JOURNAL TRANSCRIPT

---

Intestinal cell damage, inflammation and wound healing in major gastrointestinal surgery Kostan Reisinger

© Copyright K.W. Reisinger, Maastricht 2015 ISBN 978 94 6159 439 6 Cover design: Christine Muris Cover: “What goes on four feet, on two feet, and three, but the more feet it goes on, the weaker it be?” Sophocles - Oedipus Rex Production: Datawyse Maastricht Printing of this thesis was financially supported by the Nederlandse Vereniging voor Gastroenterologie. The studies in this thesis were performed at the Nutrition and Toxicology Research Institute Maastricht (NUTRIM).

All rights reserved. No part of this thesis may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage or retrieval system, without permission in writing from the author, or, when appropriate, from the publishers of publications.

Intestinal cell damage, inflammation and wound healing in major gastrointestinal surgery PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Universiteit Maastricht, op gezag van de Rector Magnificus, Prof. dr. L.L.G. Soete, volgens het besluit van het College van Decanen, in het openbaar te verdedigen op vrijdag 5 juni 2015 om 14:00 uur door

Kostan Werner Reisinger geboren op 16 oktober 1985 te Muiden

P

UM UNIVERSITAIRE

PERS MAASTRICHT

Promotores Prof. dr. M.F. von Meyenfeldt Prof. dr. L.W.E. van Heurn

Copromotor Dr. M. Poeze

Beoordelingscommissie Prof. dr. C.H.C. Dejong (voorzitter) Dr. N.D. Bouvy Prof. dr. K.C.H. Fearon, University of Edinburgh Prof. dr. E. Heineman, Universitair Medisch Centrum Groningen Dr. K. Lenaerts

Aan mijn ouders Voor Aleida, Nina en Max

Table of contents Chapter 1. General introduction

9

Part 1. Pathophysiologic aspects of complications after adult gastrointestinal surgery

23

Chapter 2. Doppler guided goal-directed fluid therapy increases intestinal perfusion in colorectal surgery

25

Chapter 3. Cyclooxygenase-2 is essential for colorectal anastomotic healing

45

Chapter 4. Functional compromise reflected by sarcopenia, frailty and nutritional depletion predicts adverse postoperative outcome after colorectal cancer surgery

61

Chapter 5. Sarcopenia is associated with an increased inflammatory response to surgery in colorectal cancer

83

Chapter 6. Loss of skeletal muscle mass during neoadjuvant therapy predicts postoperative mortality in esophageal cancer surgery

97

Part 2. Biomarkers of complications after adult gastrointestinal surgery

111

Chapter 7.

Accurate prediction of anastomotic leakage after colorectal surgery using plasma markers for intestinal damage and inflammation

113

Part 3. Pathophysiologic aspects of necrotizing enterocolitis (NEC)

131

Chapter 8. Intestinal fatty acid binding protein (I-FABP): a possible marker for gut maturation

133

Chapter 9. Breast feeding improves gut maturation compared to formula feeding in preterm babies

153

Part 4. Biomarkers of necrotizing enterocolitis (NEC)

169

Chapter 10. Non-invasive measurement of fecal calprotectin and serum amyloid A (SAA) combined with intestinal Fatty Acid Binding Protein (I-FABP) in necrotizing enterocolitis (NEC)

171

Chapter 11. Non-invasive serum amyloid A (SAA) measurement and plasma platelets for accurate prediction of surgical intervention in severe necrotizing enterocolitis (NEC)

187

Chapter 12. Non-invasive measurement of intestinal epithelial damage at time of re-feeding can predict clinical outcome after necrotizing enterocolitis

207

Chapter 13. Summary and discussion

221

Valorization

237

Samenvatting

241

Dankwoord

249

Scientific output

261

Curriculum vitae

265

Chapter 1. General introduction

Partly adapted from: Jeroen L.A. van Vugt, Kostan W. Reisinger, Joep P.M. Derikx, Djamila Boerma, Jan H.M.B. Stoot. Improving outcomes in oncological colorectal surgery. World Journal of Gastroenterology 2014. 20(35):12445-57

9

Chapter 1

1.1 Introduction Major gastrointestinal surgery is the treatment of choice in a variety of gastrointestinal diseases, such as cancer and severe inflammatory conditions of the gut. This type of surgery carries significant risk of complications, which negatively affects treatment effectiveness, survival, health-related quality of life and healthcare costs 1, 2. Furthermore, gut functioning has a central role in patients recovering from different kinds of insults, such as surgery. First, adequate digestion and absorption of nutrients is necessary to recover from the compromised physical condition imposed by surgery and anesthesia. Second, integrity of the gut barrier is essential to avoid exposure to luminal bacteria and other toxins related to postoperative complications 3-5. Compromise of these functions may lead to inadequate wound healing, postoperative morbidity and mortality. The search for improvements in perioperative care is therefore of crucial importance and should aim at fundamental aspects, prevention, and diagnosis of impaired gut function and complications. In this thesis, it is hypothesized that especially in vulnerable patients with gastrointestinal disorders, inadequate wound healing, postoperative morbidity and mortality are likely to develop. To investigate this hypothesis, patients at the extremes of age were primarily investigated, i.e. elderly patients and premature neonates. In paragraph 1.2, opportunities for improvement in adult oncological gastrointestinal surgery are discussed. First, important aspects of vulnerability of elderly patients such as frailty and nutritional compromise are considered in paragraphs 1.2.1 and 1.2.2. Second, clinical and laboratory markers are indispensible tools for timely recognition of postoperative complications, which is described in paragraph 1.2.3. Third, perioperative optimization strategies have led to improved surgical outcome in the past decennia. Several elements of perioperative optimization are however subject to debate, which necessitates further investigation as shown in paragraph 1.2.4. In paragraph 1.3, another vulnerable patient population is considered, namely premature neonates. Necrotizing enterocolitis (NEC) is the most severe gastrointestinal emergency in these patients and one of the primary indications for pediatric gastrointestinal surgery. Possibilities to improve the diagnosis, treatment and pathophysiological knowledge of NEC are discussed.

10

Introduction

1.2 Oncological gastrointestinal surgery Colorectal cancer is one of the predominant types of cancer and the fourth leading cause of cancer-associated deaths worldwide 6. In numbers: 600,000 patients died of colorectal cancer in 2008 worldwide, and disability-adjusted life-years lost from colorectal cancer were 300 per 100,000 patients, which was estimated to be 7% of the total cancer burden worldwide 7. Sixty-six per cent of patients with colorectal cancer will undergo at least one major surgical resection 8 . The perioperative course of colorectal surgery for malignancy is crucial for the clinical outcome of treatment, in terms of mortality, tolerance, efficacy, and functional recovery, and has a considerable impact on health care resources 9, 10. In the past decennia, perioperative care has improved largely due to advances in anesthetic and analgesic approaches, minimally invasive operative techniques, and the introduction of fast-track protocols 11, 12. Despite these improvements, many complications are still observed after oncological colorectal surgery, leading to prolonged hospital stay and high readmission rates with concurrent health care costs 2. Early recognition and adequate intervention of complications will attenuate severity and may eventually prevent mortality. Anastomotic leakage is among the most prevalent and detrimental complications of colorectal surgery and is a severe form of gut function disruption. Of 10,017 registered resections for colorectal cancer in the Netherlands in 2012, 691 (6.9%) were complicated by anastomotic leakage requiring re-intervention (Dutch Surgical Colorectal Audit 2012) 13, making anastomotic leakage the primary complication requiring re-intervention. In high-risk patients incidence rates can even increase to 18% 14. Anastomotic leakage is associated with high morbidity 15, mortality 16, reoperation rates 10, and duration of hospitalization 17. In cancer, anastomotic leakage is related to diminished disease-specific survival and higher recurrence rates 1, 10, 18. It is therefore imperative to find new strategies to prevent, diagnose and treat anastomotic leakage, or implement what is already known. 1.2.1 Frailty and sarcopenia Advanced age is associated with an increased incidence of cancer 19. The number of elderly cancer patients is concomitantly increasing. Fifty per cent of patients with colorectal cancer is above the age of 70 years 13. While survival of all cancer types is increasing, the improvement of cancer outcome in older patients is relatively limited 20. Higher age is an independent predictor of disease-specific

11

Chapter 1

perioperative mortality in patients undergoing surgery for colorectal cancer 21, 22. Weight loss, cachexia and nutritional compromise, especially in older patients, are associated with impaired response to chemotherapy and decreased survival 23, 24. Frail elderly undergoing colorectal surgery have a 4-fold increased risk of major postoperative complications 25. Frailty is a state of increased vulnerability towards stressors in older individuals, leading to an increased risk of developing adverse health outcomes 26. The definitions and biological characteristics of frailty are subject to debate. Weight loss, decreased muscle strength, reduced physical activity, exhaustion, and reduced walking speed are symptoms of a physical definition of frailty 25, 27, whereas comorbidity, polypharmacy, decreased physical functioning, impaired nutritional and cognitive status, depression and social support are components of a more multidimensional description of frailty 28. A simple screening instrument for frailty is the Groningen Frailty Index (GFI), based on physical, cognitive, social and emotional items 29. Skeletal muscle depletion or sarcopenia is an element of frailty in both definitions. Sarcopenia, which can easily be assessed by measurement of muscle area at the level of the third lumbar spine at CT-scan, is associated with prolonged hospital stay, infectious complications and decreased recurrence and survival rates following colorectal and liver surgery 30, 31. However, a relationship with anastomotic leakage has not yet been established. 1.2.2 Nutritional status A powerful and easily obtainable tool to assess the patient’s physical and/or mental condition before operation is the use of questionnaires. Various questionnaires have been developed to evaluate nutritional status. A poor nutritional condition correlates well with impaired quality of life and physical functioning 32. The Short Nutritional Assessment Questionnaire (SNAQ) and Malnutrition Universal Screening Tool (MUST) scores are commonly used nutritional screening tools in surgical patients. These questionnaires accurately detect malnutrition, and the MUST score predicts postoperative complications in cardiac surgery 33. Evidence for the value of nutritional screening tools to predict postoperative outcome in colorectal surgery is lacking. As one in five patients undergoing colorectal surgery is malnourished 34, the detection of nutritional depletion is of great importance, especially with neo-adjuvant therapies which can further compromise the nutritional and metabolic status. A misbalance between energy expenditure and nutritional supplementation, combined with metabolic inefficiency leading

