University of Groningen

Factors associated with outcome of surgery and hepatocellular carcinoma Alkozai, Edris M.

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record

Publication date: 2016

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA): Alkozai, E. M. (2016). Factors associated with outcome of liver surgery and hepatocellular carcinoma. Rijksuniversiteit Groningen.

Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment.

Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

Download date: 05-10-2021

FACTORS ASSOCIATED WITH OUTCOME OF LIVER SURGERY AND HEPATOCELLULAR CARCINOMA

Edris M. Alkozai

This thesis has been funded by the Mozaiek grant of the Dutch Organisation of Scientific Research (017.007.115), University of Groningen, the Jan Kornelis de Cock Foundation, and European Society of Organ Transplantation (ESOT)

Publication of this thesis was financially supported by the University of Groningen, and Groningen University Institute for Drug Exploration (GUIDE)

Their support is gratefully acknowledged.

© E.M. Alkozai, 2016, the Netherlands All rights are reserved. No parts of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means - electronic, mechanical, photocopy, recording or otherwise - without prior written permission of the author or the corresponding journal.

ISBN: 978-90-9030104-4 [druk] ISBN: 978-90-826307-0-1 [digitaal]

Printed by Ipskamp Enschede

Factors Associated with Outcome of Liver Surgery and Hepatocellular Carcinoma

Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen op gezag van de rector magnificus prof. dr. E. Sterken en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

dinsdag 6 december 2016 om 11.00 uur

door

Edris M. Alkozai

geboren op 29 april 1984 te Kabul, Afghanistan Promotores Prof. dr. J.A. Lisman Prof. dr. R.J. Porte

Copromotor Dr. M.W.N. Nijsten

Beoordelingscommissie Prof. dr. K.N. Faber Prof. dr. J.E. Tulleken Prof. dr. S.W.M. Olde Damink

Paranimfen Dr. Bakhtawar Khan Mahmoodi Dr. Freeha Arshad

TABLE OF CONTENTS

Chapter 1 Introduction and Aims of the Thesis 11

Chapter 2 Bleeding in Liver Surgery: Prevention and Treatment 21 Clinics in Liver Disease 2009; 13: 145-154

Chapter 3 Normal to Increased Thrombin Generation in Patients after a Major 37 Partial Liver Resection despite Prolonged Conventional Tests Alimentary Pharmacology and Therapeutics 2015; 41: 189-198

Chapter 4 Immediate Post-operative Low Count is Associated with 53 Delayed Liver Function Recovery after Partial Liver Resection Annals of Surgery 2010; 251: 300-306

Chapter 5 Evidence Against a Role of Serotonin in Liver Regeneration in 69 Humans Hepatology 2015; 62: 983

Chapter 6 No Evidence for Increased Platelet Activation in Patients with 75 Hepatitis B- or C-Related Cirrhosis and Hepatocellular Carcinoma. Thrombosis Research 2015; 135: 292–297

Chapter 7 Levels of Angiogenic Proteins in Plasma and are not 89 Different in Patients with Infectious Hepatitis Related Cirrhosis and Patients with Cirrhosis and Hepatocellular Carcinoma. Platelets, 2015; 26: 577-82

Chapter 8 Early Elevated Serum Gamma Glutamyl Transpeptidase after Liver 103 Transplantation is Associated with Better Survival. F1000Res. 2014; 3:85

Chapter 9 Systematic Comparison of Routine Laboratory Measurements with 117 Outcomes Identifies a Paradoxical Relation of Gamma-Glutamyl Transpeptidase With Mortality: ICU-Labome, a Large Cohort Study of Critically Ill Patients Submitted for publication

Chapter 10 Discussion and Perspectives 137 Nederlandse samenvatting 153 Acknowledgements (Dankwoord) 159 List of Publications 165 Curriculum Vitae 167

CHAPTER 1

GENERAL INTRODUCTION AND RATIONALE OF THE THESIS

I N T R O D U C T I O N A N D R A T I O N A L E

______12

The liver is the largest internal organ in the body and it has several unique functions. These include synthesis of clotting factors involved in coagulation and ,1 carbohydrate and lipid , detoxification, and bile production that is essential for absorption of fat and other lipophilic nutrients. A well-functioning liver is therefore pivotal for human’s well-being and survival.2,3 Acute and chronic liver failure represent major hurdles worldwide resulting over 1 million deaths a year.4,5 In developing countries viral infections such as hepatitis A, B, and E are the predominant causes of liver disease.6,7 In contrast, the most common etiologies of cirrhosis in western societies are hepatitis C virus infection (HCV), excessive alcohol intake, genetic and inherited disorders, autoimmune conditions, and non- alcoholic fatty liver diseases (NAFLD).5,8 Although cirrhotic patients can be asymptomatic, progressive deterioration may lead to the development of end stage liver disease (ESLD). ESLD is associated with severe complications including portal hypertension, ascites, hepatic encephalopathy,5 and development of hepatocellular carcinoma (HCC). HCC has a 5-year cumulative risk of 15% and 10% in endemic areas and developed countries, respectively.9,10 In fact, HCC is responsible for 50% to 70% of liver-related mortality in patients with compensated cirrhosis.11,12

Despite medical advances, partial liver resection and liver transplantation are the only curative options for patients with end stage acute or chronic liver failure, or cancer of the liver. However, postoperative complications still occur in many patients 13-15 because the liver remnant or grafts are too small or of poor quality to maintain the synthetic, excretory and detoxifying functions of the liver.16,17 Consequently, multiple organ failure, sepsis, and death can occur within days after surgery 17,18 unless prompt and sufficient liver regeneration occurs.

Liver regeneration starts immediately following partial liver resection.19 Hepatocyte proliferation terminates within days in humans.3 However, Kele and associates20 showed that the remnant liver did not regenerate to its preoperative total liver volume after six months of surgery. The mechanism of liver regeneration has been studied extensively,21 yet little is known about factors promoting tissue repair and liver regeneration in humans following surgical procedures involving the liver.21,22 Studies in rodents such as mice show that blood platelets play a pivotal role in liver regeneration.23-26 Induction of thrombocytosis by administration of a single dose of thrombopoietin or by splenectomy promoted liver regeneration.23 Conversely, antibody-induced depletion of platelets and inhibition of platelet function using , a P2Y12 receptor blocker significantly reduced proliferation of hepatocytes.23-26 Also, transfusion of platelets or platelet-rich plasma promotes liver regeneration in mice.27,28

Although the exact mechanism of platelet-mediated liver regeneration is yet to be determined, there is evidence that platelet derived growth factors and serotonin, which are

I N T R O D U C T I O N A N D R A T I O N A L E

______13

stored within the platelet granules, play a key role in proliferation of hepatocytes.26,29 Nevertheless, the role of platelets and bioactive molecules released by platelets in regeneration in humans is incompletely understood.

HEMOSTASIS IN PATIENTS WITH CIRRHOSIS AND HEPATOCELLULAR CARCINOMA

Given the critical role of liver in coagulation, fibrinolysis, and platelet metabolism,1 patients with ESLD frequently have substantial abnormalities in their hemostatic system. Conventional coagulation assays such as the prothrombin time (PT) / international normalized ratio (INR) and activated partial thromboplastin time (APTT) are very often prolonged in patients with liver disorders reflecting hypocoagulability in these patients.30 However, conventional coagulation tests are poor predictors of bleeding in patients with liver disease.30,31 In fact, PT and APTT assays may not accurately reflect the hemostatic status since these assays only measure the pro-coagulant activity and ignore the natural anti- coagulant systems,30,32 Importantly, many centers tend to correct these in vitro measures prophylactically prior to or during invasive procedures, exposing the patients to transfusion- related risks.33-35 In addition, the incorrect assumption that patients with liver disease are “auto-anticoagulated” is (partly) based on wrongful interpretation of routine hemostasis tests.36,37 In fact, patients with ESLD are at risk for both bleeding and thrombotic complications.

Although cirrhotic patients commonly have a decreased platelet count,38,39 the development of HCC in these patients is associated with elevated platelet count.40-42 Given the interaction between the cancer and platelets and the increased risk of VTE in patients with HCC,44,45 elevated platelet counts in cirrhotic patients in the presence of HCC might be considered a paraneoplastic manifestation of HCC.41,43 In fact, thrombotic complications are a major cause of death in patients with cancer.46 Yet, little is known about the interaction between the platelets and HCC in the presence of cirrhosis.47 In animal models, it has been shown that circulating platelets take up and sequester angiogenesis regulators in the presence of a microscopical (<1mm) human tumor.48 A recent study in humans showed that angiogenic proteins were elevated in platelets from patients with colorectal cancer,49 and platelet levels of these proteins were independent predictors of the presence of a malignancy. It has been established that platelet-induced angiogenic molecules are involved in extra-hemostatic platelet effects such as liver regeneration, tumor development and progression, and inflammation.50 However, it is not clear if the development of HCC in cirrhotic patients affects platelet biology or the levels of bioactive molecules that are stored within platelet granules.

I N T R O D U C T I O N A N D R A T I O N A L E

______14

PREDICTORS OF OUTCOME FOLLOWING LIVER SURGERY

After liver surgery, all (LFTs) may be abnormal including aspartate aminotransferase (AST), alanine aminotransferase (ALT), total , albumin, antithrombin, PT / INR, and gamma-glutamyltransfrase (GGT).51,52 Hence, LFT derangements may reflect physiologic changes rather than pathologic conditions following procedures involving the liver. Nevertheless, LFTs may predict morbidity and mortality beyond a certain threshold value at a certain time point following liver surgery.15,53 As such, total bilirubin, PT/INR, albumin, antithrombin (AT), AST, ALT, and clinical signs such as ascites and encephalopathy have been consistently used in the literature to predict outcomes following partial liver resection and liver transplantation.15,17,53 Also blood platelet number decreases transiently following procedures involving the liver. Given the key role of platelets in liver regeneration,26,54 immediate postoperative platelet count may accurately predict postoperative outcomes.

Contrary to other LFTs, an elevated (i.e., more deviated from the normal) GGT has shown to be associated with an improved hospital survival following surgical correction of a ruptured abdominal aortic aneurysm.55 However, in the general population, a chronically elevated GGT is associated with an elevated all-cause mortality related to cardiovascular disease, metabolic syndrome and cancer.56-59 GGT is a sensitive test of liver disease regardless of its causes, but it lacks specificity.60,61 GGT increases following procedures that involve the liver, and it has been suggested as a biomarker of biliary epithelial injury.62 Yet, little is known about the clinical relevance of elevated GGT following liver surgery.

AIMS OF THE THESIS

This thesis consists of clinical and pre-clinical research that aims to gain better understanding of factors influencing the outcome of liver surgery and the paradoxical role of GGT following liver transplantation and surgery. In addition, two studies evaluate the impact of HCC development on hemostasis in cirrhotic patients.

I N T R O D U C T I O N A N D R A T I O N A L E

______15

REFERENCES

1. Maria T. DeSancho and Stephen M. Pastores. The liver and coagulation. In: Juan Rodés, Jean- Pierre Benhamou, Andres T. Blei, Jürg Reichen, Mario Rizzetto, ed. Textbook of Hepatology: From Basic Science to Clinical Practice. Third Edition ed. , 2008:255-263. 2. Fausto N, Campbell JS, Riehle KJ. Liver regeneration. Hepatology 2006;43:S45-53. 3. Michalopoulos GK. Liver regeneration. J Cell Physiol 2007;213:286-300. 4. Mokdad AA, Lopez AD, Shahraz S, et al. Liver cirrhosis mortality in 187 countries between 1980 and 2010: a systematic analysis. BMC Med 2014;12:145,014-0145-y. 5. Blachier M, Leleu H, Peck-Radosavljevic M, Valla DC, Roudot-Thoraval F. The burden of liver disease in Europe: a review of available epidemiological data. J Hepatol 2013;58:593-608. 6. Cheng EY, Zarrinpar A, Geller DA, Goss JA, Busuttil RW. Liver. In: Schwartz SI, Brunicardi FC, Andersen DK, et al, eds. Schwartz's principles of surgery. 10e ed. New York: McGraw-Hill Education, 2014. 7. Bernal W, Wendon J. Acute liver failure. N Engl J Med 2013;369:2525-34. 8. Adam R, Hoti E. Liver transplantation: the current situation. Semin Liver Dis 2009;29:3-18. 9. Mazzanti R, Gramantieri L, Bolondi L. Hepatocellular carcinoma: epidemiology and clinical aspects. Mol Aspects Med 2008;29:130-43. 10. Fattovich G, Stroffolini T, Zagni I, Donato F. Hepatocellular carcinoma in cirrhosis: incidence and risk factors. Gastroenterology 2004;127:S35-50. 11. Sangiovanni A, Del Ninno E, Fasani P, et al. Increased survival of cirrhotic patients with a hepatocellular carcinoma detected during surveillance. Gastroenterology 2004;126:1005-14. 12. Benvegnu L, Gios M, Boccato S, Alberti A. Natural history of compensated viral cirrhosis: a prospective study on the incidence and hierarchy of major complications. Gut 2004;53:744-9. 13. Dokmak S, Fteriche FS, Borscheid R, Cauchy F, Farges O, Belghiti J. 2012 Liver resections in the 21st century: we are far from zero mortality. HPB (Oxford) 2013;. 14. Lesurtel M, Clavien PA. 2010 International Consensus Conference on Liver Transplantation for Hepatocellular Carcinoma: texts of experts. Liver Transpl 2011;17 Suppl 2:S1-5. 15. Balzan S, Belghiti J, Farges O, et al. The "50-50 criteria" on postoperative day 5: an accurate predictor of liver failure and death after hepatectomy. Ann Surg 2005;242:824,8, discussion 828-9. 16. Tian Y, Jochum W, Georgiev P, Moritz W, Graf R, Clavien PA. Kupffer cell-dependent TNF- alpha signaling mediates injury in the arterialized small-for-size liver transplantation in the mouse. Proc Natl Acad Sci U S A 2006;103:4598-603. 17. Rahbari NN, Garden OJ, Padbury R, et al. Posthepatectomy liver failure: a definition and grading by the International Study Group of Liver Surgery (ISGLS). Surgery 2011;149:713-24. 18. Dahm F, Georgiev P, Clavien PA. Small-for-size syndrome after partial liver transplantation: definition, mechanisms of disease and clinical implications. Am J Transplant 2005;5:2605-10. 19. Kurinna S, Barton MC. Cascades of transcription regulation during liver regeneration. Int J Biochem Cell Biol 2011;43:189-97.

I N T R O D U C T I O N A N D R A T I O N A L E

______16

20. Kele PG, de Boer M, van der Jagt EJ, Lisman T, Porte RJ. Early hepatic regeneration index and completeness of regeneration at 6 months after partial hepatectomy. Br J Surg 2012;99:1113-9. 21. Michalopoulos GK. Liver regeneration after partial hepatectomy: critical analysis of mechanistic dilemmas. Am J Pathol 2010;176:2-13. 22. Lisman T, Porte RJ. The role of platelets in liver inflammation and regeneration. Semin Thromb Hemost 2010;36:170-4. 23. Murata S, Matsuo R, Ikeda O, et al. Platelets promote liver regeneration under conditions of Kupffer cell depletion after hepatectomy in mice. World J Surg 2008;32:1088-96. 24. Matsuo R, Ohkohchi N, Murata S, et al. Platelets Strongly Induce Hepatocyte Proliferation with IGF-1 and HGF In Vitro. J Surg Res 2008;145:279-86. 25. Murata S, Hashimoto I, Nakano Y, Myronovych A, Watanabe M, Ohkohchi N. Single administration of thrombopoietin prevents progression of liver fibrosis and promotes liver regeneration after partial hepatectomy in cirrhotic rats. Ann Surg 2008;248:821-8. 26. Lesurtel M, Graf R, Aleil B, et al. Platelet-derived serotonin mediates liver regeneration. Science 2006;312:104-7. 27. Takahashi K, Kozuma Y, Suzuki H, et al. Human platelets promote liver regeneration with Kupffer cells in SCID mice. J Surg Res 2013;180:62-72. 28. Matsuo R, Nakano Y, Ohkohchi N. Platelet Administration Via the Portal Promotes Liver Regeneration in Rats After 70% Hepatectomy. Ann Surg 2011;253:759-63. 29. Murata S, Ohkohchi N, Matsuo R, Ikeda O, Myronovych A, Hoshi R. Platelets promote liver regeneration in early period after hepatectomy in mice. World J Surg 2007;31:808-16. 30. Tripodi A, Caldwell SH, Hoffman M, Trotter JF, Sanyal AJ. Review article: the prothrombin time test as a measure of bleeding risk and prognosis in liver disease. Aliment Pharmacol Ther 2007;26:141-8. 31. Lisman T, Porte RJ. Rebalanced hemostasis in patients with liver disease: evidence and clinical consequences. Blood 2010;116:878-85. 32. Ewe K, Reinhardt P, Muller H, Ohler W. The bleeding time after liver biopsy does not correlate with peripheral coagulation factors. Verh Dtsch Ges Inn Med 1978;(84):1060-2. 33. Cacciarelli TV, Keeffe EB, Moore DH, et al. Effect of intraoperative on patient outcome in hepatic transplantation. Arch Surg 1999;134:25-9. 34. Hendriks HG, van der Meer J, de Wolf JT, et al. Intraoperative blood transfusion requirement is the main determinant of early surgical re-intervention after orthotopic liver transplantation. Transpl Int 2005;17:673-9. 35. Porte RJ, Hendriks HG, Slooff MJ. Blood conservation in liver transplantation: The role of . J Cardiothorac Vasc Anesth 2004;18:31S-7S. 36. Sogaard KK, Horvath-Puho E, Gronbaek H, Jepsen P, Vilstrup H, Sorensen HT. Risk of venous thromboembolism in patients with liver disease: a nationwide population-based case-control study. Am J Gastroenterol 2009;104:96-101. 37. Senzolo M, Sartori MT, Lisman T. Should we give thromboprophylaxis to patients with liver cirrhosis and coagulopathy? HPB (Oxford) 2009;11:459-64. 38. Violi F, Basili S, Raparelli V, Chowdary P, Gatt A, Burroughs AK. Patients with liver cirrhosis suffer from primary haemostatic defects? Fact or fiction? J Hepatol 2011;55:1415-27.

I N T R O D U C T I O N A N D R A T I O N A L E

______17

39. Giannini E, Botta F, Borro P, et al. Platelet count/spleen diameter ratio: proposal and validation of a non-invasive parameter to predict the presence of oesophageal varices in patients with liver cirrhosis. Gut 2003;52:1200-5. 40. Carr BI, Guerra V. Thrombocytosis and hepatocellular carcinoma. Dig Dis Sci 2013;58:1790- 6. 41. Hwang SJ, Luo JC, Li CP, et al. Thrombocytosis: a paraneoplastic syndrome in patients with hepatocellular carcinoma. World J Gastroenterol 2004;10:2472-7. 42. Nwokediuko SC IO. Quantitative Platelet Abnormalities in Patients WithHepatitis B Virus- Related Liver Disease. Gastroenterology Research 2009;2:344-9. 43. Stone RL, Nick AM, McNeish IA, et al. Paraneoplastic thrombocytosis in ovarian cancer. N Engl J Med 2012;366:610-8. 44. Pirisi M, Avellini C, Fabris C, et al. Portal vein thrombosis in hepatocellular carcinoma: age and sex distribution in an autopsy study. J Cancer Res Clin Oncol 1998;124:397-400. 45. Mori H, Hayashi K, Uetani M, Matsuoka Y, Iwao M, Maeda H. High-attenuation recent thrombus of the portal vein: CT demonstration and clinical significance. Radiology 1987;163:353-6. 46. Timp JF, Braekkan SK, Versteeg HH, Cannegieter SC. Epidemiology of cancer-associated venous thrombosis. Blood 2013;122:1712-23. 47. Lisman T, Porte RJ. Platelet function in patients with cirrhosis. J Hepatol 2012;56:993,4; author reply 994-5. 48. Klement GL, Yip TT, Cassiola F, et al. Platelets actively sequester angiogenesis regulators. Blood 2009;113:2835-42. 49. Peterson JE, Zurakowski D, Italiano JE,Jr, et al. VEGF, PF4 and PDGF are elevated in platelets of colorectal cancer patients. Angiogenesis 2012;15:265-73. 50. Smyth SS, McEver RP, Weyrich AS, et al. Platelet functions beyond hemostasis. J Thromb Haemost 2009;7:1759-66. 51. Needham P, Dasgupta D, Davies J, Stringer MD. Postoperative biochemical liver function after major hepatic resection in children. J Pediatr Surg 2008;43:1610-8. 52. Pelton JJ, Hoffman JP, Eisenberg BL. Comparison of liver function tests after hepatic lobectomy and hepatic wedge resection. Am Surg 1998;64:408-14. 53. Mullen JT, Ribero D, Reddy SK, et al. Hepatic insufficiency and mortality in 1,059 noncirrhotic patients undergoing major hepatectomy. J Am Coll Surg 2007;204:854,62; discussion 862-4. 54. Clavien PA, Graf R. Liver regeneration and platelets. Br J Surg 2009;96:965-6. 55. Haveman JW, Zeebregts CJ, Verhoeven EL, et al. Changes in laboratory values and their relationship with time after rupture of an abdominal aortic aneurysm. Surg Today 2008;38:1091-101. 56. Lee DS, Evans JC, Robins SJ, et al. Gamma glutamyl transferase and metabolic syndrome, cardiovascular disease, and mortality risk: the Framingham Study. Arterioscler Thromb Vasc Biol 2007;27:127-33. 57. Kazemi-Shirazi L, Endler G, Winkler S, Schickbauer T, Wagner O, Marsik C. Gamma glutamyltransferase and long-term survival: is it just the liver? Clin Chem 2007;53:940-6.

I N T R O D U C T I O N A N D R A T I O N A L E

______18

58. Kengne AP, Czernichow S, Stamatakis E, Hamer M, Batty GD. Gamma-glutamyltransferase and risk of cardiovascular disease mortality in people with and without diabetes: Pooling of three British Health Surveys. J Hepatol 2012;57:1083-9. 59. Breitling LP, Claessen H, Drath C, Arndt V, Brenner H. Gamma-glutamyltransferase, general and cause-specific mortality in 19,000 construction workers followed over 20 years. J Hepatol 2011;55:594-601. 60. Lum G, Gambino SR. Serum gamma-glutamyl transpeptidase activity as an indicator of disease of liver, , or bone. Clin Chem 1972;18:358-62. 61. Betro MG, Oon RC, Edwards JB. Gamma-glutamyl transpeptidase in diseases of the liver and bone. Am J Clin Pathol 1973;60:672-8. 62. Vajdova K, Smrekova R, Kukan M, Lutterova M, Wsolova L. Bile analysis as a tool for assessing integrity of biliary epithelial cells after cold ischemia--reperfusion of rat . Cryobiology 2000;41:145-52.

CHAPTER 2

BLEEDING IN LIVER SURGERY: PREVENTION AND TREATMENT

EDRIS M. ALKOZAI TON LISMAN ROBERT J. PORTE

PUBLISHED IN: STEPHEN H. CALDWELL AND ARUN J. SANYAL, EDITORS COAGULATION AND HEMOSTASIS IN LIVER DISEASE: CONTROVERSIES AND ADVANCES

CLINICS IN LIVER DISEASE 13 2009;145–154

B L E E D I NG I N L I V E R S U R G E R Y

______22

ABSTRACT

Intraoperative blood loss and transfusion of blood products are negatively associated with postoperative outcome after liver surgery. Blood loss can be minimized by surgical methods, including vascular clamping techniques, the use of dissection devices, and the use of topical hemostatic agents. Preoperative correction of coagulation tests with blood products has not been shown to reduce intraoperative bleeding and it may, in fact, enhance the bleeding risk. Maintaining a low central venous pressure has been shown to be effective in reducing blood loss during partial liver resections, and volume contraction rather than prophylactic transfusion blood products seems justified in patients undergoing major liver surgery. Although drugs have proved to be effective in reducing blood loss during liver transplantation, systemic hemostatic drugs are of limited value in reducing blood loss in patients undergoing partial liver resections.

