (1·01, 5·77)§ 0·048 Approach  Thoracoabdominal (reference) 1·00 1·00  Abdominal 1·81 (0·98, 3·37) 0·059 2·04 (0·98, 4·23) 0·056 Substitute  Stomach (reference) 1·00 1·00  Jejunum 1·11 (0·62, 1·97) 0·729 0·65 (0·26, 1·61) 0·350  Colon 0·77 (0·18, 3·27) 0·725 1·17 (0·18, 7·55) 0·872 Anastomosis  Handsewn (reference) 1·00 1·00  Stapled 0·73 (0·46, 1·16) 0·188 0·84 (0·49, 1·45) 0·532 Complications 0·256 0·251  0 (reference) 1·00 1·00  1–2 1·33 (0·82, 2·16) 0·255 1·24 (0·72, 2·13) 0·430  > 2 1·34 (0·62, 2·91) 0·453 1·55 (0·65, 3·69) 0·325 Technical surgical complications  No (reference) 1·00 1·00  Yes* 1·17 (0·64, 2·17) 0·610 1·02 (0·46, 2·28) 0·953 Values in parentheses are 95 per cent confidence intervals. * At least one complication including: postoperative bleeding (more than 2 litres or need for reoperation), anastomotic leakage, necrosis of the substitute, septicaemia, recurrent laryngeal nerve injury, lymphatic leakage in the thorax, perforation of the stomach. † Basic model includes age, sex and tumour stage; ‡ multivariable model includes age, sex, tumour stage, histological type, tumour location, surgical approach, complications, co-morbidity, neoadjuvant therapy, preoperative body mass index, nutritional catheter jejunostomy; § multivariable model except that adjustment was made for preoperative weight loss instead of preoperative body mass index. ¶ Logistic regression analysis; # test for trend. Open in new tab The purely abdominal approach had a borderline significant twofold increased risk of postoperative malnutrition compared with the combined thoracoabdominal approach (OR 2·04 (95 per cent c.i. 0·98 to 4·23)). Type of oesophageal substitute or anastomosis did not influence the risk of malnutrition after surgery (Table 3). There was a 55 per cent increase in the risk of postoperative malnutrition among patients with more than two surgical complications compared with those without complications, but this was not statistically significant (OR 1·55 (95 per cent c.i. 0·65 to 3·69)). Technical surgical complications did not increase the risk of postoperative malnutrition (Table 3). Stratified analyses of the groups operated on with total gastrectomy and with other types of resection leaving at least a part of the stomach did not reveal any material differences between the groups (data not shown). Discussion Female sex, neoadjuvant therapy and a purely abdominal surgical approach were associated with an increased risk of malnutrition 6 months after oesophageal cancer surgery, whereas high preoperative weight loss was associated with a decreased risk of weight loss after surgery. Weight loss was in general more pronounced after operation than preoperative loss. Some methodological aspects deserve attention. The study benefits from a population-based design with a high participation rate, which reduces the risk of selection bias and facilitates generalizability. Non-participation might, however, have influenced results. Misclassification of malnutrition is an obvious problem, but the definition was based on previous research and decided before initiation of the analyses. Moreover, any such misclassification should be non-differential and cannot explain the associations identified; it would only dilute the effects. The detailed and prospective data collection reduced the risk of misclassification and allowed some ability to control for several potential confounders. Data on lifestyle factors were not available, but the adjustment for histological tumour type indirectly took into account strong risk factors for squamous cell carcinoma and adenocarcinoma, such as tobacco and alcohol use and gastro-oesophageal reflux respectively. Although patients with high-grade dysplasia do not usually have dysphagia or weight loss before surgery, this should not materially affect the results as they were so few (eight patients). To avoid the possible influence of early tumour recurrence on postoperative weight development, the study was restricted to patients who had potentially curative resections with clear margins. The choice of an early endpoint for outcome assessment (6 months after surgery) made tumour recurrence an unlikely cause of symptoms17,18. The risk of misclassification of recalled weight at the time of surgery was alleviated by the good correlation between objectively measured weight and the corresponding recalled weight, as presented in a separate validation study5. Although the nationwide approach provided a large sample size, the study had limited power to ascertain small differences between the comparison groups. Although recurrent dysphagia due to anastomotic strictures and postoperative gastro-oesophageal reflux is not rare19, relatively few studies have addressed nutritional problems after oesophagectomy. Symptoms and inadequate dietary intake owing to appetite loss, eating difficulties and odynophagia have been linked to postoperative weight loss5, but these problems were not adjusted for in the present study as they were deemed to be in the causal pathway between the studied potential risk factors and risk of postoperative malnutrition. Although only a small proportion of patients received neoadjuvant therapy, the main finding of this study was a link between neoadjuvant therapy and postoperative malnutrition. Survival is the key endpoint in the curative treatment of oesophageal cancer, but malnutrition and health-related quality of life are also important. This might be of particular relevance in the absence of convincing data to support a strong survival benefit from neoadjuvant therapy11,12. It has been reported that preoperative chemoradiotherapy may negatively influence quality of life20 and that patients receiving neoadjuvant therapy experience pronounced fatigue before surgery18. Weight loss is common in oesophageal cancer and an average preoperative decrease in weight of 10 per cent is often mentioned