12

Introduction

to protein breakdown particulary in muscle are the fundamental physiologic derangements leading to cancer-induced weight loss. The combination of energy imbalance and metabolic derangements are associated with poor clinical outcome after surgery 35, 36. This may at least partly be related to impaired wound healing when metabolic substrates are depleted. Although nutritional supplementation strategies in oncological colorectal surgery can improve handgrip strength, pulmonary function and insulin resistance 37 , nutritional support has not been proven unequivocally effective to reduce length of hospital stay and anastomotic leakage rates 38. It may be concluded that only severely malnourished patients benefit from nutritional support 39, 40. Nutritional status questionnaires may however not only be used to identify malnourished patients for nutritional support. As a tool for accurate prediction of postoperative complications, SNAQ and MUST scores could lead the way to other treatment options, for example surgery without a primary anastomosis or protection of the anastomosis using a diverting stoma. The predictive value of these scores for postoperative complications remains yet to be determined. 1.2.3 Clinical and laboratory detection of anastomotic leakage The clinical presentation of anastomotic leakage is heterogeneous and often nonspecific. Anastomotic leakage is therefore frequently diagnosed late 41. Furthermore, some leaks develop subclinically and are only detectable with radiological examination. Abdominal CT scan with intraluminal contrast has considerably improved timely recognition of anastomotic leakage, although it yields low sensitivity (68%), which may delay the diagnosis and appropriate treatment 42. Delay in recognizing and consequently treating anastomotic leakage after colorectal surgery is associated with increased mortality 43, 44. Clinical signs for accurate and early detection of anastomotic leakage have been widely investigated. Den Dulk and colleagues standardized postoperative monitoring and developed a leakage-score, consisting of general, local physical examination, laboratory investigation and dietary items. The use of this score resulted in a significantly shorter delay in the diagnosis of anastomotic leakage 45. Accurate diagnostic markers are needed to detect anastomotic leakage early after colorectal surgery. Various biomarkers have been investigated, although none has been validated clinically and studies are difficult to compare, mainly due to different definitions of anastomotic leakage 46. C-reactive protein (CRP) has been widely proposed as an early indicator to diagnose anastomotic leakage

13

Chapter 1

on postoperative day 2-4 47. However, the test characteristics are not convincingly robust, with approximately 70-80% sensitivity and specificity 47. Currently, the PRECIOUS trial investigates a step-up approach in major abdominal surgery combining CRP and CT imaging of the abdomen to diagnose severe complications, including anastomotic leakage 48. Specific plasma markers for intestinal cell damage and inflammation may provide better accuracy. 1.2.4 Perioperative optimization strategies Several meta-analyses have shown that Enhanced Recovery After Surgery (ERAS) programs result in reduced length of hospital stay and overall complications without affecting patient safety 12, 49, 50. Although strong evidence exists for many recommendations, such as antibiotic prophylaxis and preoperative bowel preparation, controversies remain around perioperative fluid therapy and use of non-steroidal anti-inflammatory drugs (NSAIDs) 39, 51. To reduce cardiopulmonary complications, restrictive fluid regimens seem superior to liberal fluid treatment 52. Both liberal and restrictive fluid therapies may induce hypoperfusion of the anastomosis by causing local edema or hypovolemia, which could be avoided by individualized, goal-directed fluid therapy, aiming at maximal stroke volumes. A recent randomized controlled trial (RCT) using esophageal Doppler monitoring for cardiac output measurement could however not prove a reduction in postoperative complications 53. The benefits of goaldirected fluid therapy in colorectal surgery are not indisputable and the effects on intestinal integrity are not known. Therefore, the effect of goal-directed fluid therapy on intestinal perfusion, damage and healing needs further exploration. Several animal and human studies have indicated that the use of NSAIDs is markedly correlated with anastomotic leakage following colorectal surgery 5456 . The ERAS guidelines state that sufficient evidence is lacking to stop using NSAIDs as a component of multimodal analgesia 39. The mechanisms by which NSAIDs exert their detrimental effects on colonic surgical wound healing are not known, which deserves further in-depth investigation. As classical wound healing, anastomotic wound healing consists of inflammation, proliferation, and remodeling of which the first two phases are principally of interest in the context of anastomotic leakage, which develops within several days after surgery 57 . Platelets first establish hemostasis, followed by influx of inflammatory cells, mainly neutrophils, monocytes and macrophages 58. Myofibroblasts, regulated by various growth factors, then create collagen-rich granulation tissue, a process

14

Introduction

in which angiogenesis plays a crucial role. Cyclooxygenase-2 (COX-2), the ratelimiting enzyme in the conversion of arachidonic acid into various prostaglandins, is involved in inflammation, proliferation and angiogenesis 59-61. Since NSAIDs block COX-2, these mechanisms can be involved in the detrimental effects of NSAIDs on gastrointestinal wound healing and should therefore be investigated.

1.3 Pediatric gastrointestinal surgery Necrotizing enterocolitis (NEC) is the most frequent indication for gastrointestinal surgery in infants, affecting predominantly premature neonates 62. NEC carries high morbidity and mortality (20-40%) 63. Early diagnosis, the detection of severe NEC requiring surgical treatment, and the timing to reintroduce enteral feeding after NEC episodes are unresolved issues. The initial clinical manifestations of NEC are non-specific and indistinguishable from other gastrointestinal disorders and non-abdominal sepsis 64. The diagnosis is further hampered by limited diagnostic accuracy of laboratory tests and currently used imaging modalities 65, 66 . Therefore, there is a high need for accurate tests to diagnose NEC and severe NEC necessitating surgery at an early stage. Furthermore, knowledge of the pathophysiological processes underlying NEC is required. The etiology of NEC remains poorly understood, although a strong correlation with formula feeding has been established 67. Growth factors abundantly present in human milk, including epidermal growth factor and insulin-like growth factor, which stimulate epithelial cell growth and cell differentiation, are supposed to play an important role in the regulation of the neonatal gastrointestinal development 68, 69. The effect of human milk on intestinal damage and maturation should be further investigated to extend knowledge on the mechanisms by which human milk is beneficial, which may ultimately lead to finding appropriate human milk substitutes. Furthermore, NEC incidence is inversely associated with gestational age 70. Intestinal barrier function is incompletely developed in the first week after birth indicated by higher intestinal permeability, especially in premature infants 71-75. Immaturity of the gut also plays a pivotal role in the development of gut-derived sepsis and feeding problems 76. The underlying mechanisms remain poorly understood, but increased intestinal permeability and inadequate intestinal immune responses in preterm neonates are associated with gastrointestinal disorders 74, 77, 78. New research should aim at defining intestinal morphological differences between premature and term infants, and differences

15

Chapter 1

in circulating levels of gut-specific markers to obtain better understanding of the immature gut. This may provide necessary knowledge for the development of protective strategies in this vulnerable patient population.

1.4 Aims to be studied As outlined in this chapter, various opportunities exist to improve outcomes in gastrointestinal surgery. The central hypothesis of this thesis was that in vulnerable patients with gastrointestinal disorders, inadequate wound healing, postoperative morbidity and mortality are likely to develop. Therefore, patients at the extremes of age, i.e. elderly patients and premature neonates with severe gastrointestinal morbidity were investigated. The aims of this thesis were to: • study factors and mechanisms underlying intestinal cell damage, inflammation and impaired wound healing in gastrointestinal surgery to find new strategies for the prevention of postoperative complications (Chapter 2-6, Pathophysiologic aspects of complications after adult gastrointestinal surgery) • investigate biomarkers to detect complications early after colorectal surgery (Chapter 7, Biomarkers of complications after adult gastrointestinal surgery) • characterize gut-specific biomarkers in the context of gut maturation to gain better understanding of the pathophysiology of NEC (Chapter 8-9, Pathophysiologic aspects of NEC) • assess the usefulness of gut-specific biomarkers in the detection, treatment and follow-up of pediatric surgery for NEC (Chapter 10-12, Biomarkers of NEC)

16

Introduction

References 1.

2. 3. 4.

5. 6.

7.

8. 9.

10.

11. 12.

13. 14.

15. 16.