B L E E D I NG I N L I V E R S U R G E R Y

______23

INTRODUCTION

Bleeding in major surgical procedures involving the liver, such as partial liver resection and liver transplantation, occurs almost inevitably. Although blood loss in patients undergoing liver surgery has decreased substantially during the last decade, excessive blood loss can still be a major concern in individual patients. Bleeding problems are not limited to surgical patients who have a cirrhotic liver; they may also occur in patients who have a normal liver. Extensive bleeding may require the transfusion of blood or blood products, which are associated with increased rates of morbidity and mortality.1-6 Although the mechanism of bleeding in surgical interventions is multifactorial, technical factors may be responsible for a significant amount of intraoperative and early postoperative bleeding.7 Besides surgical factors, abnormalities of the hemostatic system can contribute to bleeding during liver surgery. Hemostatic function is determined by the interaction of the vascular wall, platelets, coagulation factors, and fibrinolytic function. All these components of the hemostatic system may be abnormal in patients who have a compromised liver function, and this may contribute to excessive bleeding during liver surgery.8,9 However, despite the multiple laboratory abnormalities in the hemostatic system, patients who have cirrhosis can nowadays undergo major surgical procedures such as liver transplantation or partial liver resection without transfusion of blood products.9 Although part of this can be explained by important advances in surgical methods and techniques, it may also imply that the detected abnormalities in laboratory tests of the hemostatic system are (not always) clinically relevant. Indeed, several investigators have shown that preoperative conventional coagulation assays are a poor predictor of blood loss during liver transplantation.10,11 In addition, the correction of a prolonged prothrombin time with recombinant factor VIIa has not been shown to lead to a reduction in blood loss or transfusion requirements in patients undergoing major liver surgery.12,13

The main progress in reducing perioperative blood loss has been made through improved surgical and anesthetic techniques and through better understanding of hemostatic disorders in patients who have liver disease.7,14 The purpose of this article is to provide a clinically oriented guide to the prevention and treatment of bleeding in liver surgery. The authors discuss the developments in surgical, anesthesiologic, and pharmacologic strategies that have contributed to a reduction of blood loss during liver surgery in cirrhotic and noncirrhotic patients. The clinical relevance of different types of strategies may vary, depending on the stage of the operation. For example, topical hemostatic agents have a role in reducing blood loss from the hepatic resection surface after partial liver resection, whereas surgical techniques play a more important role during transsection of the liver parenchyma (Figure 1).

B L E E D I NG I N L I V E R S U R G E R Y

______24

SURGICAL STRATEGIES TO REDUCE BLOOD LOSS

Refinements in surgical techniques and better understanding of the liver anatomy have provided important contributions to the reduction of blood loss during liver surgery. In recent years, several new techniques have been developed to perform more complex surgical interventions in patients who have a pre-existing bleeding risk, such as patients who have liver cirrhosis (Box 1). In addition, improvements in the preoperative imaging and evaluation of the liver function reserve have contributed to a better selection of patients and a lower overall postoperative morbidity and mortality.15,16

Blood loss during a partial liver resection may vary during the three stages of the procedure (see Figure 1). The first stage, in which the efferent and afferent vessels of the part of the liver that needs to be resected are identified, is characterized by minor blood loss. An exception may be patients who have intra-abdominal adhesions caused by previous abdominal surgery and patients who have significant portal hypertension, who generally have a higher bleeding tendency. In general, the amount of surgical blood loss is the highest in the second stage of liver resection, when transection of the parenchyma is performed. In this stage, the quality of the liver tissue, the dissection method used, and the central venous pressure (CVP) may influence the extent of blood loss. Selective vascular occlusion techniques have an important role in controlling blood loss in this stage of the operation, as was recently discussed elsewhere.17-20

Van der Belt and colleagues20 studied the application of vascular occlusion methods by sending a questionnaire to 621 surgeons in Europe. Although the overall response rate was low (50%), this study provided good insight into current practice. Most of the responding surgeons indicated clamping of the liver. Complete inflow occlusion (ie, the Pringle maneuver) is the most frequently applied method in this situation. Similar results have been reported by Nakajima and colleagues,21 based on a survey of 231 hospitals in Japan. A disadvantage of using vascular inflow occlusion is the resulting ischemic injury of the liver. Intermittent clamping or ischemic preconditioning may decrease the amount of ischemic injury, especially in cirrhotic livers.21,22 However, intermittent clamping is also associated with more bleeding than continuous clamping.22 Nevertheless, it is the most frequently applied method of vascular occlusion in Europe.

In addition to vascular inflow occlusion techniques, several new methods and devices for transsection of the liver parenchyma have been developed (see table 1). The Cavitron Ultrasonic Surgical Aspirator (CUSA) is the most frequently used device, followed by precoagulation devices.20,21 Although most of these devices may contribute to a reduction of

B L E E D I NG I N L I V E R S U R G E R Y

______25

Figure 1: The mechanisms of bleeding and the relative amount of blood loss (dotted line) during the three surgical stages of partial liver resections. In general, most bleeding can be encountered the operation, blood loss is mainly caused by bleeding from the resection surface of the liver. Volume contraction and a low intravascular filling status (ie, low central venous pressure) are generally more effective in reducing blood loss in this stage than massive transfusion of blood products such as fresh-frozen plasma. blood loss during the transsection phase, some of them perform slowly and some groups have reported disappointing results.22,23 In a prospective randomized clinical trial, Lesurtel and colleagues24 compared four techniques of liver transsection in 100 noncirrhotic patients undergoing major liver resections. The conventional clamp-crashing technique was compared with CUSA, Hydro-jet, and a dissecting sealer.25 In this study, the clamp-crashing technique was associated with significantly lower blood loss, shorter resection time, and lower costs, compared with the other three techniques. So, all in all, the beneficial effects of these new devices are not entirely clear and more prospective studies will be needed to assess the role of these devices in liver surgery. In the absence of a strong advantage of any of these transsection devices, personal preference and local availability are the main factors that determine the use of a given device in a center.

ANESTHESIOLOGIC STRATEGIES TO REDUCE BLOOD LOSS

The impact of anesthesiologic care on blood loss and transfusion requirement in patients undergoing major liver surgery is mainly determined by (1) intraoperative fluid management, (2) the transfusion triggers used, and (3) the use of pharmacologic agents (the last of which

B L E E D I NG I N L I V E R S U R G E R Y

______26

Box 1: Surgical and anesthesiologic methods used to reduce blood loss in liver surgery

Surgical Piggyback (preservation of retrohepatic Inferior vena cava) Vascular clamping techniques - Inflow occlusion Continuous Pringle manoeuvre Intermittent Pringle manoeuvre - Total vascular occlusion - Intermittent vascular occlusion - Ischemic preconditioning Dissection devices for transection of liver parenchyma - Classical methods - Scalpel - “Finger-fracture” method - Clamp crushing - Ultrasonic dissection (CUSA) - Hydrojet dissection - Electro coagulation (Argon coagulation) - Radio frequency ablation-based devices Anaesthesiological Maintaining low CVP by using - Volume contraction - Phlebotomy - Vasodilatation - And if needed forced diuresis Blood products Use of pharmacologic agents - - Recombinant factor VIIa

will be discussed below). Transfusion of blood products may be required in the case of active and serious bleeding, but the value of the prophylactic use of blood products, such as fresh- frozen plasma (FFP), is currently being debated.25-27 The use of blood products, however, is highly variable and not always evidence based. For example, studies in patients undergoing liver transplantation have shown a large variability in the use of blood products among different centers and even among individual anesthesiologists within centers.27 Although excessive bleeding may, and should, be managed by the transfusion of blood products, such as FFP, platelet concentrates, and packed red blood cells (RBC),28 it is also becoming clear that no consensus currently exists on transfusion practice in liver surgery. Prospective, multicenter studies with predetermined hemostasis assessment and transfusion guidelines are needed and would improve our understanding of the correction and prevention of massive bleeding during liver surgery, with likely improvements in patient outcomes.29

B L E E D I NG I N L I V E R S U R G E R Y

______27

In addition to monitoring and correcting blood loss and associated metabolic abnormalities, anesthesiologists play a key role in reducing blood loss during liver surgery by maintaining a low CVP. Performance of surgical practice under low CVP is one of the strategies that have been studied intensively in liver surgery.30-33 Although already suggested by Bismuth and colleagues,33 Jones and colleagues32 were the first to show that blood loss during liver resection is almost linearly related to the CVP. Low CVP (<5 mm Hg) can be achieved by applying volume contraction, by using vasodilating agents, or by stimulation of forced diuresis (see Figure 1). Volume contraction has been suggested as a safe method of reducing blood loss during liver surgery. It can be achieved by a restrictive use of fluid and blood products, avoidance of fluid overload, and no routine correction of abnormal coagulation tests by infusion of FFP or other large-volume blood products.2,4,31 Although a low CVP is associated with reduced blood loss, it also carries a higher risk for complications such as air embolism, systemic tissue hypoperfusion, and renal failure.2,30,34,35 Schroeder and colleagues34 studied the safety of a fluid restriction policy and low CVP in liver transplant recipients by comparing outcome variables in two centers with different policies. One center had the policy to aim for a low CVP (<5 mm Hg) by using fluid restriction, whereas the second center did not take any specific measures to lower the CVP and aimed for normal CVP policy (7–10 mm Hg). Both patient groups were similar in demographics, cause of liver disease, and surgical methods. The low CVP group received lower amounts of RBC (3.8 versus 11.6 units, P<0.01), FFP (1.3 versus 14.7 units, P<0.001), and platelets (0.6 versus 2.4 units, P<0.001) compared with the normal CVP group. However, the postoperative peak serum level (3.2 versus 1.8 mg/dL, P<0.01), the need for dialysis (6.8% versus 1.2%, P<0.05), and 30-day mortality (6 [8.2%] versus 0, P<0.05) were higher in patients who had low CVP. A limitation of this study is the lack of randomization and the comparison of two centers, which may have differed in many other aspects than just a CVP target. Contrary to the study by Schroeder and colleagues, Wang and colleagues36 found no detrimental effect on maintaining a low CVP in a prospective study of 50 cirrhotic patient undergoing partial liver resection for hepatocellular carcinoma. Patients were divided into an intervention group (n = 25), in which the CVP was maintained at less than or equal to 4 mm Hg, and a control group (n = 25) with normal CVP. Intraoperative blood loss was significantly lower in the group with low CVP, compared with the control group (903 ± 180 mL versus 2329 ± 2538 mL, P<0.01). In addition, RBC and FFP transfusion requirements were significantly lower and hospital stay was shorter in the group with low CVP, whereas no negative effect was found in postoperative hepatic and renal function.

Some groups have taken the concept of fluid contraction much further than only reducing fluid infusions, and these groups even performed phlebotomy as a strategy to minimize intraoperative blood loss in patients undergoing major liver surgery.35,37 Hashimoto and colleagues37 performed a randomized controlled trial in 79 healthy participants who

B L E E D I NG I N L I V E R S U R G E R Y

______28 underwent partial liver resection for living donor liver transplantation. Participants were randomly allocated to a blood withdrawal group (n = 40, collecting a volume of blood corresponding to 0.7% of the patient’s body weight) or a control group (n = 39) with no blood withdrawal. Surgeons were blinded for the allocated groups. The CVP at the beginning of the parenchymal transsection was significantly lower in the group with blood withdrawal (median 5 [range 2–9] cm H2O versus 6 [range 2–13] cm H2O, P = 0.005) compared with controls. Blood loss during liver transection was also significantly lower in the phlebotomized group (140 [range 40–430] mL versus 230 [range 40–660] mL, P = 0 .034). However, the two showed no statistical difference in postoperative outcomes. In another prospective study, Massicotte and colleagues35 examined the effect of maintaining a low CVP through volume contraction and by using intraoperative phlebotomy in patients undergoing liver transplantation. Outcome in these patients was compared with outcome in a historical control group without phlebotomy.26 Intraoperative blood loss was significantly lower in the prospective group with a low CVP (903 ± 582 mL versus 1479 ± 1750 mL, P = 0.001), and no patient required dialysis in the postoperative period.

In general, evidence is increasing that blood loss during major liver surgery is strongly influenced by the filling status and CVP of the patient. Measures to reduce the filling status of the patient and to lower the CVP through volume contraction and no routine correction of laboratory coagulation test with large-volume blood products is effective and safe. Larger prospective studies will be needed to define the exact role and safety of blood withdrawal as a measure of reducing the CVP and minimizing blood loss during liver surgery.

PHARMACOLOGIC STRATEGIES TO REDUCE BLOOD LOSS

Several pharmacologic measures are available to treat or prevent bleeding complications during liver surgery. However, these agents should only be used as complementary to other methods in reducing blood loss. Three main categories can be recognized: topical hemostatic agents, antifibrinolytic drugs, and procoagulant drugs.38

TOPICAL HEMOSTATIC AGENTS Topical agents may be useful to stimulate hemostasis at the resection surface of the liver after parenchymal transsection. Based on their working mechanism, topical agents can be divided into three groups: agents that mimic coagulation (ie, sealants), agents that provide a matrix for endogenous coagulation (ie, collagen, gelatin, and cellulose sponges), and combined products that work as a matrix for endogenous and exogenous coagulation factors.38,39 Current scientific evidence suggests beneficial effects in reducing the time to hemostasis and in lowering the requirements for perioperative RBC transfusions.39-43

B L E E D I NG I N L I V E R S U R G E R Y

______29

Although the beneficial effects of fibrin sealants have also been confirmed in a recent Cochrane review,44 the efficacy of fibrin sealant in liver surgery has recently been questioned.45 In a large, randomized controlled trial in 300 patients undergoing partial liver resection, Figueras and associates45 found no difference in total blood loss, transfusion requirements, or postoperative morbidity between patients treated with fibrin sealants (n=150) and a control group without fibrin sealants (n = 150).

ANTIFIBRINOLYTICS Antifibrinolytics can be categorized into two groups: inhibitors of plasminogen (lysineanalogs tranexamic acid and epsilon-aminocaproic acid), and inhibitors of plasmin (serine protease inhibitors aprotinin and nafamostat mesylate). In recent years, several studies and reviews have been published on the efficacy and safety of antifibrinolytics in liver surgery and transplantation.14,38,46-49 In liver transplantation, aprotinin and tranexamic acid have been shown to result in a significant reduction in blood loss and transfusion requirements of around 30% to 40%.50 Because of recent safety concerns, especially a higher risk for renal failure and perioperative death in patients who were given aprotinin during cardiac surgery, marketing of aprotinin has recently been suspended. However, in the liver transplant population, prospective studies have not caused any safety concerns, and no increased risk for thromboembolic events or renal failure has been noted in liver transplant patients treated with aprotinin.50,51 Although antifibrinolytics have been studied extensively in liver transplantation, only two prospective studies have examined the efficacy in patients undergoing liver resections.52,53 In general, improvements in surgical technique and anesthesiologic care seem to be more important in reducing blood loss in patients undergoing partial liver resections than the use of the antifibrinolytic drugs. Antifibrinolytics may be indicated in a selected group of patients who have cirrhosis and are undergoing liver resection, but further studies in this specific group of patients will be needed.54

PROCOAGULANT DRUGS The efficacy and safety of the recombinant factor VIIa has been studied in several randomized clinical trials in cirrhotic and noncirrhotic patients undergoing partial liver resections or transplantation.12,13,55-57 Although these studies did not cause major safety concerns,38,58,59 they also failed to demonstrate a significant difference in blood loss or transfusion requirements between patients who received recombinant factor VIIa or placebo. In all of these studies, recombinant factor VIIa was used as a prophylactic drug, which may not be the most efficient use for this drug. Probably, this drug should be seen more as a drug that can be used a ‘‘rescue therapy’’ to control bleeding in situations of major bleeding where other therapies have failed. More research in this area is needed.

B L E E D I NG I N L I V E R S U R G E R Y

______30

SUMMARY

In general, perioperative blood loss and blood transfusions have a negative impact on postoperative outcome after liver surgery. Surgical technique and experience are key factors determining the amount of blood loss in liver surgery. Inflow occlusion (the Pringle maneuver) and the use of low CVP are simple and effective measures of reducing blood loss during parenchyma transsection. No superiority of one dissection device has been shown above the others, and their use depends mainly on the quality of the liver parenchyma and personal preference and experience. The emerging evidence indicates that abnormal coagulation tests do not predict bleeding in cirrhotic patients. Preprocedural correction of coagulation tests with blood products has not been shown to reduce intraoperative bleeding and it even seems counterproductive because it results mainly in an increase of the intravascular filling status of the patient, which may, in fact, enhance the bleeding risk. Factors such as portal hypertension and the hyperdynamic circulation in patients who have cirrhosis may play a more important role in the bleeding tendency of these patients. Therefore, volume contraction, rather than prophylactic transfusion blood products (ie, FFP), seems justified in patients undergoing major liver surgery. An increasing number of studies suggest that volume contraction in these patients is safe and effective in reducing perioperative blood loss and transfusion requirements. Although antifibrinolytic drugs proved to be effective in reducing blood loss during liver transplantation, topical or systemic hemostatic drugs are of limited value in reducing blood loss in patients undergoing partial liver resections.

B L E E D I NG I N L I V E R S U R G E R Y

______31

REFERENCE LIST

1. Hendriks HG, van der Meer J, de Wolf JT, et al. Intraoperative blood transfusion requirement is the main determinant of early surgical re-intervention after orthotopic liver transplantation. Transpl Int 2005;17:673-9. 2. Cacciarelli TV, Keeffe EB, Moore DH, et al. Effect of intraoperative blood transfusion on patient outcome in hepatic transplantation. Arch Surg 1999;134:25-9. 3. Stainsby D, Williamson L, Jones H, Cohen H. 6 Years of shot reporting--its influence on UK blood safety. Transfus Apher Sci 2004;31:123-31. 4. de Boer MT, Molenaar IQ, Hendriks HG, Slooff MJ, Porte RJ. Minimizing blood loss in liver transplantation: progress through research and evolution of techniques. Dig Surg 2005;22:265- 75. 5. Porte RJ, Hendriks HG, Slooff MJ. Blood conservation in liver transplantation: The role of aprotinin. J Cardiothorac Vasc Anesth 2004;18:31S-7S. 6. Ramos E, Dalmau A, Sabate A, et al. Intraoperative red blood cell transfusion in liver transplantation: influence on patient outcome, prediction of requirements, and measures to reduce them. Liver Transpl 2003;9:1320-7. 7. Marietta M, Facchini L, Pedrazzi P, Busani S, Torelli G. Pathophysiology of bleeding in surgery. Transplant Proc 2006;38:812-4. 8. Porte RJ, Knot EA, Bontempo FA. Hemostasis in liver transplantation. Gastroenterology 1989;97:488-501. 9. Lisman T, Leebeek FW. Hemostatic alterations in liver disease: a review on pathophysiology, clinical consequences, and treatment. Dig Surg 2007;24:250-8. 10. Findlay JY, Rettke SR. Poor prediction of blood transfusion requirements in adult liver transplantations from preoperative variables. J Clin Anesth 2000;12:319-23. 11. Steib A, Freys G, Lehmann C, Meyer C, Mahoudeau G. Intraoperative blood losses and transfusion requirements during adult liver transplantation remain difficult to predict. Can J Anaesth 2001;48:1075-9. 12. Lodge JP, Jonas S, Oussoultzoglou E, et al. Recombinant coagulation factor VIIa in major liver resection: a randomized, placebo-controlled, double-blind clinical trial. Anesthesiology 2005;102:269-75. 13. Planinsic RM, van der Meer J, Testa G, et al. Safety and efficacy of a single bolus administration of recombinant factor VIIa in liver transplantation due to chronic liver disease. Liver Transpl 2005;11:895-900. 14. Groenland TH, Porte RJ. Antifibrinolytics in liver transplantation. Int Anesthesiol Clin 2006;44:83-97. 15. Friedman LS. The risk of surgery in patients with liver disease. Hepatology 1999;29:1617- 23. 16. Suman A, Carey WD. Assessing the risk of surgery in patients with liver disease. Cleve Clin J Med 2006;73:398-404.

B L E E D I NG I N L I V E R S U R G E R Y

______32

17. van Gulik TM, de Graaf W, Dinant S, Busch OR, Gouma DJ. Vascular occlusion techniques during liver resection. Dig Surg 2007;24:274-81. 18. Smyrniotis V, Farantos C, Kostopanagiotou G, Arkadopoulos N. Vascular control during hepatectomy: review of methods and results. World J Surg 2005;29:1384-96. 19. Dixon E, Vollmer CM,Jr, Bathe OF, Sutherland F. Vascular occlusion to decrease blood loss during hepatic resection. Am J Surg 2005;190:75-86. 20. van der Bilt JD, Livestro DP, Borren A, van Hillegersberg R, Borel Rinkes IH. European survey on the application of vascular clamping in liver surgery. Dig Surg 2007;24:423-35. 21. Nakajima Y, Shimamura T, Kamiyama T, Matsushita M, Sato N, Todo S. Control of intraoperative bleeding during liver resection: analysis of a questionnaire sent to 231 Japanese hospitals. Surg Today 2002;32:48-52. 22. Selzner N, Rudiger H, Graf R, Clavien PA. Protective strategies against ischemic injury of the liver. Gastroenterology 2003;125:917-36. 23. Takayama T, Makuuchi M, Kubota K, et al. Randomized comparison of ultrasonic vs clamp transection of the liver. Arch Surg 2001;136:922-8. 24. Lesurtel M, Selzner M, Petrowsky H, McCormack L, Clavien PA. How should transection of the liver be performed?: a prospective randomized study in 100 consecutive patients: comparing four different transection strategies. Ann Surg 2005;242:814,22. 25. Lisman T, Caldwell SH, Porte RJ, Leebeek FW. Consequences of abnormal hemostasis tests for clinical practice. J Thromb Haemost 2006;4:2062-3. 26. Massicotte L, Sassine MP, Lenis S, Roy A. Transfusion predictors in liver transplant. Anesth Analg 2004;98:1245,51. 27. Ozier Y, Pessione F, Samain E, Courtois F, French Study Group on Blood Transfusion in Liver Transplantation. Institutional variability in transfusion practice for liver transplantation. Anesth Analg 2003;97:671-9. 28. Kang Y, Audu P. Coagulation and liver transplantation. Int Anesthesiol Clin 2006;44:17-36. 29. Lopez-Plaza I. Transfusion guidelines and liver transplantation: time for consensus. Liver Transpl 2007;13:1630-2. 30. Melendez JA, Arslan V, Fischer ME, et al. Perioperative outcomes of major hepatic resections under low central venous pressure anesthesia: blood loss, blood transfusion, and the risk of postoperative renal dysfunction. J Am Coll Surg 1998;187:620-5. 31. Smyrniotis V, Kostopanagiotou G, Theodoraki K, Tsantoulas D, Contis JC. The role of central venous pressure and type of vascular control in blood loss during major liver resections. Am J Surg 2004;187:398-402. 32. Jones RM, Moulton CE, Hardy KJ. Central venous pressure and its effect on blood loss during liver resection. Br J Surg 1998;85:1058-60. 33. Bismuth H, Castaing D, Garden OJ. Major hepatic resection under total vascular exclusion. Ann Surg 1989;210:13-9. 34. Schroeder RA, Collins BH, Tuttle-Newhall E, et al. Intraoperative fluid management during orthotopic liver transplantation. J Cardiothorac Vasc Anesth 2004;18:438-41.

B L E E D I NG I N L I V E R S U R G E R Y

______33

35. Massicotte L, Lenis S, Thibeault L, Sassine MP, Seal RF, Roy A. Effect of low central venous pressure and phlebotomy on blood product transfusion requirements during liver transplantations. Liver Transpl 2006;12:117-23. 36. Wang WD, Liang LJ, Huang XQ, Yin XY. Low central venous pressure reduces blood loss in hepatectomy. World J Gastroenterol 2006;12:935-9. 37. Hashimoto T, Kokudo N, Orii R, et al. Intraoperative blood salvage during liver resection: a randomized controlled trial. Ann Surg 2007;245:686-91. 38. Porte RJ, Leebeek FW. Pharmacological strategies to decrease transfusion requirements in patients undergoing surgery. Drugs 2002;62:2193-211. 39. Berrevoet F, de Hemptinne B. Use of topical hemostatic agents during liver resection. Dig Surg 2007;24:288-93. 40. Heaton N. Advances and methods in liver surgery: haemostasis. Eur J Gastroenterol Hepatol 2005;17 Suppl 1:S3-12. 41. Chapman WC, Clavien PA, Fung J, Khanna A, Bonham A. Effective control of hepatic bleeding with a novel collagen-based composite combined with autologous plasma: results of a randomized controlled trial. Arch Surg 2000;135:1200,4 42. Schwartz M, Madariaga J, Hirose R, et al. Comparison of a new fibrin sealant with standard topical hemostatic agents. Arch Surg 2004;139:1148-54. 43. Jackson MR. Fibrin sealants in surgical practice: An overview. Am J Surg 2001;182:1S-7S. 44. Carless PA, Henry DA, Anthony DM. Fibrin sealant use for minimising peri-operative allogeneic blood transfusion. Cochrane Database Syst Rev 2003;(2):CD004171. 45. Figueras J, Llado L, Miro M, et al. Application of fibrin glue sealant after hepatectomy does not seem justified: results of a randomized study in 300 patients. Ann Surg 2007;245:536-42. 46. Xia VW, Steadman RH. Antifibrinolytics in orthotopic liver transplantation: current status and controversies. Liver Transpl 2005;11:10-8. 47. Ozier Y, Schlumberger S. Pharmacological approaches to reducing blood loss and transfusions in the surgical patient. Can J Anaesth 2006;53:S21-9. 48. Henry DA, Carless PA, Moxey AJ, et al. Anti-fibrinolytic use for minimising perioperative allogeneic blood transfusion. Cochrane Database Syst Rev 2007;(4):CD001886. 49. Dalmau A, Sabate A, Acosta F, et al. Tranexamic acid reduces red cell transfusion better than epsilon-aminocaproic acid or placebo in liver transplantation. Anesth Analg 2000;91:29-34. 50. Molenaar IQ, Warnaar N, Groen H, Tenvergert EM, Slooff MJ, Porte RJ. Efficacy and safety of antifibrinolytic drugs in liver transplantation: a systematic review and meta-analysis. Am J Transplant 2007;7:185-94. 51. Warnaar N, Mallett SV, de Boer MT, et al. The impact of aprotinin on renal function after liver transplantation: an analysis of 1,043 patients. Am J Transplant 2007;7:2378-87. 52. Lentschener C, Benhamou D, Mercier FJ, et al. Aprotinin reduces blood loss in patients undergoing elective liver resection. Anesth Analg 1997;84:875-81.