in the literature. There is a lack of nutritional screening routines in Europe21, and assessment and treatment of malnutrition is carried out in only about half of all hospitalized patients. The present study has shown that increased preoperative weight loss is associated with a decreased risk of weight loss after surgery for oesophageal cancer. One explanation is that more attention is paid to counteract postoperative malnutrition among patients with the greatest preoperative weight loss. This suggests that unintentional weight loss is treatable, and recognition of potential risk factors for postoperative malnutrition is therefore important. It is possible that patients who have already lost much weight before surgery will have less surplus body mass from which to lose further weight, but this does not apply in the present study because adjustment was made for BMI. Patients with adenocarcinoma are often obese at diagnosis, and many patients were still overweight even with postoperative weight loss. In this context, it might need to be emphasized that involuntary postoperative weight loss is a marker of malnutrition even if the patient is obese or overweight, and weight loss is an important measure of malnutrition on its own22. It might be expected that total gastrectomy would be associated with greater malnutrition than oesophagectomy with retention of part of the stomach, but the study did not identify such a difference. Impaired function of an intrathoracic stomach, particularly delayed gastric emptying, might account for this23. Weight loss probably occurs mainly during the first 3 months after operation24,25, with a loss of body protein mass that is hard to regain, especially in older people26. The greater risk of postoperative weight loss in women might be explained by the tradition among the older generation for women to be responsible for the preparation of food27. The ideal of a slim body among women is another plausible explanation. Neoadjuvant therapy and female sex increase the risk of malnutrition after surgery for oesophageal and cardia cancer. Future research is needed, however, to determine whether weight is regained in the longer term after oesophagectomy, and whether the factors identified in this study are also associated with long-term malnutrition. Risk factors for postoperative malnutrition should be taken into consideration in the diagnostic phase, when planning treatment and during clinical follow-up. Acknowledgements Funding was provided by the Swedish Cancer Society and the National Board of Health and Welfare in Sweden. References 1 Viklund P , Lindblad M, Lu M, Ye W, Johansson J, Lagergren J. Risk factors for complications after esophageal cancer resection: a prospective population-based study in Sweden . Ann Surg 2006 ; 243 : 204 – 211 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Connors RC , Reuben BC, Neumayer LA, Bull DA. Comparing outcomes after transthoracic and transhiatal esophagectomy: a 5-year prospective cohort of 17 395 patients . J Am Coll Surg 2007 ; 205 : 735 – 740 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Daly JM , Fry WA, Little AG, Winchester DP, McKee RF, Stewart AK et al. Esophageal cancer: results of an American College of Surgeons Patient Care Evaluation Study . J Am Coll Surg 2000 ; 190 : 562 – 572 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Enzinger PC , Mayer RJ. Esophageal cancer . N Engl J Med 2003 ; 349 : 2241 – 2252 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Martin L , Lagergren J, Lindblad M, Rouvelas I, Lagergren P. Malnutrition after oesophageal cancer surgery in Sweden . Br J Surg 2007 ; 94 : 1496 – 1500 . Google Scholar Crossref Search ADS PubMed WorldCat 6 National Insitute for Health and Clinical Excellence . Nutrition Support in Adults: Oral Nutrition Support, Enteral Tube Feeding and Parenteral Nutrition. http://www.nice.org.uk/guidance/CG32/guidance/pdf/English/download.dspx [accessed 1 February 2006]. 7 Weimann A , Braga M, Harsanyi L, Laviano A, Ljungqvist O, Soeters P et al. ESPEN Guidelines on Enteral Nutrition: surgery including organ transplantation . Clin Nutr 2006 ; 25 : 224 – 244 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Stratton RJ , Green CJ, Elia M. Disease-related Malnutrition: an Evidence-based Approach to Treatment . CABI Publishing : Wallingford , 2003 . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC 9 Lochs H , Allison SP, Meier R, Pirlich M, Kondrup J, Schneider S et al. Introductory to the ESPEN Guidelines on Enteral Nutrition: terminology, definitions and general topics . Clin Nutr 2006 ; 25 : 180 – 186 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Correia MI , Waitzberg DL. The impact of malnutrition on morbidity, mortality, length of hospital stay and costs evaluated through a multivariate model analysis . Clin Nutr 2003 ; 22 : 235 – 239 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Gebski V , Burmeister B, Smithers BM, Foo K, Zalcberg J, Simes J. Survival benefits from neoadjuvant chemoradiotherapy or chemotherapy in oesophageal carcinoma: a meta-analysis . Lancet Oncol 2007 ; 8 : 226 – 234 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Malthaner RA , Collin S, Fenlon D. Preoperative chemotherapy for resectable thoracic esophageal cancer . Cochrane Database Syst Rev 2006 ; ( 3 ) CD001556 . Google Scholar OpenURL Placeholder Text WorldCat 13 Viklund P , Wengstrom Y, Rouvelas I, Lindblad M, Lagergren J. Quality of life and persisting symptoms after oesophageal cancer surgery . Eur J Cancer 2006 ; 42 : 1407 – 1414 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Lagergren J , Bergstrom R, Lindgren A, Nyren O. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma . N Engl J Med 1999 ; 340 : 825 – 831 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Sobin LH , Wittekind C. TNM Classification of Malignant Tumours (6th edn). Wiley-Liss : New York , 2002 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 16 Siewert JR , Stein HJ. Classification of adenocarcinoma of the oesophagogastric junction . Br J Surg 1998 ; 85 : 1457 – 1459 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 17 Zieren HU , Jacobi CA, Zieren J, Muller JM. Quality of life following resection of oesophageal carcinoma . Br J Surg 1996 ; 83 : 1772 – 1775 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Brooks JA , Kesler KA, Johnson CS, Ciaccia D, Brown JW. Prospective analysis of quality of life after surgical resection for esophageal cancer: preliminary results . J Surg Oncol 2002 ; 81 : 185 – 194 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Lagergren P , Avery KN, Hughes R, Barham CP, Alderson D, Falk SJ et al. Health-related quality of life among patients cured by surgery for esophageal cancer . Cancer 2007 ; 110 : 686 – 693 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Blazeby JM , Sanford E, Falk SJ, Alderson D, Donovan JL. Health-related quality of life during neoadjuvant treatment and surgery for localized esophageal carcinoma . Cancer 2005 ; 103 : 1791 – 1799 . Google Scholar Crossref Search ADS PubMed WorldCat 21 Elia M , Zellipour L, Stratton RJ. To screen or not to screen for adult malnutrition? Clin Nutr 2005 ; 24 : 867 – 884 . Google Scholar Crossref Search ADS PubMed WorldCat 22 Norman K , Pichard C, Lochs H, Pirlich M. Prognostic impact of disease-related malnutrition . Clin Nutr 2008 ; 27 : 5 – 15 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Holscher AH , Voit H, Buttermann G, Siewert JR. Function of the intrathoracic stomach as esophageal replacement . World J Surg 1988 ; 12 : 835 – 844 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Walther B , Johansson J, Johnsson F, Von Holstein CS, Zilling T. Cervical or thoracic anastomosis after esophageal resection and gastric tube reconstruction: a prospective randomized trial comparing sutured neck anastomosis with stapled intrathoracic anastomosis . Ann Surg 2003 ; 238 : 803 – 812 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Kiyama T , Mizutani T, Okuda T, Fujita I, Tokunaga A, Tajiri T et al. Postoperative changes in body composition after gastrectomy . J Gastrointest Surg 2005 ; 9 : 313 – 319 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Hickson M . Malnutrition and ageing . Postgrad Med J 2006 ; 82 : 2 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Sidenvall B , Nydahl M, Fjellstrom C. The meal as a gift—the meaning of cooking among retired women . J Appl Gerontol 2000 ; 19 : 405 – 423 . Google Scholar Crossref Search ADS WorldCat Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.

Furnée, E J B; Draaisma, W A; Broeders, I A M J; Smout, A J P M; Vlek, A L M; Gooszen, H G

doi: 10.1002/bjs.6346pmid: 18844266

Abstract Background Recurrent gastro-oesophageal reflux disease (GORD) and troublesome dysphagia after primary antireflux surgery are treated successfully by reoperation in 70 per cent of patients. Identifying predictors of outcome could allow selection of patients likely to benefit from further surgery. The aim was to identify such predictors in patients reoperated on for recurrent GORD or troublesome dysphagia. Methods Between 1994 and 2005, 83 patients (mean(s.d.) age 47·2(14·4) years; 47 men) with recurrent GORD and 47 (aged 50·7(13·4) years; 18 men) with troublesome dysphagia had further surgery. The predictive values of demographic, anatomical and manometric variables, and 24-h pH monitoring were analysed with respect to symptomatic and objective outcomes in each group. Results None of the factors included in a multivariable analysis predicted outcome after surgery for recurrent GORD. Independent predictors of symptomatic outcome after reoperation for dysphagia were amplitude of distal oesophageal contractions (odds ratio (OR) 1·613 (95 per cent confidence interval (c.i.) 1·087 to 2·393); P = 0·017), intrathoracic wrap migration (OR 0·077 (0·003 to 1·755); P = 0·108) and an abdominal approach (OR 0·012 (0·001 to 0·337); P = 0·009). Conclusion Low-amplitude distal oesophageal contractions, intrathoracic wrap migration and an abdominal approach were significant predictors of an unsuccessful symptomatic outcome after reoperation for troublesome dysphagia. Introduction Outcomes following reoperation for recurrent gastro-oesophageal reflux disease (GORD) or troublesome dysphagia after primary antireflux surgery have been reported recently by the present authors1. Successful symptomatic and objective outcomes were achieved in about 70 per cent of patients, comparable to results in other studies2–5. The reasons for failure after surgical reintervention remain unclear. Identification of factors that predict outcome may help in selection and tailoring the intervention as well as informing patients about the expected outcome of reoperation. Several studies have shown that the success rate of surgery for GORD decreases with the number of surgical reinterventions performed6,7. In one recent study, the presence of intrathoracic wrap migration was the only significant predictor of the need for further intervention8. In another, a preoperative diagnosis of Barrett's oesophagus, alone or in combination with oesophagitis, was a significant predictor of reducing regurgitation and controlling dysphagia. The need for Collis gastroplasty for a short oesophagus was a significant predictor of increased regurgitation9. In both studies, however, the number of predictive variables analysed was restricted and did not include objective factors established before reoperation, such as results of oesophageal manometry and 24-h pH monitoring. The aim of the present study was to identify preoperative and intraoperative predictors of symptomatic and objective outcome in patients having reoperation for recurrent GORD or troublesome dysphagia. Methods Between 1994 and 2005, 130 patients underwent