Krarup PM, Nordholm-Carstensen A, Jorgensen LN, et al. Anastomotic Leak Increases Distant Recurrence and Long-Term Mortality After Curative Resection for Colonic Cancer: A Nationwide Cohort Study. Ann Surg 2013. Wick EC, Shore AD, Hirose K, et al. Readmission rates and cost following colorectal surgery. Dis Colon Rectum 2011;54:1475-9. Deitch EA. Multiple organ failure. Pathophysiology and potential future therapy. Ann Surg 1992;216:117-34. Holland J, Carey M, Hughes N, et al. Intraoperative splanchnic hypoperfusion, increased intestinal permeability, down-regulation of monocyte class II major histocompatibility complex expression, exaggerated acute phase response, and sepsis. Am J Surg 2005;190:393-400. Swank GM, Deitch EA. Role of the gut in multiple organ failure: bacterial translocation and permeability changes. World J Surg 1996;20:411-7. Garcia-Granero A, Frasson M, Flor-Lorente B, et al. Procalcitonin and C-reactive protein as early predictors of anastomotic leak in colorectal surgery: a prospective observational study. Dis Colon Rectum 2013;56:475-83. Soerjomataram I, Lortet-Tieulent J, Parkin DM, et al. Global burden of cancer in 2008: a systematic analysis of disability-adjusted life-years in 12 world regions. Lancet 2012;380:1840-50. National Cancer Intelligence Network: major surgical resections report 2011. Law WL, Choi HK, Lee YM, et al. The impact of postoperative complications on longterm outcomes following curative resection for colorectal cancer. Ann Surg Oncol 2007;14:2559-66. Walker KG, Bell SW, Rickard MJ, et al. Anastomotic leakage is predictive of diminished survival after potentially curative resection for colorectal cancer. Ann Surg 2004;240:255-9. Kehlet H. Fast-track colorectal surgery. Lancet 2008;371:791-3. Zhuang CL, Ye XZ, Zhang XD, et al. Enhanced recovery after surgery programs versus traditional care for colorectal surgery: a meta-analysis of randomized controlled trials. Dis Colon Rectum 2013;56:667-78. Jaarrapportage 2012: uitkomst van zorg registratie. Dutch Surgical Colorectal Audit, 2012. Tan WP, Talbott VA, Leong QQ, et al. American Society of Anesthesiologists class and Charlson’s comorbidity index as predictors of postoperative colorectal anastomotic leak: a single-institution experience. J Surg Res 2013;184:115-9. Makela JT, Kiviniemi H, Laitinen S. Risk factors for anastomotic leakage after left-sided colorectal resection with rectal anastomosis. Dis Colon Rectum 2003;46:653-60. Snijders HS, Wouters MW, van Leersum NJ, et al. Meta-analysis of the risk for anastomotic leakage, the postoperative mortality caused by leakage in relation to the overall postoperative mortality. Eur J Surg Oncol 2012;38:1013-9.

17

Chapter 1

17. Golub R, Golub RW, Cantu R, Jr., et al. A multivariate analysis of factors contributing to leakage of intestinal anastomoses. J Am Coll Surg 1997;184:364-72. 18. Mirnezami A, Mirnezami R, Chandrakumaran K, et al. Increased local recurrence and reduced survival from colorectal cancer following anastomotic leak: systematic review and meta-analysis. Ann Surg 2011;253:890-9. 19. DePinho RA. The age of cancer. Nature 2000;408:248-54. 20. Berrino F, Verdecchia A, Lutz JM, et al. Comparative cancer survival information in Europe. Eur J Cancer 2009;45:901-8. 21. McMillan DC, McArdle CS, Morrison DS. A clinical risk score to predict 3-, 5- and 10year survival in patients undergoing surgery for Dukes B colorectal cancer. Br J Cancer 2010;103:970-4. 22. Matthiessen P, Hallbook O, Rutegard J, et al. Population-based study of risk factors for postoperative death after anterior resection of the rectum. Br J Surg 2006;93:498503. 23. Schiesser M, Kirchhoff P, Muller MK, et al. The correlation of nutrition risk index, nutrition risk score, and bioimpedance analysis with postoperative complications in patients undergoing gastrointestinal surgery. Surgery 2009;145:519-26. 24. Persson C, Glimelius B. The relevance of weight loss for survival and quality of life in patients with advanced gastrointestinal cancer treated with palliative chemotherapy. Anticancer Res 2002;22:3661-8. 25. Tan KY, Kawamura YJ, Tokomitsu A, et al. Assessment for frailty is useful for predicting morbidity in elderly patients undergoing colorectal cancer resection whose comorbidities are already optimized. Am J Surg 2012;204:139-43. 26. Fried LP, Hadley EC, Walston JD, et al. From bedside to bench: research agenda for frailty. Sci Aging Knowledge Environ 2005;2005:pe24. 27. Makary MA, Segev DL, Pronovost PJ, et al. Frailty as a predictor of surgical outcomes in older patients. J Am Coll Surg 2010;210:901-8. 28. Saxton A, Velanovich V. Preoperative frailty and quality of life as predictors of postoperative complications. Ann Surg 2011;253:1223-9. 29. Schuurmans H, Steverink N, Lindenberg S, et al. Old or frail: what tells us more? J Gerontol A Biol Sci Med Sci 2004;59:M962-5. 30. Lieffers JR, Bathe OF, Fassbender K, et al. Sarcopenia is associated with postoperative infection and delayed recovery from colorectal cancer resection surgery. Br J Cancer 2012;107:931-6. 31. van Vledder MG, Levolger S, Ayez N, et al. Body composition and outcome in patients undergoing resection of colorectal liver metastases. Br J Surg 2012;99:550-7. 32. Kruizenga HM, de Jonge P, Seidell JC, et al. Are malnourished patients complex patients? Health status and care complexity of malnourished patients detected by the Short Nutritional Assessment Questionnaire (SNAQ). Eur J Intern Med 2006;17:18994. 33. Lomivorotov VV, Efremov SM, Boboshko VA, et al. Evaluation of nutritional screening tools for patients scheduled for cardiac surgery. Nutrition 2013;29:436-42.

18

Introduction

34. Burden ST, Hill J, Shaffer JL, et al. Nutritional status of preoperative colorectal cancer patients. J Hum Nutr Diet 2010;23:402-7. 35. 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;22:235-9. 36. Deslauriers J, Ginsberg RJ, Dubois P, et al. Current operative morbidity associated with elective surgical resection for lung cancer. Can J Surg 1989;32:335-9. 37. Lidder P, Thomas S, Fleming S, et al. A randomized placebo controlled trial of preoperative carbohydrate drinks and early postoperative nutritional supplement drinks in colorectal surgery. Colorectal Dis 2013;15:737-45. 38. Oguz M, Kerem M, Bedirli A, et al. L-alanin-L-glutamine supplementation improves the outcome after colorectal surgery for cancer. Colorectal Dis 2007;9:515-20. 39. Gustafsson UO, Scott MJ, Schwenk W, et al. Guidelines for perioperative care in elective colonic surgery: Enhanced Recovery After Surgery (ERAS(R)) Society recommendations. Clin Nutr 2012;31:783-800. 40. Bozzetti F, Gavazzi C, Miceli R, et al. Perioperative total parenteral nutrition in malnourished, gastrointestinal cancer patients: a randomized, clinical trial. JPEN J Parenter Enteral Nutr 2000;24:7-14. 41. Hyman N, Manchester TL, Osler T, et al. Anastomotic leaks after intestinal anastomosis: it’s later than you think. Ann Surg 2007;245:254-8. 42. Kornmann VN, Treskes N, Hoonhout LH, et al. Systematic review on the value of CT scanning in the diagnosis of anastomotic leakage after colorectal surgery. Int J Colorectal Dis 2012. 43. Alves A, Panis Y, Trancart D, et al. Factors associated with clinically significant anastomotic leakage after large bowel resection: multivariate analysis of 707 patients. World J Surg 2002;26:499-502. 44. Macarthur DC, Nixon SJ, Aitken RJ. Avoidable deaths still occur after large bowel surgery. Scottish Audit of Surgical Mortality, Royal College of Surgeons of Edinburgh. Br J Surg 1998;85:80-3. 45. den Dulk M, Noter SL, Hendriks ER, et al. Improved diagnosis and treatment of anastomotic leakage after colorectal surgery. Eur J Surg Oncol 2009;35:420-6. 46. Hirst N, Tiernan J, Millner P, et al. Systematic review of methods to predict and detect anastomotic leakage in colorectal surgery. Colorectal Dis 2013. 47. Platt JJ, Ramanathan ML, Crosbie RA, et al. C-reactive Protein as a Predictor of Postoperative Infective Complications after Curative Resection in Patients with Colorectal Cancer. Ann Surg Oncol 2012;19:4168-77. 48. Riddez L, Hahn RG, Brismar B, et al. Central and regional hemodynamics during acute hypovolemia and volume substitution in volunteers. Crit Care Med 1997;25:635-40. 49. Varadhan KK, Neal KR, Dejong CH, et al. The enhanced recovery after surgery (ERAS) pathway for patients undergoing major elective open colorectal surgery: a metaanalysis of randomized controlled trials. Clin Nutr 2010;29:434-40.