B L E E D I NG I N L I V E R S U R G E R Y

______34

53. Wu CC, Ho WM, Cheng SB, et al. Perioperative parenteral tranexamic acid in liver tumor resection: a prospective randomized trial toward a "blood transfusion"-free hepatectomy. Ann Surg 2006;243:173-80. 54. Pereboom IT, de Boer MT, Porte RJ, Molenaar IQ. Aprotinin and nafamostat mesilate in liver surgery: effect on blood loss. Dig Surg 2007;24:282-7. 55. Lodge JP, Jonas S, Jones RM, et al. Efficacy and safety of repeated perioperative doses of recombinant factor VIIa in liver transplantation. Liver Transpl 2005;11:973-9. 56. Meijer K, Hendriks HG, De Wolf JT, et al. Recombinant factor VIIa in orthotopic liver transplantation: influence on parameters of coagulation and fibrinolysis. Blood Coagul Fibrinolysis 2003;14:169-74. 57. Hendriks HG, Meijer K, de Wolf JT, et al. Reduced transfusion requirements by recombinant factor VIIa in orthotopic liver transplantation: a pilot study. Transplantation 2001;71:402-5. 58. Vincent JL, Rossaint R, Riou B, Ozier Y, Zideman D, Spahn DR. Recommendations on the use of recombinant activated factor VII as an adjunctive treatment for massive bleeding--a European perspective. Crit Care 2006;10:R120. 59. Levy JH, Fingerhut A, Brott T, Langbakke IH, Erhardtsen E, Porte RJ. Recombinant factor VIIa in patients with coagulopathy secondary to therapy, cirrhosis, or severe traumatic injury: review of safety profile. Transfusion 2006;46:919-33.

CHAPTER 3

NORMAL TO INCREASED THROMBIN GENERATION IN PATIENTS AFTER A MAJOR PARTIAL LIVER RESECTION DESPITE PROLONGED CONVENTIONAL COAGULATION TESTS

WILMA POTZE* EDRIS M. ALKOZAI* JELLE ADELMEIJER ROBERT J. PORTE TON LISMAN

*SHARED AUTHORSHIP

PUBLISHED IN: ALIMENT PHARMACOL THER. 2015;41:189-98

T H R O M B I N G E N E R A T I O N A F T E R L I V E R R E S E C T I O N

______38

ABSTRACT

BACKGROUND: Conventional coagulation tests are frequently prolonged after liver surgery, suggesting a post-operative hypocoagulability. However, these tests are unreliable for assessment of the haemostatic status in these patients. In contrast, thrombin generation testing measures the true balance between pro- and anti-coagulant factors.

AIM: To study the perioperative coagulation status in patients undergoing hemi- hepatectomy using thrombin generation assays.

METHODS: We examined thrombin generation profiles in serial plasma samples taken from seventeen patients undergoing right hemi-hepatectomy. Results were compared to ten patients undergoing pancreatic resection and twenty-four healthy volunteers. In addition, we measured conventional coagulation tests and plasma levels of several haemostatic proteins.

RESULTS: Following liver resection, the endogenous thrombin potential (ETP) slightly decreased until post-operative day 7. However, in the presence of thrombomodulin, the ETP increased [from 542 nM*min (417–694) at baseline to 845 nM*min (789–1050) on post-operative day 3] to values higher than that in healthy subjects (558 nM*min (390– 680); P < 0.001), which contrasts with substantially prolonged PT levels. Normal to decreased thrombin generation was observed following pancreatic resection. Thrombin generation was only slightly affected by thrombomodulin after hemi-hepatectomy, while thrombin generation in healthy subjects decreased profoundly upon addition of thrombomodulin. This hypercoagulability following liver resection may be explained by decreased levels of protein C, S, and antithrombin and by elevated levels of factor VIII.

CONCLUSIONS: Thrombin generation in the presence of thrombomodulin revealed hypercoagulability in patients following liver resection. These results support the recently advocated restrictive use of plasma during liver resection and the exploration of more extensive use of post-operative thrombosis prophylaxis.

T H R O M B I N G E N E R A T I O N A F T E R L I V E R R E S E C T I O N

______39

INTRODUCTION

Major abdominal surgery is associated with a post-operative hypercoagulable state and this hypercoagulability has been recognised as a factor that contributes to the occurrence of thromboembolic complications, such as deep vein thrombosis and pulmonary embolism.1 The underlying mechanisms include increased activation of platelets, reduced concentrations of anti-coagulants, and impaired fibrinolysis,2 as well as more general risk factors for thrombosis, such as immobilisation, tissue damage and the presence of cancer.

Conventional coagulation tests, such as the activated partial thromboplastin time (APTT), and the prothrombin time (PT) and related international normalised ratio (INR), are frequently elevated after liver surgery,3-6 suggesting a post-operative hypocoagulability. These tests are, however, not reliable for assessment of the overall haemostatic status in these patients, because they only evaluate narrow aspects of haemostasis. Specifically these tests are only sensitive for circulating levels of procoagulant factors, and do not test functionality of the natural anti-coagulant systems. Indeed, venous thromboembolism also occurs after liver surgery and, in fact, the risk increases with the extent of hepatectomy.7,8 The thrombotic risk after liver resection may result from a hypercoagulable state induced by extensive tissue injury or reduction in anti-coagulant factors determined by increased consumption, blood loss or haemodilution.9-13 Despite this, elevation of routine coagulation tests (PT, APTT) following liver resection frequently leads to transfusion of fresh frozen plasma (FFP).14-16 However, several serious side effects of blood product transfusion may occur, including the risk of infection and the risk of transfusion-related acute lung injury.17 In addition, elevation of routine coagulation tests often leads clinicians to delay thrombosis prophylaxis, based on the assumption that the patient is ‘auto-anticoagulated’. This potentially increases the risk of deep vein thrombosis and pulmonary embolism in these patients.7,18 Using (TEG), which analyses all components of the haemostatic system, it was shown that after living donor liver transplantation the majority of donors were hypercoagulable in spite of elevated routine coagulation tests.10 Also patients undergoing partial liver resection for malignant disease demonstrated a brief hypercoagulable state, followed by normal clot formation as shown by TEG,5,19 and yet another study using TEG reported normocoagulability after liver resection.6

Thromboelastography is routinely used for guiding transfusion of blood products during massive bleeding and liver transplantation by many centres.20,21 However, it is still uncertain if thromboelastography-guided transfusion strategies improve outcome in patients with massive bleeding,22 or decrease blood loss during liver transplantation.23 Furthermore, although thromboelastography assesses the haemostatic status in a whole blood environment, the technique has three major drawbacks. (1) There are multiple ways to

T H R O M B I N G E N E R A T I O N A F T E R L I V E R R E S E C T I O N

______40

perform a thromboelastographic analysis. Specifically, tests with non-anticoagulated blood as well as with citrated blood are widely used, and the results from these various methods poorly correlate.24 When using citrated blood, various triggers are added to initiate coagulation (tissue factor/kaolin). Furthermore, two devices (TEG and ROTEM) are used clinically, and results from these methods are also not always in accordance.25,26 (2) Quantification of the contribution of individual components of the haemostatic system to abnormalities in the thromboelastographic tracing is not possible, although it has been demonstrated that alterations in some components of the traces are dominated by platelets, coagulation or levels.25,26 (3) Thromboelastography suffers from a unique set of pre-analytic and analytic variables that impact test reliability and reproducibility.27

The thrombin generation assay, which measures the total amount of thrombin generated during in vitro coagulation, has been successfully used to reassess the haemostatic status of patients with liver disease.28-31 This global test, which takes plasma concentrations of both pro- and anti-coagulants into account, offers a valid alternative to the conventional coagulation tests which only test functionality of some of the procoagulant factors. Thrombin generation testing has demonstrated normal or even superior thrombin generation in patients with cirrhosis,28-31 despite a prolonged PT or APTT. It has been well established that plasma levels of haemostatic proteins decrease following a partial liver resection, which has been attributed in part to a reduced synthetic capacity of the liver remnant.6,9,10 However, the net results of these changes in plasmatic coagulation have not been established. We therefore studied the perioperative coagulation status in patients undergoing right hemi- hepatectomy using both conventional coagulation tests and thrombin generation assays. Furthermore, we compared results to those of patients undergoing a pancreatic resection which is a surgical procedure of a similar extent, but without a decrease in post-operative synthetic capacity of the liver.

PATIENTS AND METHODS

PATIENTS Seventeen adult patients, who underwent a right (n=15) or extended right (n=2) hemi- hepatectomy were included in the study. The control group consisted of patients (n=10) who underwent a pylorus preserving pancreaticoduodenectomy (PPPD). Twenty-four adult healthy volunteers were included to establish reference values for the various tests performed. All patients were included in the University Medical Center Groningen, the Netherlands. Exclusion criteria were age younger than 18 years, pre-existing coagulation disorders, pre-operative anti-coagulation, and use of non-steroidal anti-inflammatory drugs or 1 week before surgery. Routine surgical and anaesthetic procedures were

T H R O M B I N G E N E R A T I O N A F T E R L I V E R R E S E C T I O N

______41

adopted. The study protocol was approved by the local medical ethical committee and informed consent was obtained from each subject before inclusion in the study.

PLASMA SAMPLES Plasma samples for analyses were, for both groups, drawn at the following time points: after induction of anaesthesia (baseline), at the end of surgery, and on the post-operative days 1, 3, 5, 7 and 30. Following surgery, all patients received standard thromboprophylaxis with (once-daily) low molecular weight (LMWH) and at each post-operative day blood was drawn just prior to the administration of the LMWH. Blood samples from each subject were drawn by venipuncture and collected into vacuum tubes containing 3.8% trisodium citrate as an anti-coagulant, at a blood to anti-coagulant ratio of 9:1. Platelet-poor plasma was prepared by double centrifugation at 2000g and 10000g, respectively, for 10 min. at 18 °C. Plasma was aliquoted, snap-frozen and stored at -80°C until use.

THROMBIN GENERATION Thrombin generation testing was performed using platelet- poor plasma (PPP) with the fluorimetric method described by Hemker, Calibrated Automated Thrombography® (CAT).32 Coagulation was activated by using a commercial trigger composed of recombinant tissue factor (TF, final concentration 5 pM) and phospholipids (final concentration 4 lM), in the presence or absence of soluble thrombomodulin (TM). These reagents were purchased from Thrombinoscope BV, Maastricht, The Netherlands. To calibrate the thrombin generation curves, thrombin Calibrator (Thrombinoscope BV) was added. A fluorogenic substrate with CaCl2 (FluCa-kit, Thrombinoscope BV, Maastricht, the Netherlands) was dispensed in each well to permit a continuous registration of thrombin generation. Fluorescence was read in time by the fluorometer Fluoroskan Ascent (ThermoFisher Scientific, Helsinki, Finland). All procedures were followed according to the protocol suggested by Thrombinoscope B.V. Thrombin generation variables analysed were endogenous thrombin potential (ETP), peak thrombin generation, lag-time (time needed for thrombin concentration to reach 1/6th of the peak concentration) and velocity index (slope between the end of lag-time and peak thrombin generation). Furthermore, a normalised thrombomodulin sensitivity ratio (TM-SR) was determined by dividing the ETP in the presence of TM divided by the ETP in the absence of TM of an individual, by the ETP in the presence of TM divided by the ETP in the absence of TM of pooled normal plasma. A TM-SR>1 reflects a decreased anti-coagulant response to TM in comparison to pooled normal plasma.

CONVENTIONAL COAGULATION TESTS The PT was assessed on an automated coagulation analyzer (ACL 500 TOP) with reagents (Recombiplastin 2G) and protocols from the manufacturer (Instrumentation Laboratory, Breda, the Netherlands). Levels of factor (F) VIII and II, antithrombin (AT),

T H R O M B I N G E N E R A T I O N A F T E R L I V E R R E S E C T I O N

______42

and protein C were measured on an automated coagulation analyzer (ACL 500 TOP) with reagents and protocols from the manufacturer (Recombiplastin 2G for FII, Hemosil (R) SynthASil for FVIII, Liquid Antithrombin reagent for AT, and IL reagent for protein C) (Instrumentation Laboratory).

Table 1: Patient characteristics Hemi-hepatectomy PPPD Controls Characteristics (n=17) (n=10) (n=24) P Sex (male) 6 (35) 7 (70) 13 (54) 0.201 Age (years) 62 [9.8] 67 [7.9] 27 [4.5] <0.001 Surgical indications Colon cancer metastasis 9 (53) 1 (10) Hepatocellular carcinoma (HCC) 2 (12) 0 Pancreatic cancer 0 7 (70) Cholangiocarcinoma 2 (12) 1 (10) Neuroendocrine tumor 1 (5.9) 1 (10) Leiomyosarcoma 1 (5.9) 0 Adrenocortical Carcinoma metastasis 1 (5.9) 0 Benign lesion 1 (5.9) 0 Length of surgery (min) 616 (480- 650) 631 (551-739) 0.414 Estimated blood loss (ml) 700 (200-1000) 400 (300- 0.980 1125) Amount of fluids administered (ml) 4441 [1580] 5225 [1239] 0.192 RBC transfusion (units) 0 0 0.902 FFP transfusion (units) 0 0 1.000 Length of hospital stay (days) 14 (9-21) 15 (11-21) 0.863 Hemoglobin before surgery (mmol/L) * 8.6 [1.2] 8.2 [1.1] 0.407 Hemoglobin after surgery (mmol/L) * 6.7 [1.3] 6.8 [1.2] 0.841

Data are expressed as number (%), mean [s.d.], or median (interquartile range). PPPD: pylorus preserving pancreaticoduodenectomy. * To convert values for haemoglobin to g/dL, multiply by 1.650.

STATISTICAL ANALYSES Values are expressed as means (with s.d.), medians (with interquartile ranges), or numbers (with percentages) as appropriate. Differences between values of pre-operative coagulation tests and thrombin generation tests, and follow-up values were evaluated by mixed linear models. To calculate differences between continuous data among independent groups, the t-test for independent samples or the Mann–Whitney U-test, as appropriate, was used. Differences between patient values and levels measured in healthy controls were compared using one-way ANOVA (with the Bonferroni post-test) or Kruskal–Wallis H test (with Dunn’s post-test) as appropriate. P-values of 0.05 or less were considered statistically significant. GraphPad Prism (San Diego, CA, USA) and IBM SPSS Statistics 20 (New York, NY, USA) were used for analyses.

T H R O M B I N G E N E R A T I O N A F T E R L I V E R R E S E C T I O N

______43

RESULTS

PATIENT CHARACTERISTICS Seventeen patients who underwent a right (n = 15) or extended right (n = 2) hemi- hepatectomy, ten patients who underwent a PPPD, and twenty-four controls were included in the study. The main characteristics of the study population are reported in Table 1. The most common indication for liver resection was liver metastases from colorectal cancer and the most common indication for PPPD was pancreatic cancer. The median estimated blood loss was 700 mL in the patients who underwent a hemi- hepatectomy and 400 mL in the patients who underwent a PPPD. None of the patients suffered from venous thrombosis within 30 days after surgery.

CONVENTIONAL COAGULATION TESTS Baseline PT was 10.8 s (9.9–11.2) [median (interquartile range)] in the controls, 11.5 s (10.9–12.0) in the patients undergoing hepatectomy (P<0.05 compared to controls), and 11.6 s (10.6–12.2) in the patients undergoing PPPD (P>0.05 compared to controls). The PT progressively increased after both liver and pancreatic surgery (Figure 1), with a peak PT of 18.3 s (16.8–21.3) in the patients undergoing hepatectomy (P<0.001 compared to baseline PT) and 14.4 s (13.5–15.1) in the patients undergoing PPPD (P=0.002 compared to baseline PT) on post-operative day 1. After post-operative day 1 the PT decreased to baseline levels on post-operative day 3 in the patients undergoing PPPD and on post-operative day 7 in the patients undergoing hepatectomy. The APTT was 34.2 s (31.1–38.5) in the controls, 37.9 s (32.3–43.5) in the patients undergoing hepatectomy (P > 0.05 compared to controls), and 30.2 s (25.6–36.3) in the patients undergoing PPPD (P>0.05 compared to controls) at baseline. The APTT slightly decreased following hepatectomy [reaching 31.8 s (28.8–34.6) on post- operative day 3; P = 0.009], but increased following PPPD [reaching 39.6 s (36.4–43.2) on post-operative day 1; P<0.001; Figure 1].

At baseline, AT levels were slightly lower in patients undergoing hepatectomy [85.0% (78.0–92.0), P<0.001)] and patients undergoing PPPD [83.0% (71.3–85.8), P<0.001] compared to the controls [109.0% (99.3–117.0)]. Levels of AT substantially and significantly decreased at the end of surgery after both the hepatectomy [52.0% (40.0– 63.5), P<0.001 compared to baseline value] and PPPD [61.0% (35.3–68.5), P<0.001 compared to baseline value]. AT levels further decreased in the patients undergoing hepatectomy until post-operative day 3, after which levels slowly increased towards baseline levels which were reached on post-operative day 30. In the patients undergoing PPPD, levels of AT progressively increased after post-operative day 1, reaching levels exceeding baseline levels on post-operative day 30 (P=0.002; Figure 1). Baseline levels of protein C were

PRESERVED THROMBIN GENERATION AFTER LIVER RESECTION

______44

Figure 1: Changes in laboratory measurements (A. prothrombin time (PT), B. activated partial thromboplastin time (APTT), C. antithrombin (AT), D. Protein C (PC), E. factor II (FII), and F. factor VIII (FVIII)) after hepatectomy and PPPD. *P<0.05 between the hepatectomy and PPPD group. ºP<0.05 vs. baseline levels in the hepatectomy group. §P<0.05 vs. baseline levels in the PPPD group. ●P<0.05 in the hepatectomy group vs. controls. ∞P<0.05 in the PPPD group vs. controls. End-OK: end of surgery, POD: postoperative day. comparable between patients undergoing hepatectomy, patients undergoing PPPD, and controls [96.0% (82.0–111.0), 94.5% (80.0–118.8) and 104.0% (93.0–112.3), respectively]. Protein C levels decreased until post-operative day 1 in both the patients undergoing hepatectomy [44.0% (31.0–51.5), P<0.001 compared to baseline value] and the patients

PRESERVED THROMBIN GENERATION AFTER LIVER RESECTION

______45

undergoing PPPD [53.0% (45.0–75.5), P<0.001 compared to baseline value]. Protein C levels decreased significantly after hepatectomy compared to PPPD (P=0.03), and the time to reach baseline levels was substantially longer after hepatectomy (levels returned to baseline on post-operative day 30 after hepatectomy, compared to post-operative day 3 after PPPD) (Figure. 1).

Levels of FII at baseline were comparable between the groups, with 99.0% (91.0–110.0) of FII in the controls, 102.0% (86.8–107.4) in the patients undergoing hepatectomy, and 93.5% (83.5–105.5) in the patients undergoing pancreatic surgery. FII levels decreased following surgery to 55.8% (50.9–67.0) (P<0.001 compared to baseline value) in patients undergoing hepatectomy and 55.0% (47.5–71.0) (P<0.001 compared to baseline value) in patients undergoing PPPD on post-operative day 1.

Baseline levels of FII were reached again on post-operative day 4 after PPPD, but not until post-operative day 30 after hepatectomy (Figure 1). FVIII levels were, at baseline, significantly higher in both the patients undergoing hepatectomy [135.6% (109.0–165.8), P<0.05] and the patients undergoing PPPD [162.5% (112.5–203.5), P<0.01] compared to the controls [92.5% (85.5–114.8)]. Levels of FVIII increased in both groups following surgery, with the highest levels of FVIII on post-operative day 5 after hepatectomy [251.1% (227.8–264.4), P<0.001 compared to baseline value], and on post-operative day 7 after PPPD [250.0% (234.0–261.0), P=0.03 compared to baseline value]. After post-operative day 7 the levels of FVIII substantially decreased in both groups although levels remained elevated compared to the healthy controls (Figure. 1).

THROMBIN GENERATION At the start of surgery, the ETP measured in the absence of TM, was comparable between the patients undergoing hepatectomy, the patients undergoing PPPD and healthy volunteers (Table 2). In the patients undergoing hepatectomy, the ETP decreased until Post-operative day 7 (P<0.001 compared to baseline value), and recovered until baseline values on day 30. In the patients undergoing PPPD, the decrease in ETP was more pronounced as compared to the decrease in the patients undergoing hepatectomy (P=0.01 compared to baseline value, and P=0.045 compared to hepatectomy on post-operative day 3).

The ETP in the patients undergoing PPPD also recovered to baseline values on post-operative day 30 (Figure 2). Despite the decrease in thrombin generation in the absence of TM, in the presence of TM thrombin generation increased following hepatectomy (Table 2; Figure 2). The ETP increased from 542 nM*min (417–694) at baseline (P>0.05 compared to controls) to 845 nM*min (789–1050) on post-operative day 3 (P<0.001 compared to baseline; P<0.001 compared to controls).

PRESERVED THROMBIN GENERATION AFTER LIVER RESECTION

______46

Table 2: Parameters derived from the thrombin generation test in patients following hepatectomy, patients following PPPD, and healthy controls. In absence of thrombomodulin In presence of thrombomodulin ETP Peak Lag-time Velocity index ETP Peak Lag-time Velocity index Controls 1009 [856-1154] 233 [191-245] 1.9 [1.7-2.0] 94 [58-105] 558 [390-680] 142 [99-176] 1.7 [1.7-2.0] 630 [47-87]

Patients following hepatectomy Baseline 1141 [960-1364] 201 [183-201] 1.8 [1.7-2.2] 68 [51-95] 542 [417-694] 128 [97-156] 1.7 [1.7-2.0] 57 [41-78] End-OK 987 [915-1223]+ 200 [170-220] 1.7 [1.3-1.8] 85 [75-97]+ 760 [593-878]*+ 161 [124-164]+ 1.6 [1.3-1.7] 77 [61-87]+ POD1 993 [781-1179]+ 194 [146-212] 2.0 [1.7-2.0] 85 [71-98] 820 [628-956]+ 167 [135-185] 2.0 [1.7-2.0] 79 [64-93] POD3 981 [931-1150]+ 203 [186 -227] 2.0 [1.7-2.0] 99 [88-106]+ 845 [789-1050]*+ 189 [177-203]*+ 2.0 [1.7-2.0] 94 [88-102]*+ POD5 907 [823-1073]+ 203 [180-218] 1.7 [1.7-2.0] 95 [88-103]+ 787 [721-896]*+ 181 [174-208]+ 1.7 [1.7-2.0] 93 [79-104]*+ POD7 883 [745-994]+ 199 [176-215] 1.7 [1.7-2.0] 89 [82-106]+ 715 [654-847] 186 [162-198]+ 1.7 [1.7-2.0] 93 [74-103]*+ POD30 1173 1080-1173] 243 [194-293] 1.8 [1.7-2.0] 103 [58-126]+ 818 [501-1101] 179 [116-248]+ 1.7 [1.7-2.0] 77 [50-115]+

Patients following PPPD Baseline 1036 [810-1189] 165 [118-205]* 1.8 [1.7-2.3] 49 [33-87]* 361 [243-680] 78 [55-168]* 1.7 [1.7-1.9] 36 [25-80]* End-OK 980 [686-1112]+ 152 [116-193]* 1.8 [1.5-2.2] 64 [36-80]* 559 [243-823] 110 [63-127]* 1.7 [1.7-1.9] 53 [32-62] POD1 772 [656-874]*+ 128 [96-162]*+ 2.3 [1.6-3.3] 47 [26-72]* 388 [178-691] 91 [35-140]* 2.0 [1.6-2.7] 42 [11-63]* POD3 868 [658-962]*+ 194 [146-212]* 1.7 [1.7-2.1] 79 [65-99]+ 627 [469-754] 154 [120-187]+ 1.8 [1.7-2.1] 70 [59-94]+ POD5 825 [524-882]*+ 182 [122-207]* 2.0 [1.9-2.2] 67 [49-104] 399 [347-701] 119 [98-185] 2.0 [1.7-2.3] 59 [47-83] POD7 733 [549-966]*+ 162 [127-212]* 2.2 [1.9-2.5] 69 [45-84]* 317 [259-601] 89 [74-165] 2.0 [1.9-2.5] 49 [32-83] POD30 979 [901-1121] 185 [149-206]* 2.2 [1.9-2.3] 70 [43-78]* 476 [331-602] 112 [74-136]* 2.0 [1.7-2.1] 48 [32-66]*

Data are expressed as median [interquartile range]. TM: thrombomodulin. ETP: endogenous thrombin potential. End-OK: end of surgery. POD: postoperative day. PPPD: pylorus preserving pancreaticoduodenectomy. *P<0.05 compared to controls; +P<0.05 compared to baseline.