surgical reintervention for recurrent GORD or troublesome dysphagia after previous antireflux surgery. Primary antireflux procedures were Nissen fundoplication in 110 patients (84·6 per cent), partial fundoplication in eight (6·2 per cent) and other procedures in 12 (9·2 per cent). After this primary intervention, 29 patients (22·3 per cent) had another surgical antireflux procedure before the re- intervention analysed in the present study. The indication for reintervention was recurrent reflux in 83 patients (63·8 per cent) (reflux group) and troublesome dysphagia in 47 (36·2 per cent) (dysphagia group). Of patients reoperated on for recurrent reflux, 30 (36 per cent) had a 360° Nissen fundoplication and one (1 per cent) a partial posterior fundoplication, all performed via an abdominal approach. The remaining 52 patients (63 per cent) had a Belsey mark IV 270° fundoplication via a left-sided thoracotomy. Reoperations for troublesome dysphagia included Belsey mark IV fundoplication in 23 patients, Nissen fundoplication in 13 and partial posterior fundoplication by an abdominal approach in 11. All patients underwent objective evaluation before reoperation comprising stationary oesophageal manometry and ambulatory 24-h pH monitoring1. A set of preoperative and intraoperative variables hypothesized to be of predictive value for symptomatic outcome was assessed separately in patients reoperated on for recurrent GORD (reflux group) and those who had surgery for troublesome dysphagia. The predictive value of these variables was determined in each group for improvement or resolution of preoperative symptoms, scored by the Visick grading system. In addition, their predictive value was determined for the development of new-onset dysphagia after reoperation for GORD, and for the recurrence of primary reflux symptoms after reoperation for dysphagia. Data on both postoperative symptoms were obtained using the Gastro-Esophageal Reflux Disease Health Related Quality of Life Scale10. In the reflux group, the predictive value of the variables was also analysed with respect to successful objective outcome, defined as the absence of pathological reflux (pH < 4 in the distal oesophagus for at most 5·8 per cent of total time, 8·2 per cent of time in the upright position and/or 3·5 per cent of time in the supine position) assessed by postoperative 24-h pH monitoring. Patient-specific characteristics (sex and age), number of previous antireflux procedures before reintervention (one or more than one), and surgical approach (abdominal or thoracic) were analysed for all outcome measures. Additionally, the predictive value of preoperative use of antisecretory drugs was determined for the effect on reflux symptoms and for objective outcome in the reflux group. The impact of anatomical abnormalities that might have led to failure of the previous antireflux procedure was determined. Intrathoracic wrap migration, wrap disruption, telescoping of the lower oesophageal sphincter (LOS) and paraoesophageal herniation were assessed as prognostic factors for all outcome measures in both groups. Further analyses were carried out after excluding the 29 patients who had already undergone at least one revisional operation. A series of objective variables, determined by oeso- phageal manometry and 24-h pH monitoring before re- operation, were evaluated. The manometric variables mean end-expiratory pressure of the LOS, hypotensive (mean LOS pressure below 0·6 kPa) or hypertensive (more than 3·5 kPa) LOS, incomplete relaxation of the LOS during swallowing (nadir LOS pressure at least 1·4 kPa), mean peak contraction amplitude in the distal part of the oesophagus, presence of low-amplitude (less than 8·5 kPa) and peristaltic wave contractions in the body of the oesophagus (peristaltic contractions during more than 70 per cent of swallowing) were analysed for their predictive value with respect to all outcome measures in both groups. The predictive value of the percentage of total, upright and supine time with pH < 4 in the distal oesophagus during preoperative 24-h pH monitoring, presence of pathological reflux during total time (more than 5·8 per cent), time in the upright position (more than 8·2 per cent) and in the supine position (more than 3·5 per cent)11, presence of isolated upright and supine pathological reflux, bipositional pathological reflux, number of reflux episodes, symptom index12 and symptom association probability13 were assessed with respect to reflux symptoms and objective outcome in patients reoperated on for recurrent GORD. Statistical analysis Values are expressed as mean(s.d.). Statistical analysis was performed using SPSS® version 12.0.1 for Windows® (SPSS, Chicago, Illinois, USA). For univariable analysis, all continuous and categorical variables were analysed separately by binary logistic regression analysis. Variables with P < 0·200 in univariable analysis were entered together into a multivariable analysis performed by binary logistic regression, using backward stepwise selection of variables. Variables with P < 0·100 in multivariable analysis were considered to be significant predictors of outcome. The odds ratio (OR), 95 per cent confidence interval (c.i.) and P value are presented for each variable. Receiver–operator characteristic (ROC) curves were constructed for outcome measures identified as significant predictors in multivariable analysis. The area under the curve (AUC) was calculated to determine the discriminative ability of the predictive model. The AUC was calculated with and without borderline significant predictors (variables with a P value between 0·100 and 0·150 in the multivariable analysis) and the predictive model with the best discriminative ability was chosen. To establish a score for predicting clinical outcome after reoperation in individual patients, points were assigned to independent predictive variables by