19

Chapter 1

50. Spanjersberg WR, Reurings J, Keus F, et al. Fast track surgery versus conventional recovery strategies for colorectal surgery. Cochrane Database Syst Rev 2011:CD007635. 51. Nygren J, Thacker J, Carli F, et al. Guidelines for perioperative care in elective rectal/ pelvic surgery: Enhanced Recovery After Surgery (ERAS(R)) Society recommendations. Clin Nutr 2012;31:801-16. 52. Rahbari NN, Zimmermann JB, Schmidt T, et al. Meta-analysis of standard, restrictive and supplemental fluid administration in colorectal surgery. Br J Surg 2009;96:33141. 53. Brandstrup B, Svendsen PE, Rasmussen M, et al. Which goal for fluid therapy during colorectal surgery is followed by the best outcome: near-maximal stroke volume or zero fluid balance? British journal of anaesthesia 2012;109:191-9. 54. Gorissen KJ, Benning D, Berghmans T, et al. Risk of anastomotic leakage with nonsteroidal anti-inflammatory drugs in colorectal surgery. Br J Surg 2012;99:721-7. 55. Holte K, Andersen J, Jakobsen DH, et al. Cyclo-oxygenase 2 inhibitors and the risk of anastomotic leakage after fast-track colonic surgery. Br J Surg 2009;96:650-4. 56. Klein M, Gogenur I, Rosenberg J. Postoperative use of non-steroidal anti-inflammatory drugs in patients with anastomotic leakage requiring reoperation after colorectal resection: cohort study based on prospective data. BMJ 2012;345:e6166. 57. Witte MB, Barbul A. General principles of wound healing. Surg Clin North Am 1997;77:509-28. 58. Thompson SK, Chang EY, Jobe BA. Clinical review: Healing in gastrointestinal anastomoses, part I. Microsurgery 2006;26:131-6. 59. Buchanan FG, Wang D, Bargiacchi F, et al. Prostaglandin E2 regulates cell migration via the intracellular activation of the epidermal growth factor receptor. J Biol Chem 2003;278:35451-7. 60. Fukata M, Chen A, Klepper A, et al. Cox-2 is regulated by Toll-like receptor-4 (TLR4) signaling: Role in proliferation and apoptosis in the intestine. Gastroenterology 2006;131:862-77. 61. Binion DG, Otterson MF, Rafiee P. Curcumin inhibits VEGF-mediated angiogenesis in human intestinal microvascular endothelial cells through COX-2 and MAPK inhibition. Gut 2008;57:1509-17. 62. Lin PW, Nasr TR, Stoll BJ. Necrotizing enterocolitis: recent scientific advances in pathophysiology and prevention. Semin Perinatol 2008;32:70-82. 63. Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011;364:255-64. 64. Fanaroff AA, Korones SB, Wright LL, et al. Incidence, presenting features, risk factors and significance of late onset septicemia in very low birth weight infants. The National Institute of Child Health and Human Development Neonatal Research Network. Pediatr Infect Dis J 1998;17:593-8. 65. Hallstrom M, Koivisto AM, Janas M, et al. Laboratory parameters predictive of developing necrotizing enterocolitis in infants born before 33 weeks of gestation. J Pediatr Surg 2006;41:792-8.

20

Introduction

66. Tam AL, Camberos A, Applebaum H. Surgical decision making in necrotizing enterocolitis and focal intestinal perforation: predictive value of radiologic findings. J Pediatr Surg 2002;37:1688-91. 67. Lucas A, Cole TJ. Breast milk and neonatal necrotising enterocolitis. Lancet 1990;336:1519-23. 68. Carver JD, Barness LA. Trophic factors for the gastrointestinal tract. Clin Perinatol 1996;23:265-85. 69. Xu RJ. Development of the newborn GI tract and its relation to colostrum/milk intake: a review. Reprod Fertil Dev 1996;8:35-48. 70. Ostlie DJ, Spilde TL, St Peter SD, et al. Necrotizing enterocolitis in full-term infants. J Pediatr Surg 2003;38:1039-42. 71. Israel EJ. Neonatal necrotizing enterocolitis, a disease of the immature intestinal mucosal barrier. Acta Paediatr Suppl 1994;396:27-32. 72. Weaver LT, Laker MF, Nelson R. Intestinal permeability in the newborn. Arch Dis Child 1984;59:236-41. 73. Beach RC, Menzies IS, Clayden GS, et al. Gastrointestinal permeability changes in the preterm neonate. Arch Dis Child 1982;57:141-5. 74. Rouwet EV, Heineman E, Buurman WA, et al. Intestinal permeability and carriermediated monosaccharide absorption in preterm neonates during the early postnatal period. Pediatr Res 2002;51:64-70. 75. van Elburg RM, Fetter WP, Bunkers CM, et al. Intestinal permeability in relation to birth weight and gestational and postnatal age. Arch Dis Child Fetal Neonatal Ed 2003;88:F52-5. 76. Neu J. Gastrointestinal development and meeting the nutritional needs of premature infants. Am J Clin Nutr 2007;85:629S-634S. 77. Neu J, Chen M, Beierle E. Intestinal innate immunity: how does it relate to the pathogenesis of necrotizing enterocolitis. Semin Pediatr Surg 2005;14:137-44. 78. Lin PW, Nasr TR, Stoll BJ. Necrotizing enterocolitis: recent scientific advances in pathophysiology and prevention. Semin Perinatol 2008;32:70-82.

21

Part 1 Pathophysiologic aspects of complications after adult gastrointestinal surgery The first aim was to study factors and mechanisms underlying intestinal cell damage, inflammation and impaired wound healing in gastrointestinal surgery to find new strategies for the prevention of postoperative complications. In the first study, it was investigated whether optimization of perioperative fluid administration based on cardiac stroke volumes led to increased splanchnic perfusion and diminished intestinal cell damage in patients undergoing colorectal surgery (chapter 2). In an animal model of colorectal anastomotic leakage, the fundamental role of cyclooxygenase-2 (COX-2) in intestinal wound healing was shown (chapter 3). Next, several aspects of functional compromise and their relationship to postoperative morbidity, mortality and inflammation were investigated in patients undergoing major gastrointestinal surgery (chapters 4-6). Hemodynamic optimization may be of high clinical importance, as a sufficient supply of oxygen and nutritions is needed to ensure adequate wound healing capacity. Furthermore, an important group of analgesics that are frequently used in major surgery inhibit COX-2, and possibly deteriorate postoperative wound healing. Finally, elderly patients are more likely to develop complications, which may be explained by loss of physiologic reserves. Preoperative selection of patients who are at high risk for developing postoperative complications is crucial for therapeutic decisions and for the selection of patients who require preoperative optimization.

23

Chapter 2. Doppler guided goal-directed fluid therapy increases intestinal perfusion in colorectal surgery

Kostan W. Reisinger, Henriette M. Willigers, Jochen Jansen, Maarten F. Von Meyenfeldt, Geerard L. Beets, Martijn Poeze. Submitted

25

Chapter 2

Abstract Background Individualized, goal-directed fluid therapy (GDFT) based on Doppler measurements of stroke volume has been proposed as a treatment strategy in terms of reducing complications, mortality and length of hospital stay in major bowel surgery. The effect of Doppler guided GDFT on intestinal damage and perfusion is studied. Methods Patients undergoing elective colorectal resection for malignancy were randomized to standard intra- and postoperative fluid therapy with or without additional Doppler guided GDFT. The primary outcome was intestinal epithelial cell damage measured by plasma levels of intestinal fatty acid binding protein (I-FABP). Global intestinal perfusion was measured by gastric tonometry expressed as regional (gastric) minus arterial CO2-gap (Pr-aCO2-gap). Results I-FABP levels were not different between the intervention group and the control group. Mean areas under the curve (AUCs) of intraoperative Pr-aCO2-gaps were significantly lower in the intervention group compared to the control group (p=0.01), indicating better global gastrointestinal perfusion in the intervention group. Moreover, the mean intraoperative Pr-aCO2-gap peak in the intervention group was 0.5 (1.0) kPa, which was significantly lower than the mean peak in the control group; 1.4 (1.4) kPa, p=0.03. Conclusion Doppler guided GDFT during and after elective colorectal surgery for malignancy increases global gastrointestinal perfusion.

26

Goal-directed fluid therapy

Introduction Perioperative fluid management in colorectal surgery is subject to debate. Although restrictive fluid regimens seem superior to liberal fluid treatment 1, euvolemia may be the golden mean. However, euvolemia has not been defined thus far, at least not in terms of generalized amounts of fluid to be administered. Individualized, goal-directed fluid therapy (GDFT) can be considered as closest to euvolemia. GDFT has been proposed as a treatment strategy in terms of reducing complications, mortality and length of hospital stay in major bowel surgery 1-6. Studies indicate that GDFT is associated with shortened length of hospital stay of 2-3 days compared to controls, and a reduction of major complications requiring intensive care admission, and gastrointestinal complications 3, 4. Typically, goaldirected fluid protocols are designed to optimize stroke volumes intra-operatively guided by oesophageal Doppler monitoring and titration of colloid fluid boluses 7, 8. The effect of GDFT on intestinal perfusion and subsequently on intestinal damage and wound healing is unknown. Furthermore, the effect of GDFT in the early postoperative period has not been investigated thus far. It is suggested that Doppler guided fluid optimization increases bowel perfusion 9, while liberal and restrictive treatment strategies may induce hypoperfusion by local oedema and hypovolemia, respectively 10. Fluid management in the first hours following surgery may be as important as intra-operative fluid management in improving tissue perfusion and oxygenation 11. In this period, hypovolemia may critically compromise perfusion 12. In major bowel surgery, this may lead to the development of gutassociated complications, i.e. anastomotic leakage, intra-abdominal abscess and sepsis. GDFT may therefore also be related to a decreased risk on gut-associated complications. However, two recent randomized controlled trials could not prove a reduction in postoperative complications in general when GDFT was compared to restrictive fluid management 13, 14. The aim of the current study was to investigate as a clinical proof of concept whether oesophageal Doppler guided GDFT approximates euvolemia by attenuating intestinal damage and improving gastrointestinal perfusion during colorectal surgery and during the first hours after colorectal surgery, compared to standard fluid therapy. It was hypothesized that GDFT decreased intestinal injury and improved gastrointestinal perfusion.