PRESERVED THROMBIN GENERATION AFTER LIVER RESECTION

______47

Figure 2: Changes in the thrombin generation assay [(A) endogenous thrombin potential (ETP) without thrombomodulin (TM), (B) endogenous thrombin potential (ETP) with TM, (C) thrombomodulin sensitivity ratio (TMSR)] after hepatectomy and PPPD. *P < 0.05 between the hepatectomy and PPPD group. °P < 0.05 vs. baseline levels in the hepatectomy group. §P < 0.05 vs. baseline levels in the PPPD group. ●P < 0.05 in the hepatectomy group vs. controls. ∞ P < 0.05 in the PPPD group vs. controls. End-OK, end of surgery; POD, post-operative day.

In comparison, total thrombin generation only slightly increased in the patients following PPPD (Figure 2). When these data were recalculated to a normalised thrombomodulin sensitivity ratio (TM-SR), it became evident that after hepatectomy thrombomodulin was less effective at regulating thrombin generation in these patients compared to controls (Figure 2). The TM-SR increased from 1.2 ± 0.4 (mean ± s.d.) at baseline to 2.1 ± 0.2 on post- operative day 3 in the patients undergoing hepatectomy (P < 0.001 compared to baseline; P < 0.001 compared to controls). In comparison, the TM-SR only slightly increased in the patients following PPPD.

Table 2 shows the other parameters derived from the thrombin generation test. In the absence of TM, also peak thrombin generation levels remained comparable to baseline values following hepatectomy, and slightly decreased in the patients following PPPD on post-operative day 1. Following surgery, the lag-time remained comparable to baseline values for both patient groups. The velocity index increased until post-operative day 30 following hepatectomy, but only slightly increased on post-operative day 3 and declined again thereafter following PPPD. In the presence of TM, peak thrombin generation also gradually increased following hepatectomy, compared to only an increase on post-operative day 3 following PPPD. Following surgery, the lag-time remained the same in both patient groups. The velocity index, however, increased in both groups following surgery, with a maximum on post-operative day 3.

PRESERVED THROMBIN GENERATION AFTER LIVER RESECTION

______48

DISCUSSION

In the present study, we observed a hypercoagulable status in samples taken from patients following a right hemi-hepatectomy, despite substantially prolonged conventional coagulation tests. Our results using thrombomodulin-modified thrombin generation testing are in line with several recent studies that found a normal to hypercoagulable state following liver surgery.5,6,10,19 In contrast to previous work, our study provides a mechanistic explanation for the hypercoagulable state, i.e. a profound and sustained post-operative deficiency of the natural anti-coagulants AT and protein C.

Samples taken after liver resection were profoundly resistant to the anti-coagulant action of thrombomoduli, the physiological activator of the natural anti-coagulant protein C. While thrombin generation substantially decreased in healthy volunteers when thrombomodulin was added to the plasma, thrombin generation only slightly decreased by the addition of thrombomodulin in plasma from patients following hemi-hepatectomy. Consequently, thrombin generation in the presence of thrombomodulin was higher than that of healthy volunteers. The thrombomodulin resistance following liver surgery is in part explained by the decreased levels of protein C and elevated levels of FVIII.28 Combined with the low levels of AT, the net effect is a normal or even supranormal thrombin generation when tested in the presence of thrombomodulin after hemi-hepatectomy, despite decreased levels of procoagulant proteins.

The decrease in coagulation factors following hemi-hepatectomy is in part explained by the decreased synthetic capacity of the liver remnant. However, we also observed a transient decrease in FII, AT, and protein C following pancreatic surgery. Therefore, it is likely that consumption of coagulation proteins as a result of surgical damage may also play a role. Finally, haemodilution may also partly explain the decrease in coagulation factor levels, and the decrease in haemoglobin following surgery combined with the limited blood loss indeed suggests our patients to be slightly haemodiluted.

Despite similar effects of both liver and pancreas surgery on levels of coagulation factors, we observed a difference in thrombin generation following hemi-hepatectomy and PPPD. Although thrombin generation tested in the presence of thrombomodulin was increased following liver surgery, thrombin generation under these experimental conditions was normal in patients following pancreatic surgery. Furthermore, when thrombin generation was tested in the absence of thrombomodulin, we observed a decreased in thrombin generation in both groups, but this decrease was much more pronounced following pancreatic surgery. De Pietri et al.6 have also shown signs of hypocoagulability in patients undergoing pancreatic surgery using TEG. As the decrease in protein C and AT were more

PRESERVED THROMBIN GENERATION AFTER LIVER RESECTION

______49

extensive following hemi-hepatectomy compared to the decrease in patients following PPPD, this may in part explain the difference in thrombin generation between both groups.

As reported before,5,6,19,28 we also observed a substantial prolongation of conventional coagulation tests in patients following liver surgery. However, these tests cannot reflect the true haemostatic status of patients undergoing liver surgery, as they are only sensitive for levels of procoagulant factors and do not take the reduction in anticoagulant factors, which also occur following liver surgery, into account. The thrombomodulin-modified thrombin generation test is, however, sensitive to all anticoagulant proteins in the plasma. Therefore, this test measures the true balance between the pro- and anti-coagulant factors. Thrombin generation testing in the presence of thrombomodulin has demonstrated normal or even superior thrombin generation in patients with chronic liver disease, despite prolonged conventional coagulation tests.28,29,31 These findings have challenged the long held dogma of chronic liver disease being associated with a bleeding tendency due to changes in the plasmatic coagulation system.33 Our present findings show comparable findings in patients that underwent liver resections, which may have important clinical consequences. First, we strongly advise not to correct routine coagulation tests by using blood product transfusion during or after hepatectomy. In fact, several serious side effects of blood product transfusion may occur, including the risk of infection and the risk of transfusion-related acute lung injury.17 Furthermore, routine thrombosis prophylaxis should not be withheld in patients undergoing liver resection, as this may result in higher rates of venous thromboembolism in post-operative liver disease patients. More extensive prophylaxis may even be required as VTE rates are still substantial in those patients receiving optimal routine thromboprophylaxis.33

In conclusion, although conventional coagulation tests point to hypocoagulability in patients undergoing liver resection, thrombin generation in the presence of thrombomodulin revealed hypercoagulability following liver resection. This hypercoagulability was associated to a profound thrombomodulin resistance which was likely attributable to decreased levels of protein C and elevated levels of FVIII. Therefore, clinicians should be aware of the limitations of the use of conventional coagulation tests to guide haemostatic management during liver surgery. Furthermore, the results of our study support the exploration of more extensive use of anti-coagulant medication in the post-operative period, as was previously suggested by others.7,18,33

PRESERVED THROMBIN GENERATION AFTER LIVER RESECTION

______50

REFERENCES

1. Mahla E, Lang T, Vicenzi MN, et al. Thromboelastography for monitoring prolonged hypercoagulability after major abdominal surgery. Anesth Analg 2001;92:572-7. 2. Gibbs NM, Crawford GP, Michalopoulos N. Postoperative changes in coagulant and anticoagulant factors following abdominal aortic surgery. J Cardiothorac Vasc Anesth 1992;6:680-5. 3. Weinberg L, Scurrah N, Gunning K, McNicol L. Postoperative changes in prothrombin time following hepatic resection: implications for perioperative analgesia. Anaesth Intensive Care 2006;34:438-43. 4. Siniscalchi A, Begliomini B, De Pietri L, et al. Increased prothrombin time and platelet counts in living donor right hepatectomy: implications for epidural anesthesia. Liver Transpl 2004;10:1144-9. 5. Barton JS, Riha GM, Differding JA, et al. Coagulopathy after a liver resection: is it over diagnosed and over treated? HPB (Oxford) 2013;15:865-71. 6. De Pietri L, Montalti R, Begliomini B, et al. Thromboelastographic changes in liver and pancreatic cancer surgery: hypercoagulability, hypocoagulability or normocoagulability? Eur J Anaesthesiol 2010;27:608-16. 7. Tzeng CW, Katz MH, Fleming JB, et al. Risk of venous thromboembolism outweighs post- hepatectomy bleeding complications: analysis of 5651 National Surgical Quality Improvement Program patients. HPB (Oxford) 2012;14:506-13. 8. Turley RS, Reddy SK, Shortell CK, Clary BM, Scarborough JE. Venous thromboembolism after hepatic resection: analysis of 5,706 patients. J Gastrointest Surg 2012;16:1705-14. 9. Bezeaud A, Denninger MH, Dondero F, et al. Hypercoagulability after partial liver resection. Thromb Haemost 2007;98:1252-6. 10. Cerutti E, Stratta C, Romagnoli R, et al. Thromboelastogram monitoring in the perioperative period of hepatectomy for adult living liver donation. Liver Transpl 2004;10:289-94. 11. Dondero F, Taille C, Mal H, et al. Respiratory complications: a major concern after right hepatectomy in living liver donors. Transplantation 2006;81:181-6. 12. Lambing A, Kuriakose P, Abouljoud MS. Hypercoagulability risks among adult living liver donors. Transplant Proc 2006;38:3579-81. 13. Shimada M, Matsumata T, Kamakura T, Hayashi H, Urata K, Sugimachi K. Modulation of coagulation and fibrinolysis in hepatic resection: a randomized prospective control study using antithrombin III concentrates. Thromb Res 1994;74:105-14. 14. Martin RC,2nd, Jarnagin WR, Fong Y, Biernacki P, Blumgart LH, DeMatteo RP. The use of fresh frozen plasma after major hepatic resection for colorectal metastasis: is there a standard for transfusion? J Am Coll Surg 2003;196:402-9. 15. Stellingwerff M, Brandsma A, Lisman T, Porte RJ. Prohemostatic interventions in liver surgery. Semin Thromb Hemost 2012;38:244-9. 16. Yamazaki S, Takayama T, Kimura Y, et al. Transfusion criteria for fresh frozen plasma in liver resection: a 3 + 3 cohort expansion study. Arch Surg 2011;146:1293-9. 17. Alter HJ, Klein HG. The hazards of blood transfusion in historical perspective. Blood 2008;112:2617-26.

PRESERVED THROMBIN GENERATION AFTER LIVER RESECTION

______51

18. Reddy SK, Turley RS, Barbas AS, et al. Post-operative pharmacologic thromboprophylaxis after major hepatectomy: does peripheral venous thromboembolism prevention outweigh bleeding risks? J Gastrointest Surg 2011;15:1602-10. 19. Louis SG, Barton JS, Riha GM, et al. The international normalized ratio overestimates coagulopathy in patients after major hepatectomy. Am J Surg 2014;207:723,7; discussion 727. 20. Johansson PI, Stensballe J, Oliveri R, Wade CE, Ostrowski SR, Holcomb JB. How I treat patients with massive hemorrhage. Blood 2014;124:3052-8. 21. Stravitz RT. Potential applications of thromboelastography in patients with acute and chronic liver disease. Gastroenterol Hepatol 2012;8:513-20. 22. Afshari A, Wikkelso A, Brok J, Moller AM, Wetterslev J. Thrombelastography (TEG) or thromboelastometry (ROTEM) to monitor haemotherapy versus usual care in patients with massive transfusion. Cochrane Database Syst Rev 2011;(3):CD007871. 23. Gurusamy KS, Pissanou T, Pikhart H, Vaughan J, Burroughs AK, Davidson BR. Methods to decrease blood loss and transfusion requirements for liver transplantation. Cochrane Database Syst Rev 2011;(12):CD009052. 24. Zambruni A, Thalheimer U, Leandro G, Perry D, Burroughs AK. Thromboelastography with citrated blood: comparability with native blood, stability of citrate storage and effect of repeated sampling. Blood Coagul Fibrinolysis 2004;15:103-7. 25. Mallett SV, Chowdary P, Burroughs AK. Clinical utility of viscoelastic tests of coagulation in patients with liver disease. Liver Int 2013;33:961-74. 26. McMichael MA, Smith SA. Viscoelastic coagulation testing: technology, applications, and limitations. Vet Clin Pathol 2011;40:140-53. 27. Karon BS. Why is everyone so excited about thromboelastrography (TEG)? Clin Chim Acta 2014;436:143-8. 28. Lisman T, Bakhtiari K, Pereboom IT, Hendriks HG, Meijers JC, Porte RJ. Normal to increased thrombin generation in patients undergoing liver transplantation despite prolonged conventional coagulation tests. J Hepatol 2010;52:355-61. 29. Tripodi A, Primignani M, Lemma L, et al. Detection of the imbalance of procoagulant versus anticoagulant factors in cirrhosis by a simple laboratory method. Hepatology 2010;52:249-55. 30. Tripodi A, Salerno F, Chantarangkul V, et al. Evidence of normal thrombin generation in cirrhosis despite abnormal conventional coagulation tests. Hepatology 2005;41:553-8. 31. Gatt A, Riddell A, Calvaruso V, Tuddenham EG, Makris M, Burroughs AK. Enhanced thrombin generation in patients with cirrhosis-induced coagulopathy. J Thromb Haemost 2010;8:1994- 2000. 32. Hemker HC, Giesen P, Al Dieri R, et al. Calibrated automated thrombin generation measurement in clotting plasma. Pathophysiol Haemost Thromb 2003;33:4-15. 33. Ejaz A, Spolverato G, Kim Y, et al. Defining incidence and risk factors of venous thromboemolism after hepatectomy. J Gastrointest Surg 2014;18:1116-24.

CHAPTER 4

IMMEDIATE POSTOPERATIVE LOW PLATELET COUNT IS ASSOCIATED WITH DELAYED LIVER FUNCTION RECOVERY AFTER PARTIAL LIVER RESECTION

EDRIS M. ALKOZAI MAARTEN W. NIJSTEN KOERT P. DE JONG MARIEKE T. DE BOER PAUL M. J. G. PEETERS MAARTEN J. SLOOFF ROBERT J. PORTE TON LISMAN

PUBLISHED IN: ANN SURG 2010;251: 300–306

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______54

ABSTRACT

OBJECTIVE: To evaluate whether a low postoperative platelet count is associated with a poor recovery of liver function in patients after partial liver resection.

Background: Experimental studies in rodents have recently suggested that blood platelets play a critical role in the initiation of liver regeneration. It remains unclear whether platelets are also involved in liver regeneration in humans.

METHODS: In a series of 216 consecutive patients who underwent partial liver resection for colorectal liver metastases, we studied postoperative mortality and liver dysfunction in relation to the immediate postoperative platelet count. All patients had normal preoperative liver function and none of them had liver fibrosis or cirrhosis. Delayed postoperative recovery of liver function was defined as serum bilirubin > 50 μmol/L or prothrombin time > 20 seconds at any time point between postoperative day 1 and 5.

RESULTS: Patients with a low (<100 x 109/L) immediate postoperative platelet count had worse postoperative liver function, higher serum markers of liver injury, and increased mortality compared with patients with normal platelet counts (≥100x109/L). A low immediate postoperative platelet count was identified as an independent risk factor of delayed postoperative recovery of liver function(OR, 11.5; 95% CI, 1.1–122.4; P = 0.04 in multivariate analysis).

CONCLUSION: After partial liver resection, a low platelet count is an independent predictor of delayed postoperative liver function recovery and is associated with increased risk of postoperative mortality. These clinical findings are in accordance with the accumulating evidence from experimental studies, indicating that platelets play a critical role in liver regeneration.

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______55

INTRODUCTION

Partial liver resection has become the treatment of choice for patients with colorectal liver metastasis.1-4 Although resection related mortality and morbidity has decreased substantially in recent years, postoperative mortality rate may still be as high as 1% to 5%.5-10 Morbidity and mortality are, among other factors, strongly related to postoperative liver insufficiency, which may be a consequence of failure of liver regeneration due to underlying liver disease or insufficient volume of residual functional hepatic reserve.11

Liver regeneration requires an orchestrated interplay of cytokines and growth factors, resulting in a time-dependent replication of different types of liver cells.12 Experimental studies suggest that blood platelets play a pivotal role in liver regeneration after partial liver resection.13-16 Depletion of platelets severely suppresses liver regeneration, whereas induction of thrombocytosis by administration of thrombopoietin or by splenectomy has been shown to accelerate liver regeneration.13-15 There is evidence that platelet-derived serotonin plays an essential role in platelet-mediated liver regeneration.14

Although the role of platelets and platelet-derived serotonin on hepatocyte proliferation has been established in vitro and in murine models, it is not known whether similar mechanisms apply in humans. Previous studies have identified an association between preoperative platelet count and outcome after liver resection, but these studies were performed in heterogeneous patient populations, including a considerable proportion of patients with chronic liver disease and subsequent thrombocytopenia due to portal hypertension and hypersplenism.8,17-19 Moreover, these studies did not consider the number of platelets immediately after surgery, when liver regeneration is initiated. These studies, therefore, do not allow an unbiased assessment of the possible relationship between platelets and liver regeneration in humans.

We here report a clinical study in which we examined the relationship between immediate postoperative platelet count and outcome after partial liver resection in patients without preexisting liver disease, specifically in patients with colorectal liver metastases. We hypothesized that patients with a low immediate postoperative platelet count, ie, at the moment when liver regeneration starts, would have a less effective liver regeneration as compared with those with a normal platelet count. We evaluated whether an immediate postoperative low platelet count was associated with poor recovery of liver function and higher risk of mortality. Primary outcome parameters were 90-day postoperative mortality, postoperative liver dysfunction, and postoperative serum markers of liver injury. In addition, we performed a uni- and multivariate analysis to identify clinical variables that are associated with delayed postoperative recovery of liver function.

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______56

PATIENTS AND METHODS

STUDY POPULATION A total of 533 consecutive liver resections were performed at the Department of Surgery of university Medical Center Groningen between January 1995 and September 2007. Only patients who underwent liver resection for colorectal liver metastasis and who did not have a preexisting liver disease (n=232) were selected for the current study. Sixteen patients were subsequently excluded from the analysis because platelet count at the day of operation was not documented. This resulted in a total of 216 patients included in the study. The baseline characteristics of the patients and variables related to the perioperative management and surgical procedure were obtained from a prospectively collected database. When necessary, the computer-stored hospital files were reviewed for other relevant clinical parameters and missing laboratory data. Patient characteristics and surgical variables for the entire series are presented in Table 1. The percentage of resected functional liver volume was calculated from published estimates of the proportion of liver volume of each individual segment.20,21 Specifically, we have used the following subdivision: segment 1 represents 2% of total liver volume, segment 2: 8%, segment 3: 8%, segment 4: 17%, segment 5: 17.5%, segment 6: 15%, segment 7: 15%, segment 8: 17.5%. National legislation and the ethical committee of our institution approve this type of retrospective studies.

LABORATORY VARIABLES To study the possible role of platelets in liver regeneration, we identified the immediate postoperative platelet count in each individual patient. Platelet count was always obtained at the day of the surgery, usually upon arrival at the intensive care unit (ICU) after surgery. On the basis of this platelet count, patients were divided into 2 groups: patients with a low platelet count (<100 x 109/L) and patients with a normal platelet count (≥100 x 109/L). In addition, the following laboratory variables were included in the analyses: serum levels of total bilirubin, creatinine, albumin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), hemoglobin, and gamma-glutamyltransferase (GGT) as well as prothrombin time (PT) and plasma levels of antithrombin (AT). Laboratory data were obtained on day 0 (immediately after surgery), between postoperative day 1 and 3, between day 4 and 6, between day 7 and 10, between day 11 and 20, and between day 21 and 30.

OUTCOME PARAMETERS The following outcome parameters were considered in this study: mortality within 90 days after surgery and delayed postoperative recovery of liver function. Delayed recovery of liver function was used as a surrogate marker for poor liver regeneration. The definition of delayed postoperative recovery of liver function was based on a modification of the criteria suggested by Balzan et al.10 and included serum bilirubin > 50 μmol/L or PT > 20 seconds at any time

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______57

Table 1: Comparison of patient characteristics and surgical variables in patients with low (<100 x 109/L) or with high (≥100 x109/L) platelet count after partial liver resection.

Total Low Platelet Count High Platelet Variables (n=216) (n=21) count (n=195) P Patient variables Age (yr), median (IQR) 67 (59–74) 71 (67–75) 66 (59–74) 0.03 Sex, man 125 (58%) 13 (62%) 112 (58%) 0.82 Location of primary tumor 0.35 Colon 135 (63%) 11 (50%) 124 (64%) Rectum 81 (37%) 10 (48%) 71 (36%) Timing of metastasis 0.62 Synchronous 65 (30%) 5 (24%) 60 (31%) Metachronous 151 (70%) 16 (76%) 16 (76%) Comorbidity Diabetes mellitus, yes 14 (7%) 0 14 (7%) 0.37 Cardiovascular disease, yes 18 (8%) 3 (14%) 15 (8%) 0.39 Hypertension, yes 30 (14%) 0 30 (15%) 0.05 COPD, yes 8 (4%) 1 (5%) 7 (4%) 0.67 Preoperative chemotherapy, yes 77 (36%) 8 (38%) 69 (35%) 0.81 Previous liver surgery, yes 9 (4%) 0 9 (5%) 0.61 Surgical variables Liver volume removed 213/216* 0.67 <35% 89 (42%) 7 (33%) 82 (43%) 35%-65% 84 (39%) 10 (48%) 74 (38%) >65% 40 (19%) 4 (19%) 36 (19%) Blood loss 197/216* <0.01 >1000 mL 109 (55%) 5 (23%) 104 (59%) 1000-5000 mL 81 (41%) 11 (52%) 70 (40%) >5000 mL 7 (4%) 5 (23%) 2 (1%) RBC transfusion, yes 215/216* 77 (36%) 19 (91%) <0.01 Length of stay in ICU 0.02 ≤2 days 148 (68%) 10 (48%) 138 (71%) 3-5 days 43 (20%) 5 (24%) 38 (19%) >5 days 25 (12%) 6 (28%) 19 (10%) Length of hospital stay 0.07 ≤15 days 116 (54%) 7 (33%) 109 (56%) >15 days 100 (46%) 14 (67%) 86 (44%) Preoperative laboratory values AST (U/L), median (IQR) 189/216* 28 (24–40) 29 (23–37) 0.98 ALT (U/L), median (IQR) 190/216* 23 (15–37) 24 (17–31) 0.62 GGT (U/L), median (IQR) 171/216* 38 (24–62) 44 (30–87) 0.24 Hb (mmol/L), median (IQR) 200/216* 8.6 (7.9–9.2) 8.6 (8.0–9.2) 0.69 Albumin (g/L), median (IQR) 170/216* 44 (39–46) 43 (40–45) 0.90 TB (μmol/L), median (IQR) 187/216* 12 (10–16) 11 (8–14) 0.12 AT (%), median (IQR) 154/216* 86 (72–102) 100 (88–112) <0.01 PT (s), median (IQR) 164/216* 13.2 (12.6–14.2) 12.6 (11.6–13.2) 0.01 Creatinine (μmol/L), median (IQR) 216/216 86 (73–96) 82 (73–95) 0.59 *Some variables were not available for all patients. Indicated are the numbers of patients for whom values were available.ALT indicates alanine aminotransferase; AT, antithrombin; AST, aspartate aminotransferase; Hb, hemoglobin; ICU, intensive care unit; IQR, interquartile range; GGT, gamma-glutamyltransferase; PT, prothrombin time; RBC, red blood cells; TB, total bilirubin point between postoperative day 1 and postoperative day 5. Hemolytic or obstructive mechanisms for high bilirubin levels were excluded. Based on these criteria, 123 patients were categorized as having adequate postoperative liver function recovery, and 93 were categorized as having delayed postoperative liver function recovery. Clinical and laboratory variables of the 2 groups were compared.

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______58

STATISTICAL ANALYSIS Statistical analyses were performed using the statistical software package SPSS 14.0 (SPSS Inc, Chicago, IL). Categorical variables were presented as numbers and percentages, and groups were compared using the Pearson χ2 test or Fisher exact test. Continuous variables were expressed as median and interquartile range (IQR), and groups were compared using the Mann-Whitney U test. Uni- and multivariate logistic regression analysis was used to identify independent risk factors for delayed postoperative liver function recovery and odds ratios (OR) with the corresponding 95% confidence intervals (95% CI) were calculated. The following variables were included in these analyses: age, sex, location of primary carcinoma, timing of metastasis, comorbidities, chemotherapy, first or second liver resection, volume of segments resected, intraoperative blood loss, perioperative red blood cell (RBC) transfusion, length of stay in the intensive care unit and hospital, and preoperative laboratory parameters. Preoperative laboratory parameters were assessed within 5 days prior to surgery. All variables that reached a P ≤ 0.1 in the univariate analysis were included in the multivariate logistic regression analysis. P < 0.05 was considered statistically significant.