dividing the logistic regression coefficients of all variables by the smallest coefficient and rounding each quotient to the nearest integer. Different cut-off points for the clinical prediction score were examined to find the score with the highest positive predictive value (fewest false positives). Results Patient characteristics and follow-up information are presented in Table 1. Overall, 118 (90·8 per cent) of 130 patients were available for symptomatic follow-up, and 63 (76 per cent) of 83 for objective follow-up after surgery for recurrent reflux. Of 43 patients in the dysphagia group, troublesome dysphagia was improved or resolved in 28 patients and daily symptoms of reflux were absent in 30. Among 75 patients reoperated on for recurrent reflux, preoperative symptoms were improved or resolved in 55 (73 per cent) and daily symptoms of dysphagia were absent in 66 (88 per cent). Pathological reflux was absent at follow-up in 42 (67 per cent) of 63 patients. Table 1 Baseline characteristics . Reflux group (n = 83) . Dysphagia group (n = 47) . Mean(s.d.) age (years) 47·2(14·4) 50·7(13·4) Sex ratio (M : F) 47 : 36 18 : 29 Mean(s.d.) time to follow-up  (months)  Subjective 61·7(37·3) 57·2(37·1)  Objective 20·3(32·8) — No. of patients at follow-up  Subjective 75 (90) 43 (91)  Objective 63 (76) — . Reflux group (n = 83) . Dysphagia group (n = 47) . Mean(s.d.) age (years) 47·2(14·4) 50·7(13·4) Sex ratio (M : F) 47 : 36 18 : 29 Mean(s.d.) time to follow-up  (months)  Subjective 61·7(37·3) 57·2(37·1)  Objective 20·3(32·8) — No. of patients at follow-up  Subjective 75 (90) 43 (91)  Objective 63 (76) — Values in parentheses are percentages unless indicated otherwise. Open in new tab Table 1 Baseline characteristics . Reflux group (n = 83) . Dysphagia group (n = 47) . Mean(s.d.) age (years) 47·2(14·4) 50·7(13·4) Sex ratio (M : F) 47 : 36 18 : 29 Mean(s.d.) time to follow-up  (months)  Subjective 61·7(37·3) 57·2(37·1)  Objective 20·3(32·8) — No. of patients at follow-up  Subjective 75 (90) 43 (91)  Objective 63 (76) — . Reflux group (n = 83) . Dysphagia group (n = 47) . Mean(s.d.) age (years) 47·2(14·4) 50·7(13·4) Sex ratio (M : F) 47 : 36 18 : 29 Mean(s.d.) time to follow-up  (months)  Subjective 61·7(37·3) 57·2(37·1)  Objective 20·3(32·8) — No. of patients at follow-up  Subjective 75 (90) 43 (91)  Objective 63 (76) — Values in parentheses are percentages unless indicated otherwise. Open in new tab Univariable analysis Univariable analysis of all potential predictive factors was performed with respect to outcome measures outlined above in each group. In patients reoperated on for recurrent reflux, none of the variables predicted the development of dysphagia after reoperation. Variables that significantly predicted the effect of reoperation on preoperative symptoms or objective outcome are shown in Table 2. Table 2 Results of univariable analysis in patients reoperated on for recurrent reflux . Odds ratio . P* . Improvement or resolution of preoperative symptoms  Female 3·346 (1·068, 10·486) 0·029  Abdominal approach during reoperation 2·323 (0·740, 7·290) 0·135  Intrathoracic wrap migration 0·417 (0·143, 1·216) 0·111  Preoperative hypotensive lower oesophageal sphincter pressure 0·356 (0·092, 1·378) 0·141  Preoperative contraction amplitude of distal oesophagus (kPa) 1·136 (0·960, 1·344) 0·114  Preoperative upright pathological reflux 0·400 (0·122, 1·311) 0·130  Preoperative symptom association probability > 95% 2·368 (0·628, 8·926) 0·190 Absence of pathological reflux at follow-up  Intrathoracic wrap migration 0·443 (0·141, 1·397) 0·166  Preoperative use of antisecretory drugs 6·882 (0·667, 71·003) 0·077  Preoperative symptom association probability > 95% 1·010 (0·997, 1·023) 0·146  Preoperative isolated supine pathological reflux 0·370 (0·108, 1·273) 0·113 . Odds ratio . P* . Improvement or resolution of preoperative symptoms  Female 3·346 (1·068, 10·486) 0·029  Abdominal approach during reoperation 2·323 (0·740, 7·290) 0·135  Intrathoracic wrap migration 0·417 (0·143, 1·216) 0·111  Preoperative hypotensive lower oesophageal sphincter pressure 0·356 (0·092, 1·378) 0·141  Preoperative contraction amplitude of distal oesophagus (kPa) 1·136 (0·960, 1·344) 0·114  Preoperative upright pathological reflux 0·400 (0·122, 1·311) 0·130  Preoperative symptom association probability > 95% 2·368 (0·628, 8·926) 0·190 Absence of pathological reflux at follow-up  Intrathoracic wrap migration 0·443 (0·141, 1·397) 0·166  Preoperative use of antisecretory drugs 6·882 (0·667, 71·003) 0·077  Preoperative symptom association probability > 95% 1·010 (0·997, 1·023) 0·146  Preoperative isolated supine pathological reflux 0·370 (0·108, 1·273) 0·113 Values in parentheses are 95 per cent confidence intervals. * Binary logistic regression analysis. Open in new tab Table 2 Results of univariable analysis in patients reoperated on for recurrent reflux . Odds ratio . P* . Improvement or resolution of preoperative symptoms  Female 3·346 (1·068, 10·486) 0·029  Abdominal approach during reoperation 2·323 (0·740, 7·290) 0·135  Intrathoracic wrap migration 0·417 (0·143, 1·216) 0·111  Preoperative hypotensive lower oesophageal sphincter pressure 0·356 (0·092, 1·378) 0·141  Preoperative contraction amplitude of distal oesophagus (kPa) 1·136 (0·960, 1·344) 0·114  Preoperative upright pathological reflux 0·400 (0·122, 1·311) 0·130  Preoperative symptom association probability > 95% 2·368 (0·628, 8·926) 0·190 Absence of pathological reflux at follow-up  Intrathoracic wrap migration 0·443 (0·141, 1·397) 0·166  Preoperative use of antisecretory drugs 6·882 (0·667, 71·003) 0·077  Preoperative symptom association probability > 95% 1·010 (0·997, 1·023) 0·146  Preoperative isolated supine pathological reflux 0·370 (0·108, 1·273) 0·113 . Odds ratio . P* . Improvement or resolution of preoperative symptoms  Female 3·346 (1·068, 10·486) 0·029  Abdominal approach during reoperation 