27

Chapter 2

Methods Patients Fifty-eight patients undergoing colorectal resection for malignancy were enrolled in this single-centre, parallel randomized clinical trial (ClinicalTrials.gov identification number: NCT01175317) between July 2010 and August 2013. Patients were eligible for inclusion if they met the following criteria: elective colorectal cancer surgery with primary anastomosis and a minimum age of 18 years. Written informed consent was obtained from all enrolled patients. Exclusion criteria were: non-malignant causes of intestinal damage (e.g. inflammatory bowel diseases, occlusive disease); use of steroids; history of oesophageal varices and other oesophageal disease; and aortic valve disease. History of oesophageal varices is a contraindication for the use of oesophageal Doppler monitoring and aortic valve disease results in unreliable Doppler measurements. Before randomization, patients were stratified according to the type of surgery, i.e. laparoscopy or open surgery. For allocation of the participants, two computer-generated lists (for laparoscopy and open surgery, respectively) of random numbers were used following a simple randomization to one of two treatment groups. An investigator who did not take part in patient enrolment nor data acquisition was in charge of these lists. The investigator responsible for patient enrolment, Doppler measurements and fluid optimization protocol telephoned the investigator in charge of the lists after obtaining informed consent. The only other person aware of each patients’ allocation was the anaesthetist responsible for the perioperative fluid management. The study was approved by the medical ethical committee of Maastricht University Medical Centre, with number 09-2-089 and conducted according to the revised version of the Declaration of Helsinki (October 2008, Seoul). The study methods were not changed after trial registration. Anaesthetic procedure Anaesthesia was induced using propofol, sufentanil, and rocuronium and maintained using sevoflurane. In the majority of patients, an epidural catheter was inserted for additional analgesia using bupivacaine. The epidural catheter was placed at Th8–10, tested with 3–5 mL of 0.25% bupivacaine with adrenaline, and continuous infusion of bupivacaine 0.25% was given for sufficient block. After induction of anaesthesia, a medically qualified investigator (K.R.) inserted a oesophageal Doppler probe (CardioQ, Deltex Medical, Chichester, UK) transnasally.

28

Goal-directed fluid therapy

In both the intervention and control group, oesophageal Doppler measurements were performed every 15 minutes during surgery. The anaesthesiologist was blinded for Doppler measurements at all times. The optimal Doppler signal was obtained according to manufacturers’ instructions and stroke volume measurements were averaged over 5 beats. All patients received a radial artery line. Prophylactic antibiotics were given to all patients using metronidazole and cefazoline as guided by protocol. Fluid treatment and study intervention All patients were allowed to drink clear fluids until 2 h before surgery. Immediately after induction of anaesthesia, stroke volume measurements were performed in all patients. In all patients, Voluven was used to replace blood loss volume in a 1:1 ratio. Voluven and Ringer lactate were infused to maintain the mean arterial pressure (MAP) above 65 mmHg. Erythrocyte concentrates were given to keep haemoglobin levels above 5-6 mmol/L depending on age and presence of cardiac disease. If the blood loss was large, plasma and thrombocytes were added. Ephedrine or phenylephrine were given if hypotension persisted despite fluid infusion. In case the inotropic support was needed over a longer time period, noradrenaline was given as continuous infusion. In the intervention group, standard fluid therapy was as described above. Furthermore, a fluid optimization protocol adapted from Wakeling and colleagues was applied 4. Immediately after induction of surgery, a 250 mL bolus of colloid fluid (Voluven) was administered. If the increase in stroke volume (SV) was 10% or more, patients were considered hypovolemic and a new 250 mL bolus was given. This procedure was repeated until SV increase was less than 10%. The maximal SV was maintained during surgery and corrected if necessary with 250 mL boluses Voluven. In the case of hypotension despite Doppler-guided volume therapy, vasoactive drugs were given as described above. Postoperatively, patients were admitted to a standard post-anaesthetic care unit for at least six hours. The Doppler probe remained in situ for a maximum of six hours or until the patient experienced significant discomfort. In all patients, basic postoperative fluid management consisted of Ringer’s solution 2-4 ml/kg/h. Furthermore, volemic status was assessed every hour by the passive leg raising test (PLR), which was performed with a standardized angle of 135º between the trunk and lower limbs as described by Monnet et al. 15. The optimal Doppler signal was obtained and stroke volume measurements were averaged over 5 beats. The

29

Chapter 2

same procedure was repeated after PLR. Doppler recordings were taken when the stroke volume reached its highest value (which was approximately 30 seconds after PLR). The maximum effect of PLR on stroke volume was seen within 1 minute in all patients. In the control group, PLR was performed but no additional fluid interventions were done. In the intervention group, if the increase in SV during PLR was 10% or more, patients were considered hypovolemic and a 250 mL bolus Voluven was given. In the control group, fluid boluses were administered based on standard hemodynamic and clinical parameters. After discharge to a specialized colorectal surgery ward (at six hours postoperatively or more), patients were treated according to a multimodal fasttrack programme with early start of feeding and mobilization. Fluid therapy was equal in both groups: as soon as swallowing was considered safe, intake of fluids and nutrition was started, and a minimum fluid intake of 2 litres per day was aimed for. If this could not be reached by oral intake, i.v. fluid suppletion was given. Pain management was achieved by patient controlled analgesia via the epidural catheter or intravenously, for a maximum of three days postoperatively, and if needed additional paracetamol or morphine was given. Hemodynamic parameters The following hemodynamic parameters were monitored with 15-minute intervals during surgery and hourly during the first six hours after surgery: heart rate; stroke volume (Doppler); MAP (arterial line); and urinary output. Blood-soaked gauzes were weighted as they were passed off the surgical field and the blood content of the blood collection system was measured to assess total blood loss at the end of surgery. Blood sampling and processing Arterial blood samples were obtained from the radial artery line at predefined time points: baseline, every 30 minutes during surgery, and every hour until six hours postoperatively. Venous blood samples were taken daily until three days after surgery. Blood samples were collected in pre-chilled EDTA containing vacuum tubes (BD vacutainer, Becton Dickinson Diagnostics, Aalst, Belgium) and immediately centrifuged at 4 ˚C (2000 g, 15 minutes). Plasma samples were stored at -80 ˚C until batch analysis.

30

Goal-directed fluid therapy

Measurement of intestinal damage Intestinal fatty acid binding protein (I-FABP) plasma levels were determined using an in-house ELISA that selectively detects human I-FABP (lower detection limit: 25 pg/ml). Gastric tonometry A gastric tonometry catheter (14F, Medi-Line, Angleur, Belgium) was introduced transnasally for measurement of intramucosal carbon dioxide pressure (PrCO2 in kPa) throughout the surgical procedure and during the six hours following surgery, using the gas-automated capnograph (Tonocap TC-200, Datex-Ohmeda, Helsinki, Finland). Gastric tonometry measurements (PrCO2, and mucosal-arterial pCO2 gap (Pr-aCO2-gap)) were done at 15-minute intervals during surgery and hourly during the first six hours following surgery. The first measurement during surgery was done at 15 minutes following the start of surgery, due to the calibration time of the device. Statistical analysis Statistical analysis was performed using Prism 5.0 for Windows (Graphpad software, Inc, San Diego, CA) and SPSS 20.0 for Windows (SPSS Inc, Chicago, IL). Normality was tested using Kolmogorov-Smirnov. The primary outcome was intestinal epithelial damage measured by plasma I-FABP levels at 1 hour postoperatively. Secondary outcomes were Pr-aCO2-gap, and the hemodynamic parameters MAP, indexed stroke volume (SVI, stroke volume corrected for body surface area) and urinary production. For all repeated measures, areas under the curve (AUC) were calculated for each patient separately, and missing data were handled using multiple imputations in SPSS. Average AUCs of the outcomes were compared using Students t test. Except for length of hospital stay (median and range), all continuous variables are presented as mean and standard deviation (SD). Dichotomous variables were compared using Chi-square test. Sample size was calculated as follows. In a previous study, patients undergoing major non-abdominal surgery had mean I-FABP levels of 443 (SD, 309) pg/mL at the end of surgery 16. In the current study, a reduction to 200 (SD, 150) pg/mL by hemodynamic optimization was estimated a priori, necessitating a sample size

31

Chapter 2

of 27 per group, with α=0.05 and 1-β=0.95. As a 5% drop-out rate was expected because of inability to achieve adequate Doppler measurements, the needed sample size was estimated at 29 for each group.

Results Patients Fifty-eight patients were randomized, of whom 27 were allocated to the intervention group (Figure 1). The study protocol was completed in all patients and none were excluded from analysis. Baseline characteristics are summarized

FIGURE 1. CONSORT diagram of the study.