RESULTS

COMPARISON OF PATIENTS WITH LOW AND NORMAL POSTOPERATIVE PLATELET COUNT Of the total of 216 patients included in this study, 21 patients had a low platelet count (<100 x109/L) immediately after surgery, while 195 patients had a normal platelet count (≥100 x109/L). A comparison of patient demographics and clinical variables in these 2 groups is presented in Table 1. Patients with a low postoperative platelet count were slightly older, had more perioperative blood loss, received more RBC transfusions, and had a longer postoperative stay in the ICU. Other important variables, such as the number of resected liver segments and preoperative laboratory values, were similar between the 2 groups, except for a slightly decreased AT and slightly longer PT in the group with a low platelet count.

IS LOW POSTOPERATIVE PLATELET COUNT ASSOCIATED WITH INCREASED MORTALITY? Overall mortality within 90 days after surgery for the entire series of patients was 4.7%. Mortality rate within 90 days after surgery was almost 4 times higher in patients with a low postoperative platelet count, compared with patients with a normal platelet count (OR, 3.9; 95% CI, 0.95–15.99, P = 0.06). A formal multivariate analysis of this outcome parameter could not be performed due the low number of patients who died postoperatively.

IS LOW POSTOPERATIVE PLATELET COUNT ASSOCIATED WITH BIOCHEMICAL EVIDENCE OF INCREASED LIVER INJURY AND DYSFUNCTION? Perioperative evolution of biochemical markers of liver cell injury and dysfunction are presented in Figure 1. Postoperatively, peak levels of AST and ALT were significantly higher

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______59

700 700

600 600 ** )

500 * ) 500

L

L

/

/ U

400 U 400

(

(

T

T 300 300 *

L

S A A 200 * 200 100 100 0 0 Pre 0 1-3 4-6 7-10 11-20 21-30 Pre1 02 1-33 4-64 7-105 11-206 21-307 Time interval (days) Time interval (days)

60 350

) **

L 300 / 50

l ** **

) o

* L 250 /

m 40

U 

( 200

(

30

n * T

i 150 b

20 G u

G 100

r *

i * l

i 10 50 B 0 0 Pre1 02 1-33 4-64 7-105 11-206 21-307 Pre1 02 1-33 4-64 7-105 11-206 21-307 Time interval (days) Time interval (days) 26 125

) *

24 %

( 100

22 ** *

)

n

i c

20 b e * 75 **

s ** * m

( 18

o r

T 50

16 h P

* t i

14 t

n 25

12 A 10 0 Pre1 02 1-33 4-64 7-105 11-206 21-307 Pre1 02 1-33 4-64 7-105 11-206 21-307 Time interval (days) Time interval (days)

50 9

) )

L 40 /

L 8

/

l

g

( o 30 * *

n ** ** * m

i 7

m m

20 (

u

b b

l 6 H

A 10

0 5 Pre1 02 1-33 4-64 7-105 11-206 21-307 Pre 0 1-3 4-6 7-10 11-20 21-30 Time interval (days) Time interval (days)

Figure 1: Evolution of laboratory variables in patients with low (gray lines, <100 x 109/L) and normal (black lines, ≥ 100x109/L) immediate postoperative platelet count. Shown are media values. **P<0.01, *P<0.05. in the patients with a low postoperative platelet count, but values of GGT were lower, compared with patients with normal platelet counts. Patients with a low platelet count had also signs of more delayed recovery of liver function, as illustrated by significantly higher levels of serum bilirubin, higher PT values, and lower levels of AT. Notably, the peak in serum bilirubin level in the group with low platelet counts was observed between day 4 and 6, whereas the peak in the group with normal platelet counts occurred between day 1 and 3. Altogether,

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______60

these data suggested that a low platelet count is associated with a more delayed postoperative recovery of liver function.

IS LOW PLATELET COUNT AN INDEPENDENT RISK FACTORS FOR DELAYED POSTOPERATIVE RECOVERY OF LIVER FUNCTION? To examine the possible relationship between platelets and recovery of liver function, we next performed a uni- and multivariate analysis of variables that are potentially associated with poor postoperative recovery of liver function. We have used a definition of poor postoperative recovery of liver function that was based on data previously published by Balzan et al10 and as described above. According to this definition, 93 (43%) patients had delayed postoperative recovery of liver function and 123 (57%) patients had no delayed recovery of liver function. Results of the univariate analysis of potential risk factors for delayed recovery are presented in Table 2. A low platelet count immediately after surgery was associated with an almost 5- fold increased risk of delayed postoperative recovery of liver function (OR, 4.9; 95% CI, 1.7– 13.9; P < 0.01). Other variables associated with delayed postoperative liver function recovery in the univariate analysis were age, RBC transfusion, the liver volume removed, preoperative levels of serum bilirubin, ALT, GGT, AT, and preoperative PT values. Interestingly, a low preoperative platelet count was not associated with delayed recovery of liver function. Mortality within 90 days after surgery was almost 7-fold higher in patients with delayed postoperative liver function recovery, compared with patients without delayed liver function recovery (OR, 6.5, 95% CI, 1.4 –30.8; P = 0.02).

Nine variables with a P 0.10 were entered into the multivariate logistic regression analysis. After multivariate analysis, low platelet count remained as a strong and independent risk factor for delayed postoperative recovery of liver function (OR, 11.5; 95% CI, 1.1–122.4; P=0.04) (Table 3). Other variables that were identified as independent risk factors for delayed postoperative liver function recovery were RBC transfusion, liver volume removed, and preoperative serum bilirubin and GGT (Table 3). As RBC transfusion may directly influence postoperative platelet count by hemodilution, we also performed a multivariate analysis without entering RBC transfusions into the model. In this analysis, the risk of delayed postoperative recovery of liver function associated with low platelet count increased to 21.1 (95% CI, 2.2–199.8). Seven patients received a perioperative platelet transfusion. As the biologic characteristics of transfused platelets may differ from that of endogenous platelets, we repeated our analyses with exclusion of these 7 patients, with no significant effect on the risk estimate (data not shown).

DISCUSSION This study provides clinical evidence that a low postoperative platelet count is associated with an increased risk of mortality and delayed recovery of liver function after partial liver resections. Furthermore, we have demonstrated that immediately after partial liver resection

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______61

Table 2. Univariate analysis of risk factors for delayed postoperative recovery of liver function No. Patients/No. Delayed Recovery of Variables OR (95% CI) P Patients Analyzed Liver Function n = 93 Age (yr), median (IQR) 216/216 Continuous 1.06 (1.0–1.1) <0.01 Sex, man 125/216 59 (47%) 1.5 (0.9–2.6) 0.17 Location of primary tumor 0.20 Rectum 81/216 30 (37%) 1.0, Reference Colon 135/216 63 (47%) 1.5 (0.8–2.6) Timing of metastasis 0.45 Synchronous 65/216 25 (39%) 1.0, Reference Metachronous 151/216 68 (45%) 1.3 (0.7–2.4) Comorbidity Diabetes mellitus, yes 14/216 7 (50%) 1.3 (0.5–4.0) 0.10 Cardiovascular disease, yes 18/216 7 (39%) 0.8 (0.3–2.2) 0.81 Hypertension,, yes 30/216 9 (10%) 0.5 (0.2–1.2) 0.12 COPD, yes 8/216 1 (1%) 0.2 (0.2–1.5) 0.11 Preoperative chemotherapy, yes 77/216 36 (47%) 1.3 (0.7–2.2) 0.47 Previous liver surgery, yes 9/216 2 (22%) 0.4 (0.7–1.8) 0.21 RBC transfusion 77/215 53 (69%) 5.6 (3.0–10.2) <0.01 Liver volume removed <0.01 <35% 89/213 20 (22%) 1.0, Reference 35%-65% 84/213 47 (50%) 4.4 (2.3–8.5) >65% 40/213 26 (28%) 6.4 (2.8–14.5) Blood platelet count on day 0 <0.01 ≥100 x 109/L 195/216 77 (39%) 1.0, Reference <100 x 109/L 21/216 16 (76%) 4.9 (1.7–13.9) Preoperative laboratory variables, (tertiles)* TB (μmol/L), median (IQR) 187/216 <0.01 Low 63 17 (27%) 1.0, reference Intermediate 68 32 (47%) 2.4 (1.2–5.2) High 56 33 (59%) 3.9 (1.8–8.4) PT (s), median (IQR) 168/216 <0.01 Low 58 18 (31%) 1.0, reference Intermediate 59 26 (44%) 1.8 (0.8–3.7) High 51 31 (61%) 3.4 (1.6–7.6) AST (U/L), median (IQR) 189/216 0.13 Low 65 22 (34%) 1.0, reference Intermediate 64 32 (50%) 1.8 (0.8–3.7) High 60 29 (48%) 3.4 (1.6–7.6) ALT (U/L), median (IQR) 190/216 0.01 Low 67 31 (46%) 1.0, reference Intermediate 64 19 (30%) 2.0 (1.0–4.0) High 59 33 (56%) 0.8 (0.9–3.8) GGT (U/L), median (IQR) 171/216 <0.01 Low 61 21 (34%) 1.0, reference Intermediate 53 18 (34%) 0.5 (0.2–1.0) High 56 38 (67%) 1.4 (0.7–3.0) Albumin (g/L), median (IQR) 170/216 0.70 Low 61 27 (44%) 1.0, reference Intermediate 53 24 (45%) 1.0 (0.5–2.2) High 56 22 (39%) 0.8 (0.4–1.7) AT (%) 154/216 0.01 Low 52 34 (65%) 1.0, reference Intermediate 54 24 (44%) 0.4 (0.2–0.9) High 48 17 (35%) 0.5 (0.1–1.7) Platelet count (100 x 109/L) 164/216 0.70 Low 57 24 (42%) 1.0, reference Intermediate 53 18 (34%) 0.7 (0.3–1.5) High 54 22 (41%) 0.9 (0.4–2.0) Creatinine (μmol/L), median (IQR) 216/216 0.46 Low 77 29 (38%) 1.0, reference Intermediate 67 32 (48%) 1.5 (0.8–2.9) High 72 32 (44%) 1.3 (6.9–2.5) Abbreviations as in Table 1.

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______62

Table 3. Multivariate Analysis of Independent Risk Factor for Delayed Postoperative Recovery of Liver Function

Variables OR 95% (CI) P Age 1.00 (0.95–1.06) 0.99 RBC transfusion, (yes vs. no) 6.62 (2.24–19.58) <0.01 Liver volume removed 2.40 (1.21–4.79) 0.01 Platelet count (_100 x109/L) 11.49 (1.08–122.41) 0.04 vs. ≥ 100 x109/L) Preoperative serum bilirubin (μmol/L)* 1.19 (1.06–1.33) <0.01 Preoperative serum GGT (mg/dL)* 1.01 (1.00–1.03) 0.03 Preoperative PT (s)* 1.15 (0.74–1.79) 0.52 Preoperative AT (%)* 0.99 (0.97–1.02) 0.58 Preoperative ALT (U/L)* 1.01 (0.99–1.03) 0.27 *Preoperative laboratory values were entered as continuous variables in the model. Abbreviations as in Table 1. patients with low platelet counts have higher serum markers of liver injury (ALT, AST) compared with patients with normal platelet counts. The combined results suggest that a certain number of platelets are required for optimal liver function recovery, presumably mediated by enhancement of liver regeneration. The results of this clinical study are in accordance with data from several experimental studies in rodents suggesting that platelets play a critical role in the initiation of liver regeneration.14,15 In these experiments, it has been shown that liver regeneration is significantly reduced in mice with severe thrombocytopenia, whereas thrombocytosis is associated with accelerated liver regeneration.15 Platelet-derived serotonin has been identified as a key mediator of liver regeneration.14

Our results are in line with previous clinical studies, suggesting that a low platelet count is associated with poor recovery and worse outcome after liver surgery.8,17-19,22 However, a major drawback of previous clinical studies has been the inclusion of heterogeneous patient populations, including patients with liver cirrhosis, primary liver cancer, and preexisting thrombocytopenia. It is well known that liver surgery in patients with cirrhosis and primary liver cancer has, in general, a poorer outcome than liver surgery for colorectal liver metastases.23- 25 To avoid possible bias from differences in underlying liver disease, we therefore performed our analyses in a homogenous group of patients with colorectal liver metastases, who did not have any underlying liver disease. Second, we have not investigated the impact of preoperative platelet counts, as was done in most previous studies, but we have used immediate postoperative platelet counts. Liver regeneration is known to be initiated very shortly after partial liver resection,26 and if platelets are critically involved in this process, the number of platelets available immediately after surgery will be of greater importance than preoperative platelet counts.

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______63

After our initial observation that patients with a low postoperative platelet count have an increased risk of mortality, we aimed to study the relation between platelet count and liver regeneration. Unfortunately, we were not able to perform a direct quantification of liver regeneration, which would have required measurements of liver volume by sequential imaging studies in the postoperative phase. Such imaging studies are not routinely performed after partial liver resections in our center. As a surrogate marker of liver regeneration, we therefore used laboratory data of serum markers of liver injury and liver function. Recovery of liver function was assumed to be directly related to adequate liver regeneration.

We have assessed perioperative changes in several laboratory parameters and compared these changes in patients with low or normal postoperative platelet counts. Although there was no difference between the 2 groups in the number of segments removed, patients with a low platelet count had significantly reduced postoperative liver function and a slower recovery of liver function after liver resection. Interestingly, we also observed prominent differences in serum markers of liver injury between the 2 groups. Not only the peak levels of serum were higher but also the normalization of serum transaminases was more delayed in the group with low platelet counts, compared with the group with normal platelet count. Although these results may suggest that platelets protect against hepatocellular damage induced by liver resection, the higher serum transaminases in the group with low platelet counts may also reflect an impaired capacity of the remaining liver to clear these enzymes from the circulation, since functional liver mass is decreased substantially in the postoperative period.

In a separate analysis, we have subsequently shown that patients with a low platelet count have a significantly higher risk of delayed postoperative recovery of liver function. In fact, in a multivariate regression analysis, the risk of delayed recovery of liver function was more than 11 times higher in patients with a low postoperative platelet count, compared with patients with normal platelet counts. There is no standard definition of delayed postoperative recovery of liver function. Therefore, we have adopted a modified definition based on the criteria proposed by Balzan et al.10 Delayed recovery of liver function was defined as serum bilirubin > 50 μmol/L and/or PT > 20 seconds at any time during one of the first 5 postoperative days, in the absence of hemolysis or biliary obstruction. According to this definition, a substantial proportion of patients in our study had delayed liver function recovery, and it should be stressed that our definition was not meant to identify patients at a “point of no return” as is achieved with more strict definitions such as the “50–50 criteria” proposed by Balzan et al.10 or the peak bilirubin > 7 mg/dL (or > 119.7 μmol/L) as has been proposed by Mullen et al.9 On the other hand, we did observe a substantially increased risk of mortality in patients that fulfilled our criteria of delayed postoperative liver function recovery.

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______64

A limitation of this study is its retrospective design. It cannot be fully excluded that the patients with a low platelet count were simply in a worse general condition and, therefore, had a more delayed recovery of liver function recovery and increased risk of mortality, than patients with normal platelet counts. However, we did not find any major differences in the preoperative characteristics in the 2 groups and also the liver volume removed was similar in the 2 groups. In addition, low platelet count remained as a strong and independent risk factor for delayed postoperative recovery of liver function in the multivariate regression analysis. The results of this retrospective study require confirmation, but when confirmed, these data may have important clinical implications. If postoperative platelet count is indeed directly related to liver regeneration and recovery of liver function, this would open new avenues to develop novel strategies to stimulate liver regeneration and to avoid liver failure after major liver resections or (partial) liver transplantation. Possible directions could be strategies to increase platelet count, for example, by preoperative administration of thrombopoietin agonists. Based on the finding that platelet-derived serotonin is a key mediator of liver regeneration, an alternative approach could be to use serotonin precursors or serotonin receptor agonists to promote liver regeneration after liver surgery.14 The current findings should not be seen as a stimulus for a more liberal use of platelet concentrates from blood donors, as several studies have shown that platelet transfusion is associated with an increased risk of postoperative morbidity and mortality.27-29 In contrast to endogenous platelets, platelets from blood donors are frequently in an activated state and may induce a range of inflammatory reactions and unwanted side effects.30,31

In conclusion, this retrospective study suggests that a low platelet count is an independent predictor of delayed postoperative liver function recovery and it is associated with increased risk of postoperative mortality after partial liver resection. These clinical findings are in accordance with the accumulating evidence from experimental studies, indicating that platelets play a critical role in liver regeneration.

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______65

REFERENCES

1. Beard SM, Holmes M, Price C, Majeed AW. Hepatic resection for colorectal liver metastases: A cost-effectiveness analysis. Ann Surg 2000;232:763-76. 2. Minagawa M, Makuuchi M, Torzilli G, et al. Extension of the frontiers of surgical indications in the treatment of liver metastases from colorectal cancer: long-term results. Ann Surg 2000;231:487-99. 3. Fong Y, Fortner J, Sun RL, Brennan MF, Blumgart LH. Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: analysis of 1001 consecutive cases. Ann Surg 1999;230:309,18. 4. Choti MA, Sitzmann JV, Tiburi MF, et al. Trends in long-term survival following liver resection for hepatic colorectal metastases. Ann Surg 2002;235:759-66. 5. Wei AC, Tung-Ping Poon R, Fan ST, Wong J. Risk factors for perioperative morbidity and mortality after extended hepatectomy for hepatocellular carcinoma. Br J Surg 2003;90:33-41. 6. Belghiti J, Hiramatsu K, Benoist S, Massault P, Sauvanet A, Farges O. Seven hundred forty- seven hepatectomies in the 1990s: an update to evaluate the actual risk of liver resection. J Am Coll Surg 2000;191:38-46. 7. Gomez D, Malik HZ, Bonney GK, et al. Steatosis predicts postoperative morbidity following hepatic resection for colorectal metastasis. Br J Surg 2007;94:1395-402. 8. Kaneko K, Shirai Y, Wakai T, Yokoyama N, Akazawa K, Hatakeyama K. Low preoperative platelet counts predict a high mortality after partial hepatectomy in patients with hepatocellular carcinoma. World J Gastroenterol 2005;11:5888-92. 9. Mullen JT, Ribero D, Reddy SK, et al. Hepatic insufficiency and mortality in 1,059 noncirrhotic patients undergoing major hepatectomy. J Am Coll Surg 2007;204:854,62; discussion 862-4. 10. Balzan S, Belghiti J, Farges O, et al. The "50-50 criteria" on postoperative day 5: an accurate predictor of liver failure and death after hepatectomy. Ann Surg 2005;242:824,8. 11. Clavien PA, Petrowsky H, DeOliveira ML, Graf R. Strategies for safer liver surgery and partial liver transplantation. N Engl J Med 2007;356:1545-59. 12. Clavien PA, Graf R. Liver regeneration and platelets. Br J Surg 2009;96:965-6. 13. Tomikawa M, Hashizume M, Highashi H, Ohta M, Sugimachi K. The role of the spleen, platelets, and plasma hepatocyte growth factor activity on hepatic regeneration in rats. J Am Coll Surg 1996;182:12-6. 14. Lesurtel M, Graf R, Aleil B, et al. Platelet-derived serotonin mediates liver regeneration. Science 2006;312:104-7. 15. Murata S, Ohkohchi N, Matsuo R, Ikeda O, Myronovych A, Hoshi R. Platelets promote liver regeneration in early period after hepatectomy in mice. World J Surg 2007;31:808-16. 16. Murata S, Matsuo R, Ikeda O, et al. Platelets promote liver regeneration under conditions of Kupffer cell depletion after hepatectomy in mice. World J Surg 2008;32:1088-96. 17. Jarnagin WR, Gonen M, Fong Y, et al. Improvement in perioperative outcome after hepatic resection: analysis of 1,803 consecutive cases over the past decade. Ann Surg 2002;236:397,406; discussion 406-7. 18. Poon RT, Fan ST, Lo CM, et al. Improving perioperative outcome expands the role of hepatectomy in management of benign and malignant hepatobiliary diseases: analysis of 1222 consecutive patients from a prospective database. Ann Surg 2004;240:698,708.

P L A T E L E T S A N D O U T C O M E O F L I V E R R E S E C T I O N

______66

19. Taketomi A, Kitagawa D, Itoh S, et al. Trends in morbidity and mortality after hepatic resection for hepatocellular carcinoma: an institute's experience with 625 patients. J Am Coll Surg 2007;204:580-7. 20. Leelaudomlipi S, Sugawara Y, Kaneko J, Matsui Y, Ohkubo T, Makuuchi M. Volumetric analysis of liver segments in 155 living donors. Liver Transpl 2002;8:612-4. 21. Abdalla EK, Denys A, Chevalier P, Nemr RA, Vauthey JN. Total and segmental liver volume variations: implications for liver surgery. Surgery 2004;135:404-10. 22. Nijsten MW, ten Duis HJ, Zijlstra JG, et al. Blunted rise in platelet count in critically ill patients is associated with worse outcome. Crit Care Med 2000;28:3843-6. 23. Tjandra JJ, Fan ST, Wong J. Peri-operative mortality in hepatic resection. Aust N Z J Surg 1991;61:201-6. 24. McCormack L, Petrowsky H, Jochum W, Furrer K, Clavien PA. Hepatic steatosis is a risk factor for postoperative complications after major hepatectomy: a matched case-control study. Ann Surg 2007;245:923-30. 25. McCormack L, Capitanich P, Quinonez E. Liver surgery in the presence of cirrhosis or steatosis: Is morbidity increased? Patient Saf Surg 2008;2:8,9493-2-8. 26. Haga J, Shimazu M, Wakabayashi G, et al. Liver regeneration in donors and adult recipients after living donor liver transplantation. Liver Transpl 2008;14:1718-24. 27. de Boer MT, Christensen MC, Asmussen M, et al. The impact of intraoperative transfusion of platelets and red blood cells on survival after liver transplantation. Anesth Analg 2008;106:32,44. 28. Pereboom IT, Lisman T, Porte RJ. Platelets in liver transplantation: friend or foe? Liver Transpl 2008;14:923-31. 29. Spiess BD. Transfusion of blood products affects outcome in cardiac surgery. Semin Cardiothorac Vasc Anesth 2004;8:267-81. 30. Cognasse F, Boussoulade F, Chavarin P, et al. Release of potential immunomodulatory factors during platelet storage. Transfusion 2006;46:1184-9. 31. Khan SY, Kelher MR, Heal JM, et al. Soluble CD40 ligand accumulates in stored blood components, primes neutrophils through CD40, and is a potential cofactor in the development of transfusion-related acute lung injury. Blood 2006;108:2455-62.

CHAPTER 5

EVIDENCE AGAINST A ROLE OF SEROTONIN IN LIVER REGENERATION IN HUMANS

EDRIS M. ALKOZAI MARTIJN VAN FAASSEN IDO P. KEMA ROBERT J. PORTE TON LISMAN

PUBLISHED IN: HEPATOLOGY. 2015;62:983.

S E R O T O N I N A N D L I V E R R E G E N E R A T I O N

______70

We read with great interest the paper by Starlinger and colleagues in which evidence for a role of serotonin in liver regeneration in humans was provided.1 We would like to report on our findings in a prospective study on serotonin levels in platelet-rich plasma in adult patients undergoing a (extended) right hemi-hepatectomy (n=16) in comparison to levels in patients undergoing a pylorus-preserving pancreaticoduodenectomy (PPPD) (n=10), and healthy controls (n=22). Patient characteristics were published elsewhere.2 We drew blood samples after induction of anesthesia, at the end of the surgery, and at postoperative day 1, 3, 5, 7, and 30. In addition, we took blood samples from the portal and from the hepatic vein just prior to the start and just after completion of parenchymal transection in the patients undergoing a hemi-hepatectomy. Serotonin levels in platelet-rich plasma were determined by liquid chromatography-tandem mass spectrometry and levels were corrected for platelet count. The study protocol was approved by the local medical ethical committee and informed consent was obtained from each participant before inclusion in the study.

8 Hemi-hepatectomy PPPD

) 7

s

t e

l + e

t 6

a

l

p 9

0 5

1

x 1

*

/ *

l 4

o *

m

n (

3 *

e

n i

n 2

o

t

o

r e

S 1

0

ls P K -1 -3 -5 -7 0 T T T T -3 P P P P ro -O -O D D D D - - t- t- t e d O O O O D re re s s n r n P P P P O P o o o P E P -P - C P H -P -P V V P H V V

Figure 1: Median serotonin levels in controls, patients undergoing hemi-hepatectomy, and patients undergoing PPPD. Shown are serotonin levels corrected for platelet count. *P<0.05 compared to the baseline levels of serotonin within the group (Friedman test). +P<0.05 vs controls. End-OK: end of surgery, HV: hepatic vein, PT: parenchyma transection, Pre-OP: Preoperative, POD: postoperative day, VP: vena porta

Serotonin levels at baseline were comparable between patients undergoing hemihepatectomy, patients undergoing PPPD, and healthy subjects (figure 1). In contrast to the Starlinger study, no changes in serotonin were observed in the early post-operative period. Only at postoperative day 5 and 7, serotonin levels clearly decreased, but importantly, the decrease was similar between the hemihepatectomy and PPPD patients. Serotonin content

S E R O T O N I N A N D L I V E R R E G E N E R A T I O N

______71

was identical between samples taken in the afferent and efferent liver prior to and after hemihepatectomy indicating that there was no detectable serotonin consumption by the liver directly after hemihepatectomy.