2·323 (0·740, 7·290) 0·135  Intrathoracic wrap migration 0·417 (0·143, 1·216) 0·111  Preoperative hypotensive lower oesophageal sphincter pressure 0·356 (0·092, 1·378) 0·141  Preoperative contraction amplitude of distal oesophagus (kPa) 1·136 (0·960, 1·344) 0·114  Preoperative upright pathological reflux 0·400 (0·122, 1·311) 0·130  Preoperative symptom association probability > 95% 2·368 (0·628, 8·926) 0·190 Absence of pathological reflux at follow-up  Intrathoracic wrap migration 0·443 (0·141, 1·397) 0·166  Preoperative use of antisecretory drugs 6·882 (0·667, 71·003) 0·077  Preoperative symptom association probability > 95% 1·010 (0·997, 1·023) 0·146  Preoperative isolated supine pathological reflux 0·370 (0·108, 1·273) 0·113 Values in parentheses are 95 per cent confidence intervals. * Binary logistic regression analysis. Open in new tab Significant predictors of the effect of reoperation for troublesome dysphagia on preoperative symptoms are presented in Table 3. Age was the only variable with predictive value regarding recurrent reflux symptoms after reoperation in this group (Table 3). Table 3 Results of univariable analysis in patients reoperated on for troublesome dysphagia . Odds ratio . P* . Improvement or resolution of preoperative symptoms  Preoperative contraction amplitude of distal oesophagus (kPa) 1·228 (0·977, 1·543) 0·049  Abdominal approach during reoperation 0·273 (0·069, 1·070) 0·052  Intrathoracic wrap migration 0·326 (0·079, 1·342) 0·119 Recurrence of primary daily reflux symptoms after reoperation  Age (years) 0·958 (0·905, 1·014) 0·125 . Odds ratio . P* . Improvement or resolution of preoperative symptoms  Preoperative contraction amplitude of distal oesophagus (kPa) 1·228 (0·977, 1·543) 0·049  Abdominal approach during reoperation 0·273 (0·069, 1·070) 0·052  Intrathoracic wrap migration 0·326 (0·079, 1·342) 0·119 Recurrence of primary daily reflux symptoms after reoperation  Age (years) 0·958 (0·905, 1·014) 0·125 Values in parentheses are 95 per cent confidence intervals. * Binary logistic regression analysis. Open in new tab Table 3 Results of univariable analysis in patients reoperated on for troublesome dysphagia . Odds ratio . P* . Improvement or resolution of preoperative symptoms  Preoperative contraction amplitude of distal oesophagus (kPa) 1·228 (0·977, 1·543) 0·049  Abdominal approach during reoperation 0·273 (0·069, 1·070) 0·052  Intrathoracic wrap migration 0·326 (0·079, 1·342) 0·119 Recurrence of primary daily reflux symptoms after reoperation  Age (years) 0·958 (0·905, 1·014) 0·125 . Odds ratio . P* . Improvement or resolution of preoperative symptoms  Preoperative contraction amplitude of distal oesophagus (kPa) 1·228 (0·977, 1·543) 0·049  Abdominal approach during reoperation 0·273 (0·069, 1·070) 0·052  Intrathoracic wrap migration 0·326 (0·079, 1·342) 0·119 Recurrence of primary daily reflux symptoms after reoperation  Age (years) 0·958 (0·905, 1·014) 0·125 Values in parentheses are 95 per cent confidence intervals. * Binary logistic regression analysis. Open in new tab Multivariable analysis In patients reoperated on for recurrent reflux, none of the predictive factors identified by univariable analysis was shown by multivariable analysis to be an independent predictor of either the effect of reoperation on reflux symptoms or objective outcome. In patients reoperated on for troublesome dysphagia, age was the only significant predictor of recurrent reflux symptoms after reoperation identified by univariable analysis, and so a predictive model could not be constructed for this outcome. With regard to the effect of reoperation on preoperative symptoms, two variables with significant predictive value by univariable analysis remained significant independent predictors of outcome in the multivariable analysis (Table 4). The odds of improvement or resolution of symptoms after reoperation significantly increased with higher preoperative amplitude of distal oesophageal contractions, whereas an abdominal approach to revisional surgery was a significant predictor of persistent dysphagia. Herniation of the wrap disappeared as a significant predictor when the analysis excluded the 29 patients who had undergone at least one previous revisional operation. As intrathoracic wrap migration was identified as a borderline significant predictor of persistent dysphagia, predictive models were tested with inclusion and exclusion of this variable. AUC values were 0·917 (95 per cent c.i. 0·814 to 1·019; P < 0·001) and 0·886 (95 per cent c.i. 0·764 to 1·008; P < 0·001) respectively, indicating that the first model had the best discriminative ability. Table 4 Results of multivariable analysis in patients reoperated on for troublesome dysphagia with respect to improvement or resolution of symptoms . Coefficient* . Odds ratio† . P# . Points‡ . Preoperative contraction amplitude of distal oesophagus (kPa) 0·478 (0·201) 1·613 (1·087, 2·393) 0·017 1§ Abdominal route of reoperation 4·444 (1·713) 0·012 (0·001, 0·337) 0·009 9 Intrathoracic wrap migration¶ 2·559 (1·593) 0·077 (0·003, 1·755) 0·108 5 . Coefficient* . Odds ratio† . P# . Points‡ . Preoperative contraction amplitude of distal oesophagus (kPa) 0·478 (0·201) 1·613 (1·087, 2·393) 0·017 1§ Abdominal route of reoperation 4·444 (1·713) 0·012 (0·001, 0·337) 0·009 9 Intrathoracic wrap migration¶ 2·559 (1·593) 0·077 (0·003, 1·755) 0·108 5 Values in parentheses are * s.e. or † 95 per cent confidence interval. ‡ Points were assigned by dividing the coefficient of each variable by 0·478 (the smallest coefficient), rounding each value to the nearest integer. § Preoperative amplitude of the distal oesophagus equals the number of points assigned for this continuous variable. ¶ A borderline significant predictor in multivariable analysis. # Binary logistic regression analysis, with backward stepwise selection of variables. Open in new tab Table 4 Results of