32

Goal-directed fluid therapy

TABLE 1. Patient characteristics

Sex

Intervention group

Control group

Patients

Patients

Male Female

21 (77.8%) 6 (22.2%)

> 70

10 (37.0%)

> 25 I II III Colon

12 (44.4%) 8 (29.6%) 14 (51.9%) 5 (18.5%) 11 (40.7%)

17 (54.8%) 4 (12.9%) 24 (77.4%) 3 (9.7%) 13 (41.9%)

Rectum/ Sigmoid

16 (59.3%)

18 (58.1%)

Age (years)

Tumour location

Mean (SD)

20 (64.5%) 11 (35.5%) 68.6 (10.8)

BMI (kg/m²) ASA

Mean (SD)

67.6 (10.0) 12 (38.7%)

25.9 (3.1)

25.3 (3.0)

Smokers Medical history

6 (22.2%) 5 (16.1%) Myocardial 1 (3.7%) 1 (3.2%) ischemia Stroke 4 (14.8%) 3 (9.7%) NIDDM 3 (11.1%) 2 (6.5%) COPD 1 (3.7%) 2 (6.5%) BMI: body mass index; ASA: American Society of Anaesthesiologists; NIDDM: non-insulindependent diabetes mellitus; COPD: chronic obstructive pulmonary disease TABLE 2. Surgery characteristics

Surgical approach Open Laparoscopy Type of surgery Right colectomy Left colectomy Sigmoid resection Rectal resection Subtotal colectomy Epidural Ostomy Operative time (minutes)

Intervention group

Control group

Patients

Patients

Mean (SD)

20 (74.1%) 7 (25.9%) 10 (37.0%) 1 (3.7%) 5 (18.5%)

21 (67.7%) 10 (32.3%) 12 (38.7%) 0 (0%) 5 (16.1%)

11 (40.7%) 0 (0%)

13 (41.9%) 1 (3.2%)

21 (77.8%) 12 (44.4%)

22 (71.0%) 14 (45.2%) 256 (101)

Mean (SD)

205 (77)

33

Chapter 2

in Table 1. No differences were observed between the intervention and control groups. Stratification for open and laparoscopic surgery was verified, and other operative characteristics are outlined in Table 2. Operative time was higher in the intervention group. More colloid fluid was given in the intervention group (Table 3), while total administered fluids did not differ between the groups. The mean time the Doppler probe was tolerated was 4 hours postoperatively. TABLE 3. Fluids during surgery

Total amount (mL)

Intervention group

Control group

Patients

Patients

Crystalloids Colloids

Total fluid (mL/kg/h) Blood loss (mL) > 750 mL 6 (22.2%) Blood transfusion 7 (25.9%) Total fluid (mL/kg/h) 1 Vasopressor use 12 (44.4%) 1 Including blood transfusions

Mean (SD) 3,000 (1,093) 1,526 (823) 14.6 (4.7) 957 (1,880)

Mean (SD) 3,026 (1,307) 952 (687) 16.2 (5.9) 461 (1,026)

4 (12.9%) 4 (12.9%) 15.8 (6.4)

16.7 (6.5) 15 (48.4%)

Hemodynamic changes A significant increase in SVI from baseline to start of surgery was accomplished by colloid administration in the group receiving hemodynamic optimization (46.5 (12.0) mL/m² to 59.8 (15.7) mL/m², p750 mL, n=10); 6.7 (8.3) kPa x h compared to 0.6 (3.3) kPa in patients without major blood loss (p=0.005, data not shown). In linear regression analysis, major blood loss was a significant independent predictor for increasing Pr-aCO2gap postoperatively (β=0.55, p=0.001), and group (intervention) showed a trend towards significance (β=-0.27, p=0.07). When only patients without major blood loss were analysed, mean postoperative AUC of the Pr-aCO2-gap was lower in the intervention group (-0.7 (3.9) kPa x h compared to 2.0 (2.3) kPa x h in the control group, p=0.04). Furthermore, the mean postoperative Pr-aCO2-gap peak in the intervention group was 0.6 (0.7) kPa in this subgroup, significantly lower than the peak in the control group; 1.4 (1.2) kPa, p=0.04. Clinical outcome Postoperative mortality, complications and length of hospital stay are summarized in Table 4. No significant differences in mortality or complications were found between the control and intervention groups. Length of hospital stay was significantly increased in the intervention group compared with the control group (median 11 (range, 4 – 50), and 8 (5 – 26), respectively, p=0.03).

37

Chapter 2

TABLE 4. Clinical outcome Intervention group

Control group

Patients

Patients

Median (range)

Median (range)

Mortality 0 (0%) 1 (3.2%) Anastomotic leakage 3 (11.1%) 2 (6.5%) Intra-abdominal abscess 2 (7.4%) 1 (3.2%) SSI 2 (7.4%) 3 (9.7%) Fascial dehiscence 0 (0%) 2 (6.5%) Pneumonia 3 (11.1%) 1 (3.2%) Urinary tract infection 2 (7.4%) 5 (16.1%) POI 3 (11.1%) 2 (6.5%) Gastroparesis 0 (0%) 3 (9.7%) Cardiac decompensation 0 (0%) 1 (3.2%) Unplanned ICU admission 1 (3.7%) 4 (12.9%) Readmission within 30 days 1 (3.7%) 3 (9.7%) Length of hospital stay (days) 11 (4 – 50) 8 (5 – 26) SSI: surgical site infection; POI: postoperative ileus; ICU: intensive care unit

Discussion This randomized controlled trial showed that Doppler guided GDFT during and after elective colorectal surgery for malignancy increases global gastrointestinal perfusion. However, no significant differences in plasma I-FABP levels between the intervention and control group were observed. In addition, both groups showed no significant intestinal damage. A strong positive effect of Doppler guided GDFT on gastrointestinal perfusion was seen during surgery. GDFT showed a marginally significant effect on postoperative gastrointestinal perfusion, and only when corrected for major intraoperative blood loss (>750 mL). It was hypothesized that plasma I-FABP levels would peak compared to baseline at about one hour after the end of surgery as was observed before in non-abdominal surgery with 17, 18 and without 16 aortic cross-clamping. The latter study investigating scoliosis repair in children showed that low MAPs are associated with increased I-FABP levels. In this study, average MAPs during surgery were 64 mmHg compared to 75 mmHg in the current study. It may therefore be speculated that patients undergoing elective colorectal surgery do not exhibit severe enough hypotension to develop intestinal damage. Nonetheless, increased gastrointestinal perfusion due to GDFT indicates

38

Goal-directed fluid therapy

a rather euvolemic status in these patients, since the gut is one of the organs that are primarily affected by the redistribution of blood to the vital organs in (beginning) hypovolemia 19. The discrepancy between the effect of GDFT on intraoperative in contrast to postoperative perfusion may be explained by the methodological differences of fluid optimization during and after surgery. This was underlined by higher mean SVI in the intervention group compared to the control group intraoperatively, but not postoperatively. Just before and during surgery, the stroke volume was optimized by fluid challenges, while after surgery, fluid responsiveness was assessed by PLR and fluid challenges were only given when the PLR was positive. Although PLR represents good sensitivity (77%) to detect fluid responsiveness, some patients that should receive a fluid bolus (23%) are inevitably missed 20. Moreover, other factors may be more important predictors of postoperative gastrointestinal perfusion, as was indicated by the strong association of major intraoperative blood loss and decreased postoperative gastrointestinal perfusion. Interestingly, the amount of fluid given did not differ between the intervention and the control group, indicating that rather timing of fluid administration was the determinant of improving global and regional hemodynamics, thereby shedding new light on the concept of euvolemia. This observation is in line with previous work describing more stable hemodynamic parameters when stroke volumebased optimization was applied even though total amount of fluid given was comparable with the control group 21. Another interesting finding was the need for fluid expansion in 89% of patients to establish maximal stroke volumes, which is in line with a previous study showing a functional volume deficit in 70% of patients undergoing different types of surgery 22. This underlines the possible benefits of patient-tailored GDFT. However, it remains unclear whether such deficits represent actual susceptibility to complications or rather increased physiological reserves. Several other randomized clinical trials have been performed on Doppler guided GDFT in major bowel surgery 2-6, 13, 14. Although some trials showed a significantly shortened length of hospital stay, decreased morbidity and increased gut function 3, 4, the largest 6, 13 and most recent 14 trials showed no advantage of Doppler guided GDFT over standard or restrictive fluid therapy. Challand and coworkers even showed that GDFT increases length of hospital stay in aerobically fit patients 6. This supports the hypothesis that sub-maximal stroke volumes reflect physiological reserves instead of a pathological deficit, and stroke volume optimization, although improving gut perfusion, can actually lead to fluid overload

39

Chapter 2

in these patients. The present study adds the application of Doppler guided GDFT in the early postoperative phase. However, no or only marginally significant effects of the intervention were seen postoperatively in terms of gastrointestinal perfusion or SVI. Therefore, GDFT has no additional effect compared with standard fluid treatment in the postoperative phase based on the current results. The present study was not designed to detect differences in clinical outcome. Although we show that stroke volume optimization improves gastrointestinal perfusion, it remains to be determined in which patients this approach leads to better clinical outcome. As noted in the Challand trial 6, patients that have low oxygen consumption levels may benefit from GDFT, however large numbers of such a selected population are needed and not easily obtainable. Moreover, caution should be taken when interpreting the gut tonometry data, for the measurements were done in the stomach as a reflection of overall gastrointestinal perfusion. In colorectal surgery, an important target of fluid therapy is to establish adequate perfusion of the gut and the anastomosis in particular, to diminish the risk of postoperative complications such as anastomotic leakage and gut-derived sepsis. It is not known whether this generalized read-out accurately correlates with local (hypo)perfusion. Other techniques, such as in vivo microscopy 23 could be accurate tools for detecting low local perfusion. In conclusion, Doppler guided GDFT increased global gastrointestinal perfusion in this study in patients undergoing elective colorectal surgery, indicating a rather euvolemic state in these patients. This provides new evidence for the implementation of GDFT in clinical practice. However, the definition of clinically important euvolemia deserves further exploration, since the clinical benefit of GDFT over other strategies was not unequivocally proven in previous trials.