Although the number of patients we studied was smaller compared to the Starlinger study, we studied more time points in a more homogeneous cohort consisting of non-cirrhotic patients undergoing a major hepatectomy, included an appropriate control group, and studied the serotonin gradient over the liver prior to and just after full parenchymal transection. Technical differences between the studies included measurement of serotonin in platelet-rich plasma versus a calculated serum-platelet poor plasma difference as studied by Starlinger. Importantly, we calculated serotonin content per platelet, thereby correcting for consumption of platelets as a result of dilution or consumption. Although we do not dispute that platelets are likely important for liver regeneration in humans,3 our data do not support the notion that platelet serotonin is key in this process.

S E R O T O N I N A N D L I V E R R E G E N E R A T I O N

______72

REFERENCES

(1) Starlinger P, Assinger A, Haegele S, Wanek D, Zikeli S, Schauer D, et al. Evidence for serotonin as a relevant inducer of liver regeneration after liver resection in humans. Hepatology 2014 Jul;60(1):257-266. (2) Potze W, Alkozai EM, Adelmeijer J, Porte RJ, Lisman T. Hypercoagulability detected by thrombomodulin-modified thrombin generation testing following major partial liver resection despite prolonged conventional coagulation tests. Aliment Pharmacol Ther. 2015 JAN;41(2):189- 98 (3) Alkozai EM, Nijsten MW, de Jong KP, de Boer MT, Peeters PM, Slooff MJ, et al. Immediate postoperative low platelet count is associated with delayed liver function recovery after partial liver resection. Ann Surg 2010 Feb;251(2):300-306.

CHAPTER 6

NO EVIDENCE FOR INCREASED PLATELET ACTIVATION IN PATIENTS WITH HEPATITIS B- OR C-RELATED CIRRHOSIS AND HEPATOCELLULAR CARCINOMA

EDRIS M. ALKOZAI ROBERT J PORTE J. ADELMEIJER ALBERTO ZANETTO PAOLO SIMIONI MARCO SENZOLO TON LISMAN

PUBLISHED IN: THROMBOSIS RESEARCH 2015;135:292–297

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______76

ABSTRACT

BACKGROUND: Cancer is a major risk factor for developing venous thromboembolism (VTE). Plasma hypercoagulability is an established risk factor for cancer-related VTE. In addition, thrombocytosis and hyperreactive platelets have been implicated in VTE and cancer progression. Cirrhosis is associated with changes in platelet number and function. The platelet activation status of patients with cirrhosis and hepatocellular carcinoma has not yet been established. Here we assessed the platelet activation status in patients with hepatitis- related cirrhosis in presence or absence of HCC.

MATERIALS AND METHODS: We performed a cross-sectional study including thirty-eight consecutive patients with hepatitis B- or C- related liver cirrhosis in presence or absence of HCC. We studied basal and agonist-induced platelet activation using flow cytometry. In addition, we studied the plasma levels of von Willebrand factor (VWF) and the VWF-cleaving protease ADAMTS13. Twenty healthy volunteers served as controls.

RESULTS: We found no evidence of basal platelet activation in patients with cirrhosis compared to controls. However, we found reduced agonist-induced platelet activation in patients. No differences in the basal and agonist-induced platelets activation status between patients with or without HCC were detected. Plasma levels of VWF were increased and the levels of ADAMTS13 activity were decreased in patients compared to controls. No differences between the levels of VWF and ADAMTS13 in patients with or without HCC were detected.

CONCLUSIONS: HCC development or recurrence in patients with hepatitis B- or C-related cirrhosis does not appear to be associated with platelet activation and changes in pivotal proteins in primary hemostasis.

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______77

INTRODUCTION

Cancer patients are in a hypercoagulable state and venous thrombotic event might be the first clinical sign of cancer.1 In fact, twenty percent of all cases of venous thromboembolism (VTE) are related to cancer, and the risk of developing VTE is four to seven fold increased in cancer patients.2 VTE in cancer patients is typically associated with plasma hypercoagulability that might be triggered by the cancer cells expressing procoagulant activity or by the host response to cancer (reviewed elsewhere3-5). Also, patient- and treatment-related factors such as chemotherapy, bed rest, infection, and surgery may play a contributory role.6-8 In addition to changes in plasmatic coagulation, quantitative and qualitative changes in platelets have been reported in cancer patients.

Elevation of the circulating platelet count 9,10 (i.e. thrombocytosis) is frequently found in patients with cancer, and is associated with increased risk of VTE11 and poor prognosis.10 Moreover, platelet depletion or inhibition of platelet function has been shown to reduce or prevent cancer growth and distance metastasis in animal models.12-14 In humans, aspirin intake has been shown to reduce death from cancer, and the incidence of metastases.15-18 The detrimental effects of platelets in cancer may involve release of growth factors stored within platelets,19,20 shielding of tumor cells from innate immune surveillance,21 or facilitation of tumor extravasation resulting in metastasis.22

Enhanced activation of platelets has been demonstrated in human studies in multiple types of cancer.23 However, little data on the platelet activation status in patients with hepatocellular carcinoma are available. In this context, this is an interesting tumor type as it frequently occurs in patients with cirrhosis. Cirrhosis is associated with complex changes in the hemostatic system, including quantitative and qualitative changes in platelets.24,25

Thrombocytopenia is common in cirrhosis as is a decreased platelet response in classical platelet aggregometry or the PFA-100.24 It has, however, also been described that platelet functionality is preserved when tested under conditions of flow.26 The in vivo activation status of platelets in patients with cirrhosis is still controversial, with older literature pointing towards a functional defect in platelets (reviewed in24), whereas more recent studies suggest in vivo platelet hyperreactivity.27 Here we studied the in vivo platelet activation status in patients with cirrhosis in presence or absence of HCC. We compared the results with values obtained in healthy individuals.

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______78

METHODS

STUDY POPULATION From February to April 2013, a total of thirty-eight consecutive patients with cirrhosis who visited the outpatient clinic or were admitted to the Hepatology ward of the Department of Gastroenterology of University Medical Center of Padua, Padua, Italy were enrolled into this study. We included all patients who had established cirrhosis, with or without HCC. Excluded were those who used platelet inhibitory drugs such as acetylsalicylic acid or P2Y12 inhibitors, patients who were undergoing eradication treatment for hepatitis B virus (HBV) or hepatitis C virus (HCV) infection, patients with an active HCC who had a treatment-free interval less than two months at the time of inclusion, and patients who were on dialysis. The diagnosis of HCC was obtained by using the European Association for the Study of the Liver criteria, i.e., two imaging procedures (Spiral CT, MRI with paramagnetic contrast injection, or ultrasound with second generation intravenous contrast (Sonovue, Bracco, Italy)) confirming the presence of a lesion.

A control group used to establish reference values for the various tests consisted of 20 adult employees of the University Medical Center Groningen who voluntarily participated in this study. Exclusion criteria for the control group were platelet inhibitory drugs usage such as acetylsalicylic acid or P2Y12 inhibitors, documented history of congenital coagulation disorders, documented history of hepatic disease, and history of viral infection (<2 weeks). This study was performed in accordance with, and was approved by the local ethical committee. Written informed consent was obtained from each patient and control.

BLOOD SAMPLES Blood from patients and controls was drawn by clean venipuncture in 3.8% citrate tubes. A sample was processed directly for flow cytometry and the remainder of the blood was processed to plasma by centrifugation. First, samples were spun at 200 g for 15 min at ambient temperature to obtain platelet-rich plasma. The platelet-rich plasma was transferred to a clean tube and recentrifuged (500 g for 15 min at ambient temperature), in the presence of iloprost (2 ng/ml) purchased from Santa Cruz Biotechnology, Dallas, TX U.S.A. The supernatant, platelet poor plasma, was then collected in 2 ml tubes, snap-frozen, and stored at -80 °C until use.

STUDY VARIABLES Patient characteristics were obtained from patient charts. These included age, gender, the severity of cirrhosis (according to the Child-Pugh classification), platelet count, the international normalized ratio (INR), creatinine levels, and total bilirubin levels. For patients with HCC we also obtained the number of tumor lesions, the size of the tumors and the total

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______79

Tabel 1: Participant characteristics Controls Cirrhosis without Cirrhosis with Demographic characteristics P (n = 20) HCC (n = 16) HCC (n = 22) Participant variables Gender, Male 11 (55%) 10 (63%) 20 (95%) 0.01 Age, mean (SD) 29 (±5) 55 (±14) 70 (±10) < 0.01 Platelet count, (G/L), median - 107 (48-159) 102 (68-152) 0.69 (IQR) INR, median (IQR) - 1.3 (1.2-1.6) 1.2 (1.1-1.2) 0.03 MELD, median (IQR) - 12 (9-16) 10 (8-11) 0.09 Etiology of liver Cirrhosis - 0.22 HBV 3 (19%) 7 (33%) HCV 13 (81%) 12 (57%) HBV + HCV 0 2 (10%) Child Pugh, n (%) - 0.02 A 4 (25%) 11 (52%) B 9 (56%) 6 (29%) C 3 (19%) 2 (10%) AFP, median (IQR) - 6 (5-23) 14 (3-24) 0.55 Tumor characteristics - - Treated (TACE/RFA), yes - 13 (59%) Number of lesions, median (IQR) - 2 (1-3) Total size, mm (IQR) - 40 (25-71)

AFP indicates Alpha fetoprotein; HBV, hepatitis B virus, HCC; Hepatocellular carcinoma, HCV, hepatitis C virus; INR, International Normalised Ratio; IQR, interquartile range; MELD: Model of End stage Liver Disease; RFA, radiofrequency ablation; SD, standard deviation, TACE, transarterial chemoembolization. mass of HCC, whether or not the patients had a vasoinvasion, the presence of portal vein thrombosis, and if they were previously treated for the HCC. We calculated the Model for End-Stage Liver Disease score (0.957 × loge (creatinine)+0.378 × loge (bilirubin) +1.12 × loge (INR) +0.643] × 10) based on the last laboratory measurements just prior to the blood draw for the assays described in the present study. When necessary, computer-stored hospital files were reviewed for other relevant clinical parameters.

FLOW CYTOMETRY The platelet activation status was assessed using flow cytometry in whole blood. Five μl of blood was added to a 50 μl mixture in HEPES-buffered saline (10 mM HEPES [N-2- hydroxyethylpiperazine-N’-2-ethanesulfonic acid], 137 mM NaCl, 2.68 mM KCl, 0.42 mM NaH2PO4, 1.7mMMgCl2, 5mMd-glucose, pH 7.4) containing vehicle or activators and a R- phycoerythrin (R-PE) labeled monoclonal antibody against human P-selectin (#555524, BD Pharmingen™, Franklin Lakes, NJ, 2 μl). Platelets were kept in a resting state or activated by adenosine diphosphate (ADP, 15 μM, Stago, Asnières, France), thrombin receptor activating peptide (TRAP6, 15 μM, Bachem, Bubendorf, Switserland), or a combination of both. After 20 min of incubation, the samples were fixed with 500 μL 0.2% formyl saline (0.2% formaldehyde in 0.9% NaCl) and kept at 4 °C in the dark until the analysis. A negativecontrol (mouse IgG1 monoclonal antibodies conjugated to R-PE (#555749, BD Pharmigen) was used.

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______80

The patient samples were analyzed on a Cytomics FC500 flow cytometer (Beckman Coulter, Fullerton, CA, USA) at the University of Padua. Control samples were analyzed using a FACS Calibur flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA). All samples were analyzed within six hours after processing. Samples were gated on the basis of their forward and sideward scatter properties. The percentage of platelets expressing P-selectin were recorded. The proportion of platelets positive for P-selectin after activation were corrected for the proportion of platelet positive for P-selectin in the non-activated sample of each patient and control. The increase in the proportion of P-selectin positive platelets obtained in this manner represents the extent to which platelets in a given sample can be activated by a given activator (i.e., the platelet activatability).

PLASMA LEVELS OF VWF, ADAMTS13 AND SP-SELECTIN VWF antigen (VWF) levels were determined with an in-house enzyme-linked immunosorbent assay (ELISA) using commercially available polyclonal antibodies against VWF (DAKO, Glostrup, Denmark). Plasma a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13 (ADAMTS13) activity was assessed using the FRETS-VWF73 assay (Peptanova, Sandhausen, Germany) based on the method described by Kokame and associates. 28 VWF and ADAMTS levels in pooled normal plasma were set at 100%, and values obtained in test plasmas were expressed as a percentage of pooled normal plasma. Soluble P-selectin (sP-selectin) levels were determined using a commercially available sP- selectin ELISA (R&D Systems, Abingdon, UK), and values were both expressed as the sP- selectin concentration in plasma and as plasma concentrations corrected for the platelet count in platelet-rich plasma.

STATISTICAL ANALYSIS Statistical analyses were performed using the statistical software package SPSS 20 (IBM SPSS, Chicago, IL). Categorical variables are shown as numbers and percentages and groups were compared using Pearson's chi-squared test. Continuous variables are presented as means with standard deviation (SD) or as medians with interquartile range (IQR) based on their distribution. Continuous variables were compared using a standard t- test or the Mann Whitney U test, as appropriate. P values less than 0.05 were considered statistically significant.

RESULTS

STUDY POPULATION Patient characteristics are summarized in Table 1. Included were thirty-eight patients with hepatitis-related cirrhosis of whom twenty-two (58%) had HCC. The patients that had HCC were more frequently male than patients that did not have HCC (95% vs 63%). There was no

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______81

significant difference between the patients with and without HCC regarding the type of hepatitis as underling cause of cirrhosis (HBV or HCV, 33% and 57% in patients with HCC vs 19% and 81% in the patients without HCC). Concomitant HBV and HCV infection was present in two patients (10%) who had HCC. Thirteen (59%) out of twenty-two patients that had an active HCC due to a recurrence or a residual tumor at the time of inclusion had previously been treated. Twelve patients were treated with trans-arterial chemoembolisation, one patient with radiofrequency ablation. The median time between treatment and the blood draw for the present study was 357 (IQR 127-608) days. None of the patients had evidence for vasoinvasion and none of the patients had portal vein thrombosis.

A Non-stimulated B ADP-stimulate

100 100 *** *

g

g

n

i

n i

80 s 80

%

s

s

%

-

s

e

-

r

e

s

r

p t

60 s 60

t

p

x

e

l

e

x

e

l

e

e

t

e

P t

a 40 40

P

2

l

a

l

2

6

a

l

p

6

D p

D 20 20

C

C  0 0 HCC non-HCC Controls HCC non-HCC Controls

C TRAP-stimulated D ADP & TRAP-stimulated

100 *** *** 100 ** **

g

g

n

n

i

i s

80 s 80

s

s

%

%

e

e

-

-

r

r

p

p s

60 s 60

t

t

x

x

e

e

e

e

l

l

e

e

P

P t

40 t 40

2

2

a

a

l

l

6

6

p

p D

20 D 20

C

C

  0 0 HCC non-HCC Controls HCC non-HCC Controls

Figure 1: Basal and agonist-induced platelet activation in patients with cirrhosis with or without HCC. The basal platelet activation status (A) and the agonist-induced platelet activatability in patients and controls using ADP (B), TRAP (C), and the combination of ADP and TRAP (D) as assessed by flow cytometry for P-selectin in patients with or without HCC and healthy controls. Shown is the percentage of P-selectin positive platelets. Agonist-induced platelet activatability values were corrected for baseline values. Horizontal lines represent medians. * P < 0.05, ** P < 0.01, *** P < 0.001, all v.s. control.

The patients were classified according to the Child-Pugh classification.29 The patients who had HCC appeared to have slightly milder liver disease compared to those who did not have HCC. Furthermore, the patients with HCC were significantly older, had a lower international normalized ratio and MELD score, but had similar platelet counts compared to the patients without HCC. The patients with HCC had a median of 2 lesions (IQR, 1-3), a cumulative

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______82

tumor size of 4 cm (IQR, 3-7) and a serum alpha-fetoprotein (AFP) of 12.5 ng/ml (IQR, 3.8- 23.5). One patient had an AFP level of 2.750 ng/ml, and a cumulative tumor size of 20 cm. No patients had portal vein thrombosis, extrahepatic manifestation of HCC, or macroscopic vasoinvasion.

IS BASAL OR AGONIST-INDUCED PLATELET ACTIVATION INCREASED IN PATIENTS WITH HCC? Figure 1 shows the basal platelet activation status and the agonist induced platelet activatability (defined as the increase in the proportion of P-selectin positive platelets after activation) in patients and controls. The number of p-selectin positive platelets at Baseline was not different between patients and controls. No differences were detected in the basal platelet activation status between patients that did or did not have HCC. When platelets were activated in vitro using ADP, TRAP or the combination of both, the percentage of platelets expressing P-selectin were significantly lower in patients than controls when corrected for baseline values. However, no differences were detected between patients that did or did not have HCC.

ARE PLASMA MARKERS OF PRIMARY HEMOSTASIS INCREASED IN PATIENTS WITH HCC? We next assessed plasma levels of soluble P-selectin as a marker of in vivo platelet activation (Fig. 2). Plasma sP-selectin levels were similar between the patients and controls or between the patients with or without HCC. However, when the levels of sP- selectin were corrected for the circulating platelet count, levels of sP-selectin were significantly higher in patients without HCC. Plasma levels of VWF were significantly higher in patients compared to controls, but levels were similar between patients with and without HCC. ADAMTS13 activity was lower in patients compared to controls, although the difference between controls and patients without HCC did not reach statistical significance. ADAMTS13 activity was comparable between patients with or without HCC.

DISCUSSION

In this study we detected no differences in basal platelet activation status and platelet activatability between patients that had hepatitis-related cirrhosis and patients that had hepatitis-related cirrhosis and HCC. In addition the VWF/ADAMTS13 unbalance associated with cirrhosis, which we previously showed to promote platelet function,30,31 was similar in patients with and without HCC. The basal platelet activation Status was similar between patients and controls. We did, however, detect substantially decreased platelet activatability in patients. Based on previous studies on the platelet activation status in different types of cancer, we anticipated that the platelet activation status in patients with cirrhosis who developed HCC would be increased in comparison to patients with cirrhosis without HCC.32- 38 However, our results indicate that patients with cirrhosis and HCC might not have a

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______83

A B ** l 50 m 600

/ ***

g

6

u

0

40 )

1

n

i

g

r

t

p

e c

( 400

p e

30

l

s

t

n

e

i

e

s

t

l

-

c e

P 20

e t

l

s a

200

e

l

a

s

p P

m 10

s

s

a l

P 0 0 HCC Non-HCC Controls HCC Non-HCC Controls

C D *** ***

2000 ) 200

% *

(

y

t 160

1500 i

v

)

i

t

% c

( 120

a

F 1000

3 1

W 80

S

V T

500 M

40

A D

0 A 0 HCC Non-HCC Controls HCC Non-HCC Controls

Figure 2: Plasma indices of primary hemostasis in patients with cirrhosis with or without HCC. Plasma levels of soluble P-selectin (A), soluble P-selectin corrected for platelet count (B), VWF (C), and ADAMTS13 (D) in patients with or without HCC and healthy controls. Horizontal lines represent medians. * p < 0.05, ** p < 0.01, *** P < 0.001, all v.s. control (except in panel B, ** P < 0.01 HCC vs non-HCC). relative hyperreactivity of their primary hemostatic system compared to patients with cirrhosis without HCC. These results indicate that the complexity of the alterations in the primary hemostatic system of cirrhosis may overwhelm the changes in platelet function as a result of the HCC. Platelets are increasingly recognized as important players in cancer growth and metastasis.12,13,39 Given the clinical benefit of aspirin in metastasis and cancer- related death, platelets may be an attractive therapeutic target for cancer in general. Recently, Sitia and colleagues 13 demonstrated that inhibition of platelet function prevents or delays the onset of HCC in a mouse model of chronic immune-mediated hepatitis B. The authors postulated that platelets are key players of immune-mediated necroinflammatory process that might be responsible for the onset of HCC in chronically infected patients. Given the role of platelets in inflammation and their multiple roles in liver diseases,40 it may be that platelets are active players in disease progression by delivery of angiogenic proteins within the liver. We have recently shown that the intraplatelet levels of angiogenic proteins including VEGF, bFGF, and HGF were elevated in patients with hepatitis B or C related cirrhosis that did or did not have HCC.41 However, as we have now demonstrated that the capacity to activate platelets in patients with HCC is decreased, the question arises whether secretion of such factors optimally occurs in this setting. In other words, would in vivo

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______84

secretion of these factors be decreased as a result of attenuated platelet activation, then the role of platelets in HCC development may be less than the role of platelets in other cancer types, and antiplatelet drugs may consequently be less effective in HCC patients compared to patients with different tumor types. On the other hand, our in vitro activation assay may not fully reflect physiology. We have previously demonstrated that platelet activation when studied under conditions of flow is similar in patients with cirrhosis compared to controls, despite decreased platelet activation capacity in assays performed under static conditions.26

Our data on the basal platelet activation status and in vitro activatability of patients with cirrhosis contrasts with recent findings in which platelet hyperactivity in cirrhosis was suggested.27 Differences in methodology as outlined by us previously may be responsible for these discrepant results.42 We demonstrate that basal platelet activation is not increased in cirrhosis, and that platelets cannot be fully activated, which is in line previous findings.43-45 Although the increased sP-selectin levels corrected for the platelet count (Fig. 2B) may suggest enhanced in vivo platelet activation it should be noted that sP-selectin is likely cleared by the liver, which, in patients with cirrhosis, may lead to falsely elevated levels.

Our patients had significantly higher plasma levels of VWF than controls. This finding is in concordance with our previous study showing increasing levels of VWF in patients with cirrhosis that correlated with increasing severity of liver disease.30 Notably, the levels of VWF were comparable in patients in presence or absence of HCC. It has been suggested that VWF plasma levels may be of prognostic significance in patients with solid tumors,46-48 but the prognostic significance of VWF in HCC is absent, perhaps because of the effects of the underlying cirrhotic disease on VWF levels.

We found lower levels of ADAMTS13 compared to controls although the difference between the controls and patients without HCC did not reach statistical significance. Lower levels of ADAMTS13 activity have been reported in patients with liver disease.49 ADAMTS13 levels have been reported as an independent risk factor of development of HCC within a year.50 This is in contrast to our finding, since we found no difference in the levels of ADAMTS13 activity between the patient in presence or absence of HCC. In conclusion, we found no evidence of increased platelet activation and altered activatability in patients with hepatitis B- or C related cirrhosis that developed HCC compared to patients without HCC. In addition, no changes in the VWF/ADAMTS13 axis were detected. HCC development or recurrence in hepatitis B or C-related cirrhosis may thus not be associated with changes in primary hemostasis.

ACKNOWLEDGEMENTS: We thank Dr. Claudia M. Radu for expert laboratory assistance.

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______85

REFERENCES 1. Prandoni P, Lensing AW, Buller HR, et al. Deep-vein thrombosis and the incidence of subsequent symptomatic cancer. N Engl J Med 1992;327:1128-33. 2. Timp JF, Braekkan SK, Versteeg HH, Cannegieter SC. Epidemiology of cancer-associated venous thrombosis. Blood 2013;122:1712-23. 3. Gomes M, Khorana AA. Risk assessment for thrombosis in cancer. Semin Thromb Hemost 2014;40:319-24. 4. Piccioli A, Falanga A, Baccaglini U, Marchetti M, Prandoni P. Cancer and venous thromboembolism. Semin Thromb Hemost 2006;32:694-9. 5. Falanga A, Donati MB. Pathogenesis of thrombosis in patients with malignancy. Int J Hematol 2001;73:137-44. 6. Khorana AA, Kuderer NM, Culakova E, Lyman GH, Francis CW. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood 2008;111:4902-7. 7. Sallah S, Wan JY, Nguyen NP. Venous thrombosis in patients with solid tumors: determination of frequency and characteristics. Thromb Haemost 2002;87:575-9. 8. Kroger K, Weiland D, Ose C, et al. Risk factors for venous thromboembolic events in cancer patients. Ann Oncol 2006;17:297-303. 9. Hwang SJ, Luo JC, Li CP, et al. Thrombocytosis: a paraneoplastic syndrome in patients with hepatocellular carcinoma. World J Gastroenterol 2004;10:2472-7. 10. Stone RL, Nick AM, McNeish IA, et al. Paraneoplastic thrombocytosis in ovarian cancer. N Engl J Med 2012;366:610-8. 11. Simanek R, Vormittag R, Ay C, et al. High platelet count associated with venous thromboembolism in cancer patients: results from the Vienna Cancer and Thrombosis Study (CATS). J Thromb Haemost 2010;8:114-20. 12. Gay LJ, Felding-Habermann B. Contribution of platelets to tumour metastasis. Nat Rev Cancer 2011;11:123-34. 13. Sitia G, Aiolfi R, Di Lucia P, et al. Antiplatelet therapy prevents hepatocellular carcinoma and improves survival in a mouse model of chronic hepatitis B. Proc Natl Acad Sci USA 2012;109:E2165-72. 14. Ho-Tin-Noe B, Goerge T, Wagner DD. Platelets: guardians of tumor vasculature. Cancer Res 2009;69:5623-6. 15. Rothwell PM, Fowkes FG, Belch JF, Ogawa H, Warlow CP, Meade TW. Effect of daily aspirin on long-term risk of death due to cancer: analysis of individual patient data from randomised trials. Lancet 2011;377:31-41. 16. Rothwell PM, Price JF, Fowkes FG, et al. Short-term effects of daily aspirin on cancer incidence, mortality, and non-vascular death: analysis of the time course of risks and benefits in 51 randomised controlled trials. Lancet 2012;397:1602-12. 17. Rothwell PM, Wilson M, Price JF, Belch JF, Meade TW, Mehta Z. Effect of daily aspirin on risk of cancer metastasis: a study of incident cancers during randomised controlled trials. Lancet 2012;397:1591-601. 18. Algra AM, Rothwell PM. Effects of regular aspirin on long-term cancer incidence and metastasis: a systematic comparison of evidence from observational studies versus randomised trials. Lancet Oncol 2012;13:518-27. 19. Cho MS, Bottsford-Miller J, Vasquez HG, et al. Platelets increase the proliferation of ovarian cancer cells. Blood 2012;120:4869-72.