multivariable analysis in patients reoperated on for troublesome dysphagia with respect to improvement or resolution of symptoms . Coefficient* . Odds ratio† . P# . Points‡ . Preoperative contraction amplitude of distal oesophagus (kPa) 0·478 (0·201) 1·613 (1·087, 2·393) 0·017 1§ Abdominal route of reoperation 4·444 (1·713) 0·012 (0·001, 0·337) 0·009 9 Intrathoracic wrap migration¶ 2·559 (1·593) 0·077 (0·003, 1·755) 0·108 5 . Coefficient* . Odds ratio† . P# . Points‡ . Preoperative contraction amplitude of distal oesophagus (kPa) 0·478 (0·201) 1·613 (1·087, 2·393) 0·017 1§ Abdominal route of reoperation 4·444 (1·713) 0·012 (0·001, 0·337) 0·009 9 Intrathoracic wrap migration¶ 2·559 (1·593) 0·077 (0·003, 1·755) 0·108 5 Values in parentheses are * s.e. or † 95 per cent confidence interval. ‡ Points were assigned by dividing the coefficient of each variable by 0·478 (the smallest coefficient), rounding each value to the nearest integer. § Preoperative amplitude of the distal oesophagus equals the number of points assigned for this continuous variable. ¶ A borderline significant predictor in multivariable analysis. # Binary logistic regression analysis, with backward stepwise selection of variables. Open in new tab Points were assigned to these three variables, based on the logistic regression coefficients, to establish a clinical prediction score for each patient. One point was assigned to contraction amplitude of the distal oesophagus because this variable had the smallest coefficient (Table 4); the number of points equalled the amplitude (1 point equivalent to 1 kPa). Based on the logistic regression coefficients of the other two predictors, 5 points were subtracted if the wrap had migrated intrathoracically and 9 if an abdominal approach had been used. The mean(s.d.) clinical prediction score was 1·6(5·3). Different cut-off points were examined and a clinical prediction score of 2 was chosen for the final predictive model, because this cut-off yielded a positive predictive value of 93 per cent (13 of 14 patients) and a negative predictive value of 65 per cent (11 of 17) for outcome after reoperation for troublesome dysphagia (Table 5). Table 5 Symptomatic outcome prediction in patients reoperated on for troublesome dysphagia according to clinical prediction score . Successful outcome . Unsuccessful outcome . Clinical prediction score ≥ 2 13 of 14 1 of 14 Clinical prediction score < 2 6 of 17 11 of 17 . Successful outcome . Unsuccessful outcome . Clinical prediction score ≥ 2 13 of 14 1 of 14 Clinical prediction score < 2 6 of 17 11 of 17 Open in new tab Table 5 Symptomatic outcome prediction in patients reoperated on for troublesome dysphagia according to clinical prediction score . Successful outcome . Unsuccessful outcome . Clinical prediction score ≥ 2 13 of 14 1 of 14 Clinical prediction score < 2 6 of 17 11 of 17 . Successful outcome . Unsuccessful outcome . Clinical prediction score ≥ 2 13 of 14 1 of 14 Clinical prediction score < 2 6 of 17 11 of 17 Open in new tab Discussion None of the anatomical and physiological factors analysed in this study proved to be of predictive value for symptomatic and objective outcome in patients reoperated on for recurrent reflux. Only three factors—preoperative amplitude of distal oesophageal contractions, operative approach and intrathoracic wrap migration—were significant predictors of the effect of reoperation on dysphagia. Smith and co-workers8 also identified intrathoracic wrap migration as a significant predictor of outcome after reoperation. For primary antireflux surgery, others found that complete or partial response to acid suppression therapy and abnormal score on 24-h oesophageal pH monitoring significantly predicted successful symptomatic outcome14,15. Winslow and co-workers16 noted that patients with isolated upright reflux had a worse symptomatic outcome after primary antireflux surgery than those with isolated supine or bipositional reflux. These factors were also analysed in the present study, but none was identified as a predictor of outcome after reoperation for recurrent GORD. Several studies of primary Nissen fundoplication have shown that impaired motility, defined as inadequate peristalsis and/or low-amplitude contraction of the distal part of the oesophagus, is not associated with higher rates of dysphagia at follow-up, even if a total fundoplication is created17–19. This contrasts with the present finding that the contraction amplitude of the distal oesophagus was an independent predictor of the effect of reoperation on dysphagia. Apparently, there is a subgroup of patients who cannot be identified before primary surgery, but are probably at risk of developing dysphagia from the start if construction of the wrap was not optimal (too long or not ‘floppy’ enough) or the hiatal repair was (a little) too tight. Therefore, surgeons should be aware that the outcome of reoperation may be disappointing in terms of postoperative dysphagia in patients with low-amplitude contraction of the distal oesophagus established before reoperation. Herniation of the wrap and an abdominal approach appeared to be independent predictors of an unsuccessful symptomatic outcome in patients reoperated on for dysphagia. Wrap migration was analysed because it had previously been described as a negative predictor of outcome8, although there is no obvious explanation for its predictive value. The predictive value of the abdominal approach might be because attention is directed particularly at the wrap as a possible cause of troublesome dysphagia during abdominal reoperation, where inspection of the hiatus may be restricted. With a thoracic approach, the posterior crural closure sutures have to be removed to mobilize the (remnants of) the primary wrap into the thorax and create a new wrap, which is subsequently replaced under the diaphragm. Afterwards, the hiatus is closed again, with