40

Goal-directed fluid therapy

References 1.

2.

3.

4.

5.

6.

7.

8.

9. 10. 11.

12. 13.

14.

Rahbari NN, Zimmermann JB, Schmidt T, et al. Meta-analysis of standard, restrictive and supplemental fluid administration in colorectal surgery. Br J Surg 2009;96:33141. Conway DH, Mayall R, Abdul-Latif MS, et al. Randomised controlled trial investigating the influence of intravenous fluid titration using oesophageal Doppler monitoring during bowel surgery. Anaesthesia 2002;57:845-9. Noblett SE, Snowden CP, Shenton BK, et al. Randomized clinical trial assessing the effect of Doppler-optimized fluid management on outcome after elective colorectal resection. Br J Surg 2006;93:1069-76. Wakeling HG, McFall MR, Jenkins CS, et al. Intraoperative oesophageal Doppler guided fluid management shortens postoperative hospital stay after major bowel surgery. Br J Anaesth 2005;95:634-42. Senagore AJ, Emery T, Luchtefeld M, et al. Fluid management for laparoscopic colectomy: a prospective, randomized assessment of goal-directed administration of balanced salt solution or hetastarch coupled with an enhanced recovery program. Dis Colon Rectum 2009;52:1935-40. Challand C, Struthers R, Sneyd JR, et al. Randomized controlled trial of intraoperative goal-directed fluid therapy in aerobically fit and unfit patients having major colorectal surgery. Br J Anaesth 2012;108:53-62. Klotz KF, Klingsiek S, Singer M, et al. Continuous measurement of cardiac output during aortic cross-clamping by the oesophageal Doppler monitor ODM 1. Br J Anaesth 1995;74:655-60. Donati A, Munch C, Marini B, et al. Transesophageal Doppler ultrasonography evaluation of hemodynamic changes during videolaparoscopic cholecystectomy. Minerva Anestesiol 2002;68:549-54. Roy N, Maw A, Stuart-Smith K. Fluid optimization guided by oesophageal Doppler significantly improves bowel perfusion. Br J Anaesth 2011;107:1012-3. Holte K, Sharrock NE, Kehlet H. Pathophysiology and clinical implications of perioperative fluid excess. Br J Anaesth 2002;89:622-32. Poeze M, Ramsay G, Greve JW, et al. Prediction of postoperative cardiac surgical morbidity and organ failure within 4 hours of intensive care unit admission using esophageal Doppler ultrasonography. Crit Care Med 1999;27:1288-94. Brandstrup B. Fluid therapy for the surgical patient. Best Pract Res Clin Anaesthesiol 2006;20:265-83. Brandstrup B, Svendsen PE, Rasmussen M, et al. Which goal for fluid therapy during colorectal surgery is followed by the best outcome: near-maximal stroke volume or zero fluid balance? Br J Anaesth 2012;109:191-9. Srinivasa S, Taylor MH, Singh PP, et al. Randomized clinical trial of goal-directed fluid therapy within an enhanced recovery protocol for elective colectomy. Br J Surg 2013;100:66-74.

41

Chapter 2

15. Monnet X, Rienzo M, Osman D, et al. Passive leg raising predicts fluid responsiveness in the critically ill. Crit Care Med 2006;34:1402-7. 16. Derikx JP, van Waardenburg DA, Thuijls G, et al. New Insight in Loss of Gut Barrier during Major Non-Abdominal Surgery. PLoS One 2008;3:e3954. 17. Vermeulen Windsant IC, Hellenthal FA, Derikx JP, et al. Circulating intestinal fatty acid-binding protein as an early marker of intestinal necrosis after aortic surgery: a prospective observational cohort study. Ann Surg 2012;255:796-803. 18. Hanssen SJ, Derikx JP, Vermeulen Windsant IC, et al. Visceral injury and systemic inflammation in patients undergoing extracorporeal circulation during aortic surgery. Ann Surg 2008;248:117-25. 19. Moore FA. The role of the gastrointestinal tract in postinjury multiple organ failure. Am J Surg 1999;178:449-53. 20. Lamia B, Ochagavia A, Monnet X, et al. Echocardiographic prediction of volume responsiveness in critically ill patients with spontaneously breathing activity. Intensive Care Med 2007;33:1125-32. 21. Rinehart J, Chung E, Canales C, et al. Intraoperative stroke volume optimization using stroke volume, arterial pressure, and heart rate: closed-loop (learning intravenous resuscitator) versus anesthesiologists. J Cardiothorac Vasc Anesth 2012;26:933-9. 22. Bundgaard-Nielsen M, Jorgensen CC, Secher NH, et al. Functional intravascular volume deficit in patients before surgery. Acta Anaesthesiol Scand 2010;54:464-9. 23. Goedhart PT, Khalilzada M, Bezemer R, et al. Sidestream Dark Field (SDF) imaging: a novel stroboscopic LED ring-based imaging modality for clinical assessment of the microcirculation. Opt Express 2007;15:15101-14.

42

Chapter 3. Cyclooxygenase-2 is essential for colorectal anastomotic healing

Kostan W. Reisinger, Dirk H.S.M. Schellekens, Joanna W.A.M. Bosmans, Bas Boonen, Maarten F. von Meyenfeldt, Prapto Sastrowijoto, Joep P.M. Derikx, Joep Grootjans, Martijn Poeze. Submitted

45

Chapter 3

Abstract Background Cyclooxygenase-2 (COX-2) is a key enzyme in gastrointestinal homeostasis, affecting angiogenesis, profileration, inflammation and apoptosis. In this study the effects of COX-2 in colonic surgical wound healing were investigated. We provide evidence that COX-2 is essential for neovascularization of the colonic anastomosis and thereby plays a crucial role in colonic anastomotic wound healing. Methods Mice of different COX-2 genotypes were subjected to a model of colonic anastomotic leakage, and received vehicle, diclofenac or prostaglandin E2 (PGE2). Endpoints for anastomotic healing were incidence of anastomotic leakage and mortality. Angiogenesis was assessed in anastomotic tissue. Results All mice receiving diclofenac developed anastomotic leakage. Of mice completely lacking COX-2, 92% developed anastomotic leakage, compared to 25% of wildtypes (p=0.003). Supplementation with PGE2 decreased the anastomotic leakage rate in COX-2-/- mice from 92% to only 46% (p=0.02). Quantification of the amount of blood vessels showed that only 2 vessels/mm2 were stained in anastomotic tissue of COX-2-/- mice compared to 6 vessels/mm2 in wildtype mice (p=0.03). This effect could partly be reversed by administration of PGE2 to COX-2-/- mice. Conclusion COX-2 functioning is essential for intestinal wound healing after colonic surgery, which is mediated by PGE2 production. In addition, angiogenesis is significantly impaired in the absence of COX-2 and PGE2. Future research should aim at improving COX-2 and PGE2 functioning in situations when adequate intestinal wound healing is of high importance.

46

Cyclooxygenase-2

Introduction The enzyme cyclooxygenase-2 (COX-2, also known as prostaglandin-endoperoxide syntase 2 (Ptgs2)) plays an important role in preserving gut homeostasis. In the colon, COX-2 is expressed constitutively in mesenchymal stem cells, producing prostaglandins which is believed to have an immunomodulatory role 1. These mesenchymal stem cells may act as monitors of the colonic environment and are important in colonic wound healing 2. In addition, COX-2 expression can be induced in macrophages and myofibroblasts upon exposure to proinflammatory cytokines and bacterial products 1, 3, 4. Cyclooxygenase regulates the conversion of arachidonic acid into prostaglandins, of which PGE2 is reported to restore intestinal integrity in experimental models of intestinal inflammation and damage 1, 2, 4-6 . Furthermore, COX-2 and PGE2 are critically involved in vascular endothelial growth factor (VEGF)-induced angiogenesis 7, 8. Data from human studies underline the importance of COX-2 in colonic wound healing as the use of NSAIDs, especially those with strong COX-2-inhibiting properties, is markedly correlated with anastomotic leakage, i.e. inadequate wound healing following colorectal surgery 9-12. These are however retrospective studies and a recent meta-analysis could not prove an unambiguous detrimental effect of NSAIDs 13. These studies emphasize the importance of precautious NSAID use in patients with intestinal anastomoses, however omission of NSAIDs from postsurgical care in these patients is not yet standard clinical practice 14, 15 . Studies showing the mechanisms by which NSAIDs exert their effects on colonic surgical wound healing are essential to ban the use of NSAIDs for pain management after colonic surgery. In this study the effects of COX-2 and PGE2 in colonic surgical wound healing were investigated. We provide evidence that COX-2 is essential for neovascularization of the colonic anastomosis and thereby plays a crucial role in colonic anastomotic wound healing. In line, blocking COX-2 either by using COX2-/- mice or by administration of NSAIDs was not only associated with significantly higher anastomic leakage rates but also with increased mortality. These data further emphasize that COX-2 inhibitors should be avoided in patients undergoing colonic surgery.