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______86

20. Italiano JE,Jr, Richardson JL, Patel-Hett S, et al. Angiogenesis is regulated by a novel mechanism: pro- and antiangiogenic proteins are organized into separate platelet alpha granules and differentially released. Blood 2008;111:1227-33. 21. Nieswandt B, Hafner M, Echtenacher B, Mannel DN. Lysis of tumor cells by natural killer cells in mice is impeded by platelets. Cancer Res 1999;59:1295-300. 22. Schumacher D, Strilic B, Sivaraj KK, Wettschureck N, Offermanns S. Platelet-derived nucleotides promote tumor-cell transendothelial migration and metastasis via P2Y2 receptor. Cancer Cell 2013;24:130-7. 23. Riedl J, Pabinger I, Ay C. Platelets in cancer and thrombosis. Hamostaseologie 2014;34:54- 62. 24. Witters P, Freson K, Verslype C, et al. Review article: blood platelet number and function in chronic liver disease and cirrhosis. Aliment Pharmacol Ther 2008;27:1017-29. 25. Violi F, Basili S, Raparelli V, Chowdary P, Gatt A, Burroughs AK. Patients with liver cirrhosis suffer from primary haemostatic defects? Fact or fiction? J Hepatol 2011;55:1415-27. 26. Lisman T, Adelmeijer J, de Groot PG, Janssen HL, Leebeek FW. No evidence for an intrinsic platelet defect in patients with liver cirrhosis--studies under flow conditions. J Thromb Haemost 2006;4:2070-2. 27. Basili S, Raparelli V, Riggio O, et al. NADPH oxidase-mediated platelet isoprostane over- production in cirrhotic patients: implication for platelet activation. Liver Int 2011;31:1533-40. 28. Kokame K, Nobe Y, Kokubo Y, Okayama A, Miyata T. FRETS-VWF73, a first fluorogenic substrate for ADAMTS13 assay. Br J Haematol 2005;129:93-100. 29. Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646-9. 30. Lisman T, Bongers TN, Adelmeijer J, et al. Elevated levels of von Willebrand Factor in cirrhosis support platelet adhesion despite reduced functional capacity. Hepatology 2006;44:53-61. 31. Hugenholtz GC, Adelmeijer J, Meijers JC, Porte RJ, Stravitz RT, Lisman T. An unbalance between von Willebrand factor and ADAMTS13 in acute liver failure: implications for hemostasis and clinical outcome. Hepatology 2013;58:752-61. 32. Roselli M, Mineo TC, Martini F, et al. Soluble selectin levels in patients with lung cancer. Int J Biol Markers 2002;17:56-62. 33. Ferroni P, Roselli M, Martini F, et al. Prognostic value of soluble P-selectin levels in colorectal cancer. Int J Cancer 2004;111:404-8. 34. Blann AD, Gurney D, Wadley M, Bareford D, Stonelake P, Lip GY. Increased soluble P-selectin in patients with haematological and breast cancer: a comparison with fibrinogen, plasminogen activator inhibitor and von Willebrand factor. Blood Coagul Fibrinolysis 2001;12:43-50. 35. Caine GJ, Lip GY, Stonelake PS, Ryan P, Blann AD. Platelet activation, coagulation and angiogenesis in breast and prostate carcinoma. Thromb Haemost 2004;92:185-90. 36. Starlinger P, Moll HP, Assinger A, et al. Thrombospondin-1: a unique marker to identify in vitro platelet activation when monitoring in vivo processes. J Thromb Haemost 2010;8:1809-19. 37. Gong L, Cai Y, Zhou X, Yang H. Activated platelets interact with lung cancer cells through P- selectin ligand-1. Pathol Oncol Res 2012;18:989-96. 38. Mantur M, Kemona H, Kozlowski R, Kemona-Chetnik I. Effect of tumor stage and nephrectomy on CD62P expression and sP-selectin concentration in renal cancer. Neoplasma 2003;50:262-5.

P L A T E L E T A C T I V A T I O N I N P A T I E N T S W I T H H C C

______87

39. Erpenbeck L, Schon MP. Deadly allies: the fatal interplay between platelets and metastasizing cancer cells. Blood 2010;115:3427-36. 40. Lisman T, Porte RJ. The role of platelets in liver inflammation and regeneration. Semin Thromb Hemost 2010;36:170-4. 41. Alkozai EM, Lisman T, Porte RJ, Nijsten MW. Early elevated serum gamma glutamyl transpeptidase after liver transplantation is associated with better survival. F1000Res 2014;3:85. 42. Lisman T, Porte RJ. Pitfalls in assessing platelet activation status in patients with liver disease. Liver Int 2012;32:1027; author reply 1028. 43. Laffi G, Cominelli F, Ruggiero M, et al. Altered platelet function in cirrhosis of the liver: impairment of inositol lipid and arachidonic acid metabolism in response to agonists. Hepatology 1988;8:1620-6. 44. Laffi G, Marra F, Gresele P, et al. Evidence for a storage pool defect in platelets from cirrhotic patients with defective aggregation. Gastroenterology 1992;103:641-6. 45. Laffi G, Marra F, Failli P, et al. Defective signal transduction in platelets from cirrhotics is associated with increased cyclic nucleotides. Gastroenterology 1993;105:148-56. 46. Schellerer VS, Mueller-Bergh L, Merkel S, et al. The clinical value of von Willebrand factor in colorectal carcinomas. Am J Transl Res 2011;3:445-53. 47. Gadducci A, Baicchi U, Marrai R, Del Bravo B, Fosella PV, Facchini V. Pretreatment plasma levels of fibrinopeptide-A (FPA), D-dimer (DD), and von Willebrand factor (vWF) in patients with ovarian carcinoma. Gynecol Oncol 1994;53:352-6. 48. Zietek Z, Iwan-Zietek I, Paczulski R, Kotschy M, Wolski Z. von Willebrand factor antigen in blood plasma of patients with urinary bladder carcinoma. Thromb Res 1996;83:399-402. 49. Mannucci PM, Canciani MT, Forza I, Lussana F, Lattuada A, Rossi E. Changes in health and disease of the metalloprotease that cleaves von Willebrand factor. Blood 2001;98:2730-5. 50. Ikeda H, Tateishi R, Enooku K, et al. Prediction of hepatocellular carcinoma development by plasma ADAMTS13 in chronic hepatitis B and C. Cancer Epidemiol Biomarkers Prev 2011;20:2204-11.

CHAPTER 7

LEVELS OF ANGIOGENIC PROTEINS IN PLASMA AND PLATELETS ARE NOT DIFFERENT BETWEEN PATIENTS WITH HEPATITIS B/C - RELATED CIRRHOSIS AND PATIENTS WITH CIRRHOSIS AND HEPATOCELLULAR CARCINOMA

EDRIS M. ALKOZAI ROBERT J PORTE JELLE ADELMEIJER ALBERTO ZANETTO PAOLO SIMIONI MARCO SENZOLO TON LISMAN

PUBLISHED IN: PLATELETS. 2015;26:577-82.

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______90

ABSTRACT

INTRODUCTION: Increasing evidence suggests that levels of angiogenic proteins within blood platelets change at the earliest stages of cancer development and may thus provide a promising diagnostic and prognostic tool. Patients with cirrhosis have increased risk of developing hepatocellular carcinoma (HCC).

AIMS: We aimed to study whether development of HCC in hepatitis related cirrhosis results in changes in platelet levels of angiogenic proteins.

MATERIALS and methods: We studied the intraplatelet levels of vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), platelet-derived growth factor (PDGF), hepatocyte growth factor (HGF), endostatin, platelet factor 4 (PF4) and thrombospondin type 1 (TSP-1) in 38 consecutive patients with hepatitis B- or C- related liver cirrhosis with or without HCC in addition to plasma levels of the same proteins. Twenty healthy volunteers were included to establish reference values for the various tests.

RESULTS: Intraplatelet levels of VEGF, bFGF, HGF and endostatin were significantly higher in patients compared to controls. Intraplatelet levels of PDGF, PF4 and TSP-1 were comparable between patients and controls. Plasma levels of VEGF, bFGF and endostatin were comparable between patients and controls. Plasma levels of PDGF, PF4 and TSP-1 were decreased in patients, but this difference disappeared when levels were corrected for platelet count. Intraplatelet and plasma levels of all proteins assessed were comparable between patients with and without HCC.

CONCLUSION: The intraplatelet levels of some angiogenic proteins are elevated in cirrhosis, but do not discriminate between patients with and without HCC. Thus, intraplatelet levels of angiogenic proteins do not seem useful as diagnostic or prognostic biomarker of HCC in cirrhotic patients.

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______91

INTRODUCTION

Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related death worldwide.1 Although numerous treatment modalities are available for HCC,2 only 30%–40% of patients are eligible for curative interventions, and in these patients, the five-year survival is reported between 66% and 90%.3-5 Therefore, early detection of tumor growth and recurrence could save many lives.

In general, tumor growth beyond one to two millimeters in size is dependent on angiogenesis.6 Angiogenesis is regulated by a continuous interplay of stimulatory and inhibitory proteins.7 Since HCC is a hypervascularized cancer,8,9 the levels of angiogenic proteins may represent a desirable diagnostic and prognostic tool.10,11 Therefore, several studies suggested that serum or plasma levels of angiogenesis regulatory proteins are of prognostic value.12-14 However, the plasma and serum levels of these proteins may not be accurate predictors since the biovariability of these proteins may vary widely over time.15-18 It has been well established that platelets play a key role in tumor growth and metastasis.19- 24 Furthermore, it has been demonstrated that platelets actively sequester angiogenic proteins from the blood circulation.25,26 In mice, platelet levels of angiogenic proteins were already elevated when tumors were smaller than one millimeter, without an increase of levels of these proteins in plasma.26 Based on these results, it has been postulated that the concentration of angiogenic proteins within platelets is a sensitive and early marker for the presence of tumor, and that platelet levels are superior predictors of progression or recurrence of cancer compared to levels measured in plasma or serum. Indeed, one study in humans showed that some, but not all, angiogenic proteins were elevated in platelets from patients with colorectal cancer,27 and platelet (but not plasma) levels of these proteins were independent predictors of the presence of tumor.

To our knowledge, no study has yet been performed investigating platelet levels of angiogenic proteins in patients with HCC. Since cirrhosis, especially in combination with Hepatitis B or C virus (HBV or HCV) infection, is a major risk factor for developing HCC, we hypothesized that the intraplatelet levels of angiogenic proteins might be predictors of the presence of HCC in patients with hepatitis-related cirrhosis. Therefore, we studied platelet and plasma levels of various angiogenic proteins in patients with hepatitis-related cirrhosis in presence and absence of HCC.

METHOD

STUDY POPULATION From February to April 2013, a total of 38 consecutive patients with cirrhosis who visited the outpatient clinic or were admitted to the Hepatology ward of the Department of Gastroenterology of University Medical Center of Padua, Padua, Italy, were enrolled into this

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______92

study. We included all patients who had established cirrhosis due to HBV or HCV infection, with or without HCC. Excluded were those who used platelet inhibitory drugs such as acetylsalicylic acid or P2Y12 inhibitors, patients who were undergoing eradication treatment for HBV or HCV infection, patients with an active HCC who had a treatment-free interval less than two months at the time of inclusion and patients who were on dialysis. The diagnosis of HCC was obtained by using the European Association for the Study of the Liver criteria, i.e. two imaging procedures (spiral CT, MRI with paramagnetic contrast injection or ultrasound with second generation intravenous contrast (Sonovue, Bracco, Italy)) confirming the presence of a lesion.

We included 20 adult employees of the University Medical Center Groningen, The Netherlands, to establish reference values for the tests performed. Exclusion criteria for the control group were platelet-inhibitory drugs usage such as acetylsalicylic acid or P2Y12 inhibitors documented history of congenital coagulation disorders, documented history of hepatic disease and history of viral infection (52 weeks). This study was performed in accordance with the declaration of Helsinki and was approved by the local ethical committee. Written informed consent was obtained from each individual.

STUDY VARIABLES Patient characteristics and variables were obtained from patient charts. These included age, gender, the severity of liver cirrhosis (according to the Child-Pugh classification) and platelet count. We calculated the Model for End-Stage Liver Disease (MELD) score based on the last laboratory measurements prior to the blood draw. For patients with HCC, we also obtained the number of tumor lesions, the size of the tumors and the total mass of HCC, the presence of vasoinvasion and if the patients were previously treated for the HCC. When necessary, computer-stored hospital files were reviewed for other relevant clinical parameters.

PLATELET COUNT, ISOLATION OF BLOOD PLATELETS AND PREPARATION OF PLASMA Blood was drawn from the antecubital vein through 20-gauge needles into vacuum 3.2% sodium citrate (9:1, v/v) tubes. Blood platelet count was determined using a Beckman Coulter LH755 Analyser (Miami, FL). Platelets were isolated by differential centrifugation. First, samples were spun at 200 g for 15 min at ambient temperature to obtain platelet-rich plasma (PRP). PRP was transferred to a clean tube and recentrifuged (500 g for 15 min at ambient temperature) in the presence of iloprost (2 ng/ml) purchased from Santa Cruz Biotechnology (Dallas, TX). The supernatant, platelet-poor plasma, was then collected in 2ml tubes, snap-frozen and stored at -80°C until use. The platelet pellet was resuspended in Hepes-Tyrode buffer (10mM HEPES [N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid], 137mM NaCl, 2.68mM KCl, 0.42mM NaH2PO4, 1.7mM MgCl2, 5mM d-glucose, pH 7.4) and recentrifuged (500 g for 15 min, ambient temperature) in the presence of iloprost (2 ng/ml). Then, the supernatant was discarded, and the platelet pellet was resuspended in

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______93

Hepes-Tyrode buffer. The platelets count in this suspension was determined and platelets were lysed by repeated free–thaw cycles and stored at -80°C until use.

Tabel 1: Participant characteristics Controls Cirrhosis without Cirrhosis with Demographic characteristics (n = 20) HCC (n = 16) HCC (n = 22) P Participant variables Gender, Male 11 (55%) 10 (63%) 20 (95%) 0.01 Age, mean (SD) 29 (±5) 55 (±14) 70 (±10) < 0.01 Platelet count, (G/L), median (IQR) - 107 (48-159) 102 (68-152) 0.69 INR, median (IQR) - 1.3 (1.2-1.6) 1.2 (1.1-1.2) 0.03 MELD, median (IQR) - 12 (9-16) 10 (8-11) 0.09 Etiology of liver Cirrhosis - 0.22 HBV 3 (19%) 7 (33%) HCV 13 (81%) 12 (57%) HBV + HCV 0 2 (10%) Child Pugh, n (%) - 0.02 A 4 (25%) 11 (52%) B 9 (56%) 6 (29%) C 3 (19%) 2 (10%) AFP, median (IQR) - 6 (5-23) 14 (3-24) 0.55 Tumor characteristics - - Treated (TACE/RFA), yes - 13 (59%) Number of lesions, median (IQR) - 2 (1-3) Total size, mm (IQR) - 40 (25-71) AFP indicates Alpha fetoprotein; HBV, hepatitis B virus, HCC; Hepatocellular carcinoma, HCV, hepatitis C virus; INR, International Normalised Ratio; IQR, interquartile range; MELD: Model of End stage Liver Disease; RFA, radiofrequency ablation; SD, standard deviation, TACE, transarterial chemoembolization.

LEVELS OF ANGIOGENIC PROTEINS IN PLASMA AND WITHIN PLATELETS Plasma levels of vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), platelet-derived growth factor (PDGF), hepatocyte growth factor (HGF), endostatin, platelet factor 4 (PF4) and thrombospondin type 1 (TSP-1) were assessed using commercially available enzyme-linked immunosorbent assays (Duoset, R&D systems, Abingdon, UK) as previously described.18 Levels of these proteins in platelet lysates were determined with the same assay. Platelet residues in the lysates were spun down prior to the assay. Levels of the angiogenic proteins in the platelet lysates were corrected for the content of the lysates as described previously 18 using a commercially available monoclonal antibody (MAB1501R; Milipore, Amsterdam, The Netherlands). For detection, this antibody was biotinylated using a commercially available biotinylation kit (Pierce, Rockford, IL).

All assays were validated by spiking experiments in which a known concentration of the protein was added to plasma or platelet lysate from a single healthy donor after which the recovery of the protein was determined by ELISA. Recoveries for the various tests ranged

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______94

from 83 to 94% in plasma and from 65 to 96% in platelet lysate. Accuracy of the measurements was further ensured by control samples on each ELISA plate. Between plate coefficients of variation varied between 3 and 19% for the various tests.

STATISTICAL ANALYSIS Statistical analyses were performed using the statistical software package SPSS 20 (IBM SPSS, Chicago, IL). Categorical variables are shown as numbers and percentages and groups were compared using Pearson’s chi-squared test. Continuous variables are presented as means with standard deviation or as medians with interquartile range (IQR) based on their distribution. Continuous variables were compared using a standard t-test or the Mann–Whitney U-test, as appropriate. A P < 0.05 was considered statistically significant.

RESULTS

STUDY PPOPULATION Patient characteristics are summarized in Table 1. Included were 38 patients with hepatitis- related cirrhosis of whom 22 (58%) also had HCC. One patient with HCC was excluded due to the loss of samples during pre-analytical handling. The patients that had HCC were more frequently male than patients that did not have HCC (95% vs. 63%, respectively). There was no significant difference between the patients with or without HCC regarding the type of hepatitis as underling cause of liver cirrhosis (HBV or HCV, 33% and 57% in the patients with HCC vs. 19% and 81% in the patients without HCC). Concomitant HBV and HCV infection was present in two (10%) patients who had HCC.

Thirteen (59%) of twenty two patients that had an active HCC due to a recurrence or a residual tumor at the time of inclusion had previously been treated, ten patients with trans- arterial chemoembolization and one patient with radiofrequency ablation. The patients were classified according to the Child-Pugh classification.28 The patients who had HCC appeared to have milder liver disease compared to those who did not have HCC. However, the difference between the groups was not statistically significant.

Furthermore, the patients with HCC were significantly older, had a lower INR and MELD score, but had similar platelet counts compared to the patients without HCC. The patients with HCC had a median of two lesions (IQR 1–3), a cumulative tumor size of 40mm (IQR: 30–70) and a serum alpha fetoprotein (AFP) of 12.5 ng/ml (IQR: 3.8–23.5). One patient had an AFP level of 2.750 ng/ml and a cumulative tumor size of 85 mm. No patient had portal vein thrombosis, extrahepatic manifestation of HCC or vasoinvasion.

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______95

LEVELS OF PROANGIOGENIC PROTEINS IN PLASMA AND PLATELETS FROM PATIENTS WITH HEPATITIS-RELATED CIRRHOSIS WITH AND WITHOUT HCC Figure 1 shows levels of proteins involved in stimulation of angiogenesis in plasma and platelets from patients and controls. Plasma levels of VEGF were below the detection limit in the majority of patients and controls. Plasma levels of bFGF were not different between patients and controls. Plasma PDGF levels were lower in patients compared to controls, and plasma HGF was higher in patients as compared to the controls. No differences in plasma levels of any of the proangiogenic factors were present between patients with HCC compared to the patients who did not have HCC.

Plasma Platelets Plasma Platelets 2050 80 *** *** 70

2000 n

i

t

) l

750 c 25

A

m

/ 600

g 20

g

u n

450 /

(

g 15

300 p

G

F

150 F 10

E

G V

120 E

V 5 60 0 0 Non-HCC HCC Controls Non-HCC HCC Controls

1000 20 *** ***

700 n

i

)

t

l c

m 400 15

/

A

g g

n 100

(

u /

100 10

g

F

p G

80 F

60 F b G 5

40 F

20 b 0 0 Non-HCC HCC Controls Non-HCC HCC Controls

500 **

** 500

n

) i

l 400

t 400

c

m

/

A

g 300 g

n 300

(

u

/ F

200 g

p G

200

D

F P 100 G

D 100 P 0 0 Non-HCC HCC Controls Non-HCC HCC Controls

20 *** *** 50 *** ***

n

i t

) 40 l

15 c

A

m

/

g

g 30

u

n /

( 10

g F

p 20

G

F H

5 G

10 H

0 0 Non-HCC HCC Controls Non-HCC HCC Controls

Figure 1: Plasma and intraplatelet levels of angiogenesis inhibitory proteins in patients with or without HCC and control subjects. Horizontal bars indicate median. ** P < 0.01, *** P < 0.001, all v.s. control

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______96

Platelet levels of VEGF, bFGF and HGF were all higher in patients compared to controls, but no differences were detected between patients with and without HCC. Platelet levels of PDGF were not different between patients and controls. Plasma and platelet levels of all proteins were not different between HCC patients who were treatment-naıëve and patients who had been treated for their HCC prior to inclusion in the study (all with P>0.2).

LEVELS OF ANTIANGIOGENIC PROTEINS IN PLASMA AND PLATELETS FROM PATIENTS WITH HEPATITIS-RELATED CIRRHOSIS WITH AND WITHOUT HCC Figure 2 shows levels of antiangiogenic proteins in plasma and platelets from patients and controls. Patients had decreased plasma levels of PF4 and TSP-1 compared to controls, and plasma levels of endostatin were similar between patients and controls. No differences were present in plasma levels of antiangiogenic proteins between patients with and without HCC. Platelet levels of PF4 and TSP-1 were similar between patients and controls, and platelet endostatin levels were slightly, but significantly, increased in patients patients compared to controls. No differences in platelet antiangiogenic proteins were detected between patients who did and those who did not have HCC. Plasma and platelet levels of all proteins were not different between HCC patients who were treatment-naıëve and patients who had been treated for their HCC prior to inclusion in the study (all with P>0.2).

DISCUSSION

In this study, we found no differences in platelet and plasma levels of seven angiogenic proteins between patients that had hepatitis-related cirrhosis and patients that had hepatitis-related cirrhosis and HCC. We did, however, detect differences in platelet or plasma levels of several of the proteins studied between patients and healthy controls. We hypothesized to find increased intraplatelet levels of the angiogenic markers based on studies showing increased intraplatelet levels of these proteins in patients with colorectal cancer27 and on studies showing that intraplatelet levels of these proteins increase already at very small tumor sizes in mice.25,26 The clear absence of a difference between intraplatelet levels of angiogenic proteins in patients with HCC and those without indicates that intraplatelet levels of angiogenic proteins cannot be used as a biomarker for the presence of cancer in all types of cancer. Nevertheless, the small cohort size, heterogeneity of the patient population and the fact that the majority of HCC patients were not ‘‘treatment- naıëve’’ does not allow the firm conclusion that no differences between patients with cirrhosis and those with cirrhosis and HCC exist.

The apparent absence of a difference between angiogenic proteins in patients with and without HCC may be explained by the angiogenic response occurring in patients with cirrhosis who have not (yet) developed HCC. It has been well established that the progression of fibrosis to cirrhosis is associated with a profound angiogenic response

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______97

Plasma Platelets

200 100 *** ***

n

i

t

)

c l

A 80

m

/ 150

g

g

u

/

n

( g

60

p

n

i

t 100

n

i

a t

t 40

a

s

t

o

s

d o

n 50

d 20

E

n E 0 0 Non-HCC HCC Controls Non-HCC HCC Controls

*** *** 2500 1000 900

2000 n

i 350

t

) l

c 300

A

m

/ 1500 g

g 250

u

n

/

(

g 200 4

1000 p

F

4 150

P F

500 P 100 50 0 0 Non-HCC HCC Controls Non-HCC HCC Controls

4000 ** ** 50

3000

n i

t 40

) 2000

l

c

A

m

/ 2000 g

g 30

u

n

/

(

1500 g

1

p -

20

P

1 -

S 1000

P T S 10 500 T 0 0 Non-HCC HCC Controls Non-HCC HCC Controls

Figure 2: Plasma and intraplatelet levels of angiogenesis inhibitory proteins in patients with or without HCC and control subjects. Horizontal bars indicate median. ** P < 0.01, *** P < 0.001, all v.s. control. initiated by multiple mechanisms including changes in blood flow, hypoxia and inflammation.29,30 Importantly, since our data show that plasma levels of the angiogenic proteins are not elevated, it is likely that the angiogenic proteins that are induced in cirrhosis are taken up by platelets, similar to the uptake of these proteins in patients with cancer. Although it has not been established whether platelets play a role in the cirrhosis associated angiogenic response, our data suggest that intraplatelet angiogenic proteins are elevated in cirrhosis. Given the role of platelets in inflammation and their multiple roles in liver diseases,31 it may be that platelets are active players in disease progression by delivery of angiogenic proteins within the liver.