potentially better exposure than during abdominal reoperation. It cannot be inferred from the present results that the thoracic approach is superior. It should be acknowledged, however, that control of the hiatus is a crucial part of revisional surgery. Another factor potentially contributing to the predictive value of the abdominal approach could be that thoracic reoperations were performed by the senior author, who has extensive experience in and a preference for this approach. Based on the scoring system described in the present study, an abdominal approach is likely to be avoided in patients with intrathoracic wrap migration. Indeed, the policy in this hospital is to select a thoracic approach if preoperative endoscopy and barium swallow indicate that intrathoracic wrap migration has occurred. Preoperative investigations, however, cannot reliably assess the anatomical situation at the level of the hiatus. Therefore, laparoscopy is used as the first step in deciding whether to continue the procedure laparoscopically, through a laparotomy or by conversion to thoracotomy. The model developed in this study was based on a relatively small number of patients and the results should be interpreted with caution as there may have been a risk of overestimation of the identified significant predictors of outcome. In contrast, the lack of significance of the other variables analysed may also be explained by the limited sample size. The present findings require validation in a larger population. References 1 Furnée EJ , Draaisma WA, Broeders IA, Smout AJ, Gooszen HG. Surgical reintervention after antireflux surgery for gastroesophageal reflux disease: a prospective cohort study in 130 patients . Arch Surg 2008 ; 143 : 267 – 274 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Awad ZT , Anderson PI, Sato K, Roth TA, Gerhardt J, Filipi CJ. Laparoscopic reoperative antireflux surgery . Surg Endosc 2001 ; 15 : 1401 – 1407 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Bais JE , Horbach TL, Masclee AA, Smout AJ, Terpstra JL, Gooszen HG. Surgical treatment for recurrent gastro-oesophageal reflux disease after failed antireflux surgery . Br J Surg 2000 ; 87 : 243 – 249 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Deschamps C , Trastek VF, Allen MS, Pairolero PC, Johnson JO, Larson DR. Long-term results after reoperation for failed antireflux procedures . J Thorac Cardiovasc Surg 1997 ; 113 : 545 – 550 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Stirling MC , Orringer MB. Surgical treatment after the failed antireflux operation . J Thorac Cardiovasc Surg 1986 ; 92 : 667 – 672 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Little AG , Ferguson MK, Skinner DB. Reoperation for failed antireflux operations . J Thorac Cardiovasc Surg 1986 ; 91 : 511 – 517 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Skinner DB . Surgical management after failed antireflux operations . World J Surg 1992 ; 16 : 359 – 363 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Smith CD , McClusky DA, Rajad MA, Lederman AB, Hunter JG. When fundoplication fails: redo? Ann Surg 2005 ; 241 : 861 – 869 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Iqbal A , Awad Z, Simkins J, Shah R, Haider M, Salinas V et al. Repair of 104 failed antireflux operations . Ann Surg 2006 ; 244 : 42 – 51 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Velanovich V , Vallance SR, Gusz JR, Tapia FV, Harkabus MA. Quality of life scale for gastroesophageal reflux disease . J Am Coll Surg 1996 ; 183 : 217 – 224 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 11 Richter JE , Bradley LA, DeMeester TR, Wu WC. Normal 24-hr ambulatory esophageal pH values. Influence of study center, pH electrode, age, and gender . Dig Dis Sci 1992 ; 37 : 849 – 856 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Wiener GJ , Richter JE, Copper JB, Wu WC, Castell DO. The symptom index: a clinically important parameter of ambulatory 24-hour esophageal pH monitoring . Am J Gastroenterol 1988 ; 83 : 358 – 361 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 13 Weusten BL , Roelofs JM, Akkermans LM, van Berge-Henegouwen GP, Smout AJ. The symptom-association probability: an improved method for symptom analysis of 24-hour esophageal pH data . Gastroenterology 1994 ; 107 : 1741 – 1745 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Campos GM , Peters JH, DeMeester TR, Oberg S, Crookes PF, Tan S et al. Multivariate analysis of factors predicting outcome after laparoscopic Nissen fundoplication . J Gastrointest Surg 1999 ; 3 : 292 – 300 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Jackson PG , Gleiber MA, Askari R, Evans SR. Predictors of outcome in 100 consecutive laparoscopic antireflux procedures . Am J Surg 2001 ; 181 : 231 – 235 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Winslow ER , Frisella MM, Soper NJ, Clouse RE, Klingensmith ME. Patients with upright reflux have less favorable postoperative outcomes after laparoscopic antireflux surgery than those with supine reflux . J Gastrointest Surg 2002 ; 6 : 819 – 829 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Beckingham IJ , Cariem AK, Bornman PC, Callanan MD, Louw JA. Oesophageal dysmotility is not associated with poor outcome after laparoscopic Nissen fundoplication . Br J Surg 1998 ; 85 : 1290 – 1293 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Munitiz V , Ortiz A, Martinez de Haro LF, Molina J, Parrilla P. Ineffective oesophageal motility does not affect the clinical outcome of open Nissen fundoplication . Br J Surg 2004 ; 91 : 1010 – 1014 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Zornig C , Strate U, Fibbe C, Emmermann A, Layer P. Nissen vs Toupet laparoscopic fundoplication . Surg Endosc 2002 ; 16 : 758 – 766 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) Copyright © 2008 British Journal of Surgery Society Ltd. Published by John Wiley & Sons, Ltd.