47

Chapter 3

Materials and methods Mice All animal experiments were approved by the Maastricht University Animal Experiments Committee. COX-2+/- mice were obtained from the Jackson Laboratory (Bar Harbor, ME) to produce littermate wildtype (WT), COX-2+/- and COX-2-/- mice. For all experiments, COX-2-/- mice were used simultaneously with their wildtype and/or heterozygous littermates, resulting in various group sizes. Model A previously described murine model of colonic anastomotic leakage was used and adapted to obtain an anastomotic leakage rate of 25-33%, to obtain a model that resembles clinical practice 16. Briefly, 10- to 12-week-old mice were anesthetized using isoflurane and buprenorphine. After a 1 cm midline laparotomy, the cecum was exteriorized and the right colon was microscopically transected, without damaging blood supply. An end-to-end anastomosis was performed with 7 interrupted sutures (Prolene 8-0, Ethicon, Somerville, NJ). The colon was repositioned and the abdomen was closed in two layers of interrupted sutures (Vicryl 4-0, Ethicon, and Ethilon 4-0, Ethicon, respectively). The colon was kept moist with sterile NaCl during the procedure. The surgical procedure was performed by two medically trained researchers that had equal skill and experience with the surgical procedure. Study design WT, COX-2+/- and COX-2-/- mice underwent laparotomy with colonic anastomosis, and received vehicle (PBS), 16,16-dimethyl PGE2 (dmPGE2, Cayman Chemical, Ann Arbor, MI), a stable analogue of PGE2, or diclofenac sodium (Cayman Chemical) by intraperitoneal injection. Dosage of dmPGE2 was 100 μg/kg bodyweight, twice daily, and dosage of diclofenac was 10 mg/kg bodyweight, twice daily. Vehicle, dmPGE2 and diclofenac treatment was started one day prior to laparotomy and continued until the end of the experiments. Mice were sacrificed at 5 days postoperatively to ensure detection of anastomotic leakage, or when humane endpoints were reached.

48

Cyclooxygenase-2

Endpoints Anastomotic leakage was defined as one of the following: fecal peritonitis or abscess formation around the anastomosis. Anastomotic leakage was diagnosed by two independent investigators who were blinded for genotypes. Tissue preparation In anesthetized mice, a 1 cm segment of colonic tissue surrounding the anastomosis was dissected in longitudinal direction, thereby dividing it into equal parts, and fixed in formalin for immunohistochemistry purposes, or snap frozen in liquid nitrogen for qPCR. Immunohistochemistry For immunohistochemistry, sections were deparaffinized in xylene and rehydrated in graded ethanol to distilled water. Endogenous peroxidase activity was blocked using 0.6% hydrogen peroxide in methanol for 30 min. Non-specific antibody binding was blocked using 5% bovine serum albumin (BSA) and sections were incubated with primary antibody rabbit anti-human myeloperoxidase (MPO, DakoCytomation, Glostrup, Denmark), which cross-reacts with mouse). Biotinconjugated swine anti-rabbit IgG (DakoCytomation) was used followed by incubation with the streptavidin-biotin-HRP system (DakoCytomation). Binding of primary antibody was visualized with 3,3′-diaminobenzidine-tetrahydrochloridedihydrate (Sigma, St Louis, MO) and counterstained with haematoxylin. No staining was detected in slides incubated without primary antibody. For CD31 staining, an enzymatic antigen retrieval step was used prior to the nonspecific antibody blocking step using 0.1% trypsin (Difco Laboratories, Detroit, MI) in 0.1% CaCl2 solution for 20 minutes at 37ºC. Sections were incubated overnight at 4ºC with rat anti-mouse CD31 primary antibody (BD Pharmingen, Breda, the Netherlands). After washing, biotinylated rabbit anti-rat IgG (DakoCytomation) was used as secondary antibody, followed by incubation with Brightvision poly HRP antirabbit IgG (Immunologic, Duiven, the Netherlands). Binding of primary antibodies was visualized with 3,3’-diaminobenzidinetetrahydrochloride-dihydrate (Sigma). Sections were counterstained with haematoxylin. No staining was detected in slides incubated without primary antibody. The CD31 staining was quantified in

49

Chapter 3

a blinded way by two observers by counting all CD31 positive vessels at 200x magnification and expressed as the number of vessels per total area (mm2) using ImageJ (NIH Software, Bethesda, MD). qPCR RNA was isolated from snap-frozen anastomotic tissue samples with AllPrep DNA/RNA/Protein kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. In short, samples were crushed with a pestle and mortar in liquid nitrogen. Disruption and homogenisation of the tissue was performed using an Ultra Turrax Homogeniser (IKA Labortechnik, Staufen, Germany) in lysis buffer containing β-mercaptoethanol (Promega, Madison, WI). RNeasy spin columns were used to bind RNA. Columns were washed and RNA was eluted in RNase-free water. To analyse gene expression, qPCR was performed. All samples were treated with RNAse (Promega) to ensure complete removal of genomic DNA. Quantity was measured using the NanoDrop spectrophotometer (Thermo Scientific, Wilmington, DE). Only RNA samples with a clearly visible S28 and S18 on agarose gel were considered as intact RNA and were used. Total cDNA was synthesized using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA). qPCR reactions were performed on 10 ng cDNA with 300 nM of gene-specific forward and reverse primers and 1 × Absolute qPCR SYBR Green Fluorescein Mix (Bioline, London, United Kingdom) using the MyIQ system (Bio-Rad). Primers used were m-VegfA-f1 (TATTCAGCGGACTCACCAGC); and m-VegfA-r1 (CCTCCTCAAACCGTTGGCA). Gene expression levels were calculated with IQ5 software using a ∆Ct relative quantification model. The geometric mean of two internal control genes (β2microglobulin and cyclophilin A) was calculated and used as a normalization factor. Statistics Statistical analysis was performed using Prism 5.0 for Windows (Graphpad software, Inc, San Diego, CA) and SPSS 20.0 for Windows (SPSS Inc, Chicago, IL). Normality was tested using Kolmogorov-Smirnov. All continuous variables are presented as mean and standard error of the mean (SEM) and compared using students t-test. Dichotomous variables were compared using Chi-square test. Survival was analyzed by logrank test.

50

Cyclooxygenase-2

Results Diclofenac is associated with high rates of colonic anastomotic leakage in an experimental model To study whether the observed retrospective human data on colonic anastomic leakage in patients receiving COX-2 inhibitors were also observed in mice, anastomotic leakage rates in mice receiving either vehicle or NSAIDs were determined. Three out of eleven mice in the wildtype group developed anastomotic leakage (27%). Intriguingly, all 9 mice receiving diclofenac (100%) developed anastomotic leakage, p=0.001 (Figure 1A). In line, survival was significantly reduced in the group receiving diclofenac compared to the group receiving vehicle (hazard ratio (HR), 17.9 (95% confidence interval (CI), 3.7 – 87.4), p38°C or 90 bpm; (3) respiratory rate >20 breaths/min or PaCO2 12x109/L, 10% immature (band) forms plus documented infection and hypotension despite adequate fluid resuscitation (in case of septic shock) 18), and postoperative mortality (within 30 days postoperatively or within same period of hospital admission). The 30-day mortality for patients discharged within this period was verified by checking the municipal personal records database. Pre-operative CT-based muscle measurements All patients underwent a CT-scan of the abdomen as part of routine pre-operative assessment, which was delivered on CD-ROM or DVD. Measurements were performed using Osirix® Version 3.3 (32-bit; http://www.osirix-viewer.com). The cross-sectional skeletal muscle surface (cm²) assessment of sarcopenia was performed at the level of the third lumbar vertebra (L3) on two consecutive transversal coupes on which both vertebral spines were visible 19. The ‘Grow Region (2D/3D Segmentation)’ tool in the menu of the program facilitated to automatically select all skeletal muscle mass in one coupe. The distinction between different tissues is based on Hounsfield Units (HU). A threshold range

64

Functional compromise and outcome of surgery

of -30 HU to +110 HU was used for skeletal muscle. Muscles measured were: psoas, paraspinal, transverse abdominal, external oblique, internal oblique and rectus abdominis muscles (Figure 1). Hand-adjustment of the selected areas was performed if necessary and the muscle area was calculated automatically 19. The averages of the two measurements were used for calculations. Two investigators independently measured all L3-muscle area surface parameters (J.v.V. and J.T.). A third investigator (K.R.) performed a random control measurement on 10% of the CT-scans. All investigators did not have specific skills in radiology. Sarcopenia The L3-muscle area surfaces were normalized for patient height to calculate the L3-muscle index and expressed in cm²/m². The cut-off values used for sarcopenia were 52.4 cm²/m² for men and 38.5 cm²/m² for women, based on the method of Prado et al. 13. Groningen Frailty Indicator (GFI) The GFI has been developed as a simple screening instrument for frailty 16. The GFI screens on physical, cognitive, social and emotional items (Appendix 1). The maximum score is 15 points. Patients scoring 5 or more points were considered frail 16. A trained nurse routinely performed GFI-scores at pre-operative consultation or hospital admission in patients aged 70 or older. Short Nutritional Assessment Questionnaire (SNAQ) The SNAQ is a valid and reproducible tool to detect malnourished hospitalized patients without the need to calculate percentage weight loss or BMI 17. The maximum score is 5 points (Appendix 2). Patients with a score of 3 points or more on the SNAQ were classified as malnourished (requiring nutritional support and supervision by a dietician) 17. A trained nurse routinely performed SNAQ-scores at pre-operative consultation or hospital admission. Statistical analysis Frequencies are presented as absolute numbers and percentages. Continuous data are presented as mean (standard error of the mean [SEM]). Normality was tested using Kolmogorov-Smirnov. Differences between groups were analyzed

65

Chapter 4

with the Pearson chi-square test for dichotomous parameters. Odds ratios and 95% confidence intervals were calculated by logistic regression analysis. For the calculation of significant predictors of mortality and complications, univariate analyses with clinically relevant parameters were performed. Significant predictors (p