The additional angiogenic response that occurs when patients with cirrhosis develop HCC may thus simply be too small to result in additional increases in intraplatelet levels of

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______98

angiogenic proteins. Our study population consisted of patients with cirrhosis who were followed up on regular basis with active HCC surveillance programs. The median cumulative tumor size in our cohort was 40 mm, and none of the patients had extrahepatic manifestations or macroscopic vasoinvasion suggesting that the tumors were detected early due to the regular follow up. The levels of angiogenic factors produced by these relatively small tumors may also have been too small to result in detectable increases in intraplatelet levels of these proteins. Our results are, however, at variance with previously published data showing serum levels of HGF to be elevated in patients with cirrhosis and HCC compared to patients with cirrhosis alone.32,33

Plasma levels of the angiogenic proteins were not different between patients and controls, except for levels of PDGF, PF4 and TSP-1, which were all decreased in patients compared to controls. All three proteins are synthesized by megakaryocytes, and the decreased plasma levels in patients with cirrhosis may simply be a reflection of the decreased circulating platelet count in cirrhosis. Indeed, when plasma levels of PDGF, PF4 and TSP-1 were normalized for the platelet count, the differences between patients and controls fully disappeared (data not shown).

In conclusion, intraplatelet levels of some angiogenic proteins are elevated in cirrhosis, but do not distinguish between patients with cirrhosis who do and who do not have HCC. Intraplatelet levels of angiogenic proteins thus cannot be used as biomarkers of HCC in patients with cirrhosis. Whether platelet levels of angiogenic proteins are prognostic in patients with HCC in the absence of cirrhosis requires further study.

ACKNOWLEDGEMENTS We thank Dr. Claudia M. Radu for expert laboratory assistance.

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______99

REFERENCES

1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011;61:69-90. 2. Vivarelli M, Montalti R, Risaliti A. Multimodal treatment of hepatocellular carcinoma on cirrhosis: An update. World J Gastroenterol 2013;19:7316-26. 3. Chan SC. Liver Transplantation for Hepatocellular Carcinoma. Liver Cancer 2013;2:338-44. 4. Tzanis D, Shivathirthan N, Laurent A, et al. European experience of laparoscopic major hepatectomy. J Hepatobiliary Pancreat Sci 2013;20:120-4. 5. Soubrane O, Goumard C, Laurent A, et al. Laparoscopic resection of hepatocellular carcinoma: a French survey in 351 patients. HPB (Oxford) 2013;. 6. Gimbrone MA,Jr, Leapman SB, Cotran RS, Folkman J. Tumor dormancy in vivo by prevention of neovascularization. J Exp Med 1972;136:261-76. 7. Folkman J. Angiogenesis: an organizing principle for drug discovery? Nat Rev Drug Discov 2007;6:273-86. 8. Yao DF, Wu XH, Zhu Y, et al. Quantitative analysis of vascular endothelial growth factor, microvascular density and their clinicopathologic features in human hepatocellular carcinoma. Hepatobiliary Pancreat Dis Int 2005;4:220-6. 9. El-Assal ON, Yamanoi A, Soda Y, et al. Clinical significance of microvessel density and vascular endothelial growth factor expression in hepatocellular carcinoma and surrounding liver: possible involvement of vascular endothelial growth factor in the angiogenesis of cirrhotic liver. Hepatology 1998;27:1554-62. 10. Pang R, Poon RT. Angiogenesis and antiangiogenic therapy in hepatocellular carcinoma. Cancer Lett 2006;242:151-67. 11. Almog N, Klement GL. Platelet proteome and tumor dormancy: can platelets content serve as predictive biomarkers for exit of tumors from dormancy? Cancers (Basel) 2010;2:842-58. 12. Zhong C, Wei W, Su XK, Li HD, Xu FB, Guo RP. Serum and tissue vascular endothelial growth factor predicts prognosis in hepatocellular carcinoma patients after partial liver resection. Hepatogastroenterology 2012;59:93-7. 13. Poon RT, Ng IO, Lau C, et al. Serum vascular endothelial growth factor predicts venous invasion in hepatocellular carcinoma: a prospective study. Ann Surg 2001;233:227-35. 14. Kim SJ, Choi IK, Park KH, et al. Serum vascular endothelial growth factor per platelet count in hepatocellular carcinoma: correlations with clinical parameters and survival. Jpn J Clin Oncol 2004;34:184-90. 15. Banks RE, Forbes MA, Kinsey SE, et al. Release of the angiogenic cytokine vascular endothelial growth factor (VEGF) from platelets: significance for VEGF measurements and cancer biology. Br J Cancer 1998;77:956-64. 16. Dittadi R, Meo S, Fabris F, et al. Validation of blood collection procedures for the determination of circulating vascular endothelial growth factor (VEGF) in different blood compartments. Int J Biol Markers 2001;16:87-96. 17. Jelkmann W. Pitfalls in the measurement of circulating vascular endothelial growth factor. Clin Chem 2001;47:617-23. 18. Peterson JE, Zurakowski D, Italiano JE,Jr, et al. Normal ranges of angiogenesis regulatory proteins in human platelets. Am J Hematol 2010;85:487-93.

P L A T E L E T A N G I O G E N I C F A C T O R S I N L I V E R C A N C E R

______100

19. Ho-Tin-Noe B, Goerge T, Wagner DD. Platelets: guardians of tumor vasculature. Cancer Res 2009;69:5623-6. 20. Cho MS, Bottsford-Miller J, Vasquez HG, et al. Platelets increase the proliferation of ovarian cancer cells. Blood 2012;120:4869-72. 21. Nash GF, Turner LF, Scully MF, Kakkar AK. Platelets and cancer. Lancet Oncol 2002;3:425- 30. 22. Ho-Tin-Noe B, Goerge T, Cifuni SM, Duerschmied D, Wagner DD. Platelet granule secretion continuously prevents intratumor hemorrhage. Cancer Res 2008;68:6851-8. 23. Karpatkin S, Pearlstein E, Salk PL, Yogeeswaran G. Role of platelets in tumor cell metastases. Ann N Y Acad Sci 1981;370:101-18. 24. Erpenbeck L, Schon MP. Deadly allies: the fatal interplay between platelets and metastasizing cancer cells. Blood 2010;115:3427-36. 25. Cervi D, Yip TT, Bhattacharya N, et al. Platelet-associated PF-4 as a biomarker of early tumor growth. Blood 2008;111:1201-7. 26. Klement GL, Yip TT, Cassiola F, et al. Platelets actively sequester angiogenesis regulators. Blood 2009;113:2835-42. 27. Peterson JE, Zurakowski D, Italiano JE,Jr, et al. VEGF, PF4 and PDGF are elevated in platelets of colorectal cancer patients. Angiogenesis 2012;15:265-73. 28. Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973;60:646-9. 29. Coulon S, Heindryckx F, Geerts A, Van Steenkiste C, Colle I, Van Vlierberghe H. Angiogenesis in chronic liver disease and its complications. Liver Int 2011;31:146-62. 30. Fernandez M, Semela D, Bruix J, Colle I, Pinzani M, Bosch J. Angiogenesis in liver disease. J Hepatol 2009;50:604-20. 31. Lisman T, Porte RJ. The role of platelets in liver inflammation and regeneration. Semin Thromb Hemost 2010;36:170-4. 32. Yamagamim H, Moriyama M, Matsumura H, et al. Serum concentrations of human hepatocyte growth factor is a useful indicator for predicting the occurrence of hepatocellular carcinomas in C-viral chronic liver diseases. Cancer 2002;95:824-34. 33. Costantini S, Capone F, Maio P, et al. Cancer biomarker profiling in patients with chronic hepatitis C virus, liver cirrhosis and hepatocellular carcinoma. Oncol Rep 2013;29:2163-8.

CHAPTER 8

ELEVATED SERUM GAMMA GLUTAMYL TRANSPEPTIDASE AFTER LIVER TRANSPLANTATION IS ASSOCIATED WITH BETTER SURVIVAL

EDRIS M ALKOZAI TON LISMAN ROBERT J PORTE MAARTEN W NIJSTEN

PUBLISHED IN: F1000 RESEARCH 2014, 3:85

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______104

ABSTRACT

BACKGROUND: Gamma glutamyl transpeptidase (GGT) is a membrane bound enzyme that plays a key role in the synthesis of the antioxidant glutathione. Epidemiological studies have linked high GGT with an increased risk of morbidity and cardiovascular mortality. In contrast, GGT is usually elevated in liver transplant recipients that experience good outcomes.

AIMS: To study if and how GGT is correlated with mortality following liver transplantation. Methods: We analyzed the prognostic relevance of serum GGT levels during the early and late postoperative period after liver transplantation in 522 consecutive adults. We also studied alanine aminotransferase, aspartate aminotransferase, and total bilirubin levels.

RESULTS: Early after transplantation, the peak median (interquartile range) GGT levels were significantly higher in patients who survived more than 90 days compared to non-survivors: 293 (178-464) vs. 172 (84-239) U/l, p<0.0001. In contrast, late after transplantation, GGT levels were significantly lower in patients who survived more than 5 years than those who did not (p<0.01). The pattern of GGT levels also differed from those of alanine aminotransferase, aspartate aminotransferase, and total bilirubin early after transplantation, while these patterns were congruent late after transplantation. Kaplan- Meier survival analysis showed that early after transplantation the higher the GGT levels, the better the 90-day survival (p<0.001). In contrast, late after transplantation, higher GGT levels were associated with a lower 5-year survival (p<0.001).

CONCLUSIONS: These paradoxical findings may be explained by the time-dependent role of GGT in glutathione metabolism. Immediate postoperative elevation of GGT may indicate a physiological systemic response to oxidative stress induced by ischemia-reperfusion injury and liver transplantation while chronic elevation reflects a pathological response.

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______105

INTRODUCTION

Gamma glutamyl transpeptidase (GGT) is a membrane-bound enzyme that is essential for the synthesis of glutathione (GSH), a key antioxidant.1 In clinical practice elevated serum GGT is generally used as an indicator of liver disease, such as biliary obstruction, alcohol consumption, and exposure to certain medical drugs.1 Recently, several epidemiological studies have shown that a higher serum GGT level, even within the normal range, is associated with cardiovascular risk factors such as hypertension, hypertriglyceridemia, obesity, type 2 diabetes mellitus and stroke, as well as certain types of cancer.2-10 In contrast to these studies, we observed that after surgery for ruptured abdominal aortic aneurysm 11 or after liver resection,12 GGT is transiently increased in patients who had a good outcome. In these short-term observational studies GGT level was inversely related to other liver laboratory parameters such as aspartate aminotransferase (ALT), alanine aminotransferase (AST) as well as total bilirubin (TBI).11,12 We observe that in the early postoperative period after a liver transplantation (LT) a transient gradual increase in GGT also occurs. How early and late postoperative changes in serum GGT are related to survival is not known.

Here we present a study in which we assessed the relationship between early (postoperative day seven) elevated GGT levels with early and late survival (i.e. survival within the 90 days and 5 years post-LT, respectively). We also evaluated the relationship between late (six months postoperatively) elevated GGT levels with late survival. In addition, we studied the early and late post-LT kinetics of GGT, AST, ALT, and TBI in patients who survived more than 90-days and in patients who did not. Likewise, these kinetics were compared with long-term survival.

MATERIALS AND METHODS

STUDY POPULATION We conducted a single center cohort study on 522 first liver transplant patients. All LTs performed between January 1990 and August 2009 were included; excluded were pediatric patients (age <17 years), second or subsequent LT, and first LT patients who underwent a re-LT within 90 days of their first LT. Since obstructive mechanisms such as non- anastomotic stricture (NAS), acute graft rejection, and cholestatic disorders might influence GGT and TBI levels, we also specifically repeated our analysis with and without such patients. This study was performed in accordance with Dutch legislation and the local ethical committee guidelines.

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______106

STUDY VARIABLES Patient characteristics and variables related to the perioperative management and the surgical procedure were obtained from a prospectively collected database. These included age, sex, body mass index, Karnofsky score, indication for LT, preoperative MELD (Model for End-Stage Liver Disease) score (calculated from preoperative laboratory measurements), length of hospital stay, cold ischemia time, warm ischemia time, duration of operation, combined transplant (kidney or lung), acute rejection, graft type, the number of units of allogeneic and autologous red blood cell units (RBC with 1 U containing 300 ml) and fresh frozen plasma (FFP with 1 U containing 250 ml), donor type, type of venous and bile duct anastomosis, and NAS within 90 days and within one year. When necessary, the hospital files were reviewed to complete all relevant clinical parameters.

THE KINETICS OF SERUM GGT AND OTHER LIVER FUNCTION VARIABLES EARLY AND LATE POST-LT We studied the levels of serum GGT and other liver variables postoperatively in two ways. Early postoperatively, up to postoperative day (POD) 30, we studied the levels of GGT (reference values; 0-40 U/l), ALT (0-45 U/l), AST (0-40 U/l), and TBI (0-17 µmol/l) over time in patients who survived more than 90 days after LT compared to those who did not. Late postoperatively (i.e. 90 days and beyond), we evaluated the levels of these variables at three months, six months, and one year in patients who survived more than five years compared to those patients who did not.

SURVIVAL ANALYSIS AND ELEVATED GGT EARLY AND LATE FOLLOWING LT To evaluate the clinical relevance of early and late elevated GGT levels, we generated tertiles of low, intermediate, and high GGT levels based on equal percentiles. GGT levels at POD 7 were used to study the relationship between early elevated GGT levels and both 90- day and five-year survival. GGT levels at six months following LT were used to study the relationship of late elevated GGT with five-year survival. The last observation date for the status of patient survival for the study cohort was August 23, 2012.

STATISTICAL ANALYSIS Statistical analyses were performed using the statistical software package SPSS 20 (IBM SPSS, Chicago, IL). Categorical variables are shown as numbers and percentages. Continuous variables are presented as means with standard deviation (SD) or as medians with interquartile range (IQR) based on their distribution. Continuous variables that were not normally distributed were compared using the Mann Whitney U test. We studied early LT mortality based on GGT levels at POD 7 as a categorical variable, using tertiles (low, intermediate, and high). Similarly, we assessed the late LT mortality based on GGT levels at six months post-LT using tertiles. Patient survival was analyzed with Kaplan-Meier analysis and the differences between the groups were assessed with the log-rank test. A P <0.05 was considered statistically significant.

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______107

Table1. Characteristics of study population Demographic characteristics Total n=522 Recipient variables Gender, male 283 (54%) Age, years, median (IQR) 48 (37-56) BMI, mean (SD)* 25 (5) Karnofsky score, median (IQR) 60 (30-70) MELD, median (IQR)* 17 (13-24) Indication for LT Post necrotic cirrhosis 254 (49%) Cholestatic liver diseases 155 (30%) Acute liver failure 38 (7%) Metabolic disease 51 (10%) Miscellaneous 24 (5%) In hospital stay, days (IQR) 30 (21-46) ICU stay, days (IQR) 3 (2-7) Transplantation variables Operation length (minutes), mean (SD) 581 (119) WIT (minutes), mean (SD) 53 (14) CIT (minutes), mean (SD) 573 (192) Combined transplant (kidney or lung) 18 (3%) Acute 90 day rejection 178 (34%) Mild, not treated 75 (42%) Moderate severe, treated 103 (58%) Graft type* Full size 503 (97%) Split or reduced size 18 (3%) ABO compatibility* Identical 489 (95%) Compatible 28 (5%) Blood loss (liters), median (IQR) 5.0 (2.1-8.5) RBC transfusion (unit=300 ml), median (IQR) 6 (2-11) Donor type* Heart beating 435 (91%) Non-heart beating 38 (8%) Living donor 3 (1%) Venous anastomosis Piggyback 302 (58%) Classic 220 (42%) Bile duct anastomosis Duct to duct 456 (87%) Rou-x-en Y 61 (12%) Duct-duodenostomy 5 (1%) Non-anastomotic stricture, 90-day, yes 30 (6%) *Some variables were not available for all patients. CIT, cold ischemia time; ICU, intensive care unit; IQR, interquartile range; INR, international normalized ratio; SD, standard deviation; WIT, warm ischemia time.

RESULTS

STUDY POPULATION We performed a total of 968 consecutive LTs in our center between January 1990 and August 2009. After excluding pediatric LTs (age <17 yr; n=290), patients who were re- transplanted within the 90 days of their first LT (n=39), second or subsequent LTs (n=101), patients with a lack of follow up data (n=11), and patients that died intraoperatively (n=5)

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______108

due to brain death, cardiac failure, or Table 2. Causes of mortality within the 90 uncontrollable bleeding, 522 patients were days after LT. included in our analyses. The median age Frequency was 48 years (37-56), 54% of the patients Cause of death n=43 (%) were males, mean (SD) BMI was 24.8 (± Sepsis 16 (37%) 5.3), median Karnofsky score was 60 (30- Multi organ failure 6 (14%) 70) and median MELD score was 17 (13- Brain death (metabolic and 4 (9%) 24) for the study population. Indications for hepatic encephalopathy) LT were post necrotic cirrhosis (49%), Transplant failure 3 (7%) cholestatic liver disease (30%), metabolic Pulmonary embolism 3 (7%) Bleeding 3 (7%) disease (10%), acute liver failure (7%), and Cardiac failure 3 (7%) miscellaneous (5%). Patient characteristics Respiratory failure 1 (2%) and the surgical variables of the entire Graft vs. host disease 1 (2%) group of 522 patients are summarized in Unspecified 3 (7%) Table 1.

MORTALITY RATES AND MAJOR CAUSES OF EARLY MORTALITY The overall mortality within 90 days, one year, and five years for the study cohort was 8%, 12%, and 21%, respectively. Sepsis was the major cause of mortality (37%) within 90 days followed by multi organ failure (14%), and brain death (9%). Table 2 details all causes of mortality within the first 90 days.

Figur 1: Evaluation of early post-LT laboratory variables (i.e. 0 - 30 POD) in early mortality. Curves represent patients who survived more than 90 days (gray) and those who did not (black). Median values are shown.*p < 0.05; **p ≤ 0.001. GGT, gamma glutamyl transpeptidase; ALT, aspartate aminotransferase; AST, alanine aminotransferase; TBI, total bilirubin.

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______109

EARLY POST-LT GGT AND OTHER LIVER VARIABLES AND EARLY MORTALITY Early postoperative laboratory variables are shown in Figure 1. Postoperatively, GGT levels increased gradually, reaching a maximum at POD 9 and decreased thereafter. Notably, the increase in GGT levels was significantly more pronounced, i.e. deviated more from the normal range, in patients who survived more than 90 days, as compared to those who did not: 297 (178-464) vs. 172 (69-271) U/l, p<0.0001, respectively. This pattern was different from that of postoperative levels of TBI, AST, and ALT. TBI was consistently and significantly lower in patients who survived more than 90 days following LT as compared to those who did not. AST and ALT levels increased rapidly until POD 1 and POD 2, respectively, followed by rapid normalization thereafter. Contrary to GGT, the peak levels of AST and ALT were significantly lower in patients who survived more than 90 days following LT as compared to those who did not: AST 659 (326-1267) vs. 1201 (451-1990) U/l, p=0.01, and ALT 527 (280-1080) vs. 1082 (529-2631) U/l, p=0.001, respectively. Thirty patients developed NAS within the 90 days post-LT. Since these patients may present with abnormal high TBI and GGT levels, we repeated the analysis with exclusion of these 30 patients with no significant effect on the GGT levels (graphical representation not shown). Also exclusion of patients who were treated for developing an acute rejection (n=103) and those who underwent LT for cholestatic liver disease (n=155) did not significantly affect the outcomes (graphical representation not shown).

LATE POST-LT GGT AND OTHER LIVER VARIABLES AND MORTALITY Late postoperative laboratory variables are shown in Figure 2. We studied the changes in GGT levels over time in patients who survived more than five years as compared to those who did not survive. Compared to patients who died within five years post-LT, those who survived more than five years had significantly lower GGT at three months post-LT; 95 (42- 244) vs. 212 (92-400) U/l, p=0.001, six months post-LT; 70 (31-189) vs. 281 (103-438) U/l, p<0.001, and 1 year post-LT; 57 (25-153) vs. 124 (45-431) U/l, p=0.003, respectively. Notably, late post-LT the GGT levels showed the same patterns as TBI, AST, and ALT, i.e. higher levels in non-survivors.

KAPLAN-MEIER SURVIVAL ANALYSIS We also studied the clinical relevance of early versus late elevated GGT levels with Kaplan- Meier survival analysis. Figure 3 plots the overall 90-day and 5-year overall survival for the study cohort based on GGT-tertiles. A high GGT level at POD 7 was significantly associated with better early survival following LT (Figure 3A). The overall 90-day survival was 98% for high GGT (≥ 351 U/l), compared to 94% for intermediate GGT levels (188 and 350 U/l), and 87% for the low GGT (≤ 187 U/l), at POD 7 (Figure 3A). Similarly, five-year overall survival was 86%, 83%, and 73% for high, intermediate, and low GGT at POD 7 (p =0.003; Figure 3B), respectively. Remarkably, the differences in five- year survival mainly developed during

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______110

the first three months post-LT with almost no difference in survival curves thereafter (Figure 3B). In sharp contrast with early GGT, a high GGT level six months post-LT was associated with lower five-year survival (Figure 3C). The overall survival within five years was 71% for elevated GGT (> 163 U/l), compared to 91% for intermediate GGT levels (44 and 163 U/l), and 93% for the low GGT (< 43 U/l), p<0.001.

Figure 2: Evaluation of three, six, and twelve months post-LT laboratory variables in late mortality. Bars represent patients who survived more than five years (gray) and those who did not (black). Median values are shown. *p < 0.05; **p ≤ 0.001. GGT, gamma glutamyl transpeptidase; ALT, aspartate aminotransferase; AST, alanine aminotransferase; TBI, total bilirubin.

DISCUSSION

In this study, we evaluated the changes in GGT over time following liver transplantation (LT) and the clinical relevance of these changes for early and late survival. We found that a transiently elevated GGT early after LT was associated with increased survival rates within the first 90-days. In contrast, late elevation of GGT was associated with decreased five-year survival following LT. Although the early GGT elevations was also associated with five-year survival, this difference mainly developed during the first 90-days post-LT. This peculiar effect was not observed for TBI, AST, and ALT since higher levels for these parameters at POD 7 and six months were associated with increased mortality at both 90 days and 5 years after LT. To our knowledge, this is the first study showing the short and long term kinetics of GGT and the clinical relevance of an early elevated serum GGT in LT recipients. Previously, we have reported improved outcome in patients with significantly increased levels of GGT in the early post-operative period following liver resection12 and surgical repair of a ruptured abdominal aortic aneurysm.11 However, those studies were not designed to address changes in GGT

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______111

progression over time. While we acknowledge that association does not necessarily indicate causation, these data support the hypothesis that high GGT in an early post-LT setting may be a marker of some protective process. Although the precise mechanism responsible for an elevated serum GGT early post-LT is yet to be determined, experimental studies have demonstrated that cellular GGT modulates crucial redox-sensitive functions, such as antioxidant and antitoxic defenses, cellular proliferation, and apoptotic balance.13

Cellular GGT is a key enzyme in the gamma- glutamyl cycle resulting in production of intracellular GSH,14-16 an important antioxidant agent that protects the cells against reactive oxygen species (ROS).17-19 GSH has been shown to protect the liver against ischemia reperfusion injury in animal models.16,20,21 Hepatic ischemia can cause elevation of serum GGT with peak blood levels within 20 and 30 hours after restoration of hepatic arterial blood flow.18,22 Reperfusion is associated with a surge of ROS, which may overwhelm host natural antioxidant defenses.21 The oxidative stress from the ROS formed after reperfusion may lead to increased cellular death by damaging membrane lipids through peroxidation, disrupting normal enzymatic activities, and diminishing Figure 2: Kaplan–Meier analysis showing 23 mitochondrial oxidative metabolism. Cardin the opposite relationships of early and and colleagues24 studied oxidative stress in late GGT levels with early and late patients with chronic hepatitis C virus infection. mortality. Panels A and B represent survival analysis post-LT in relation to GGT Surprisingly, the authors observed an association tertiles at POD 7, within 90 days and five between GGT and 8-hydroxydeoxyguanosine (8- years respectively (P=0.003). Panel C OHdG), a marker of oxidative DNA damage. demonstrates five-year survival in relation to GGT tertiles at six months post-LT Patients who had a high level of 8-OHdG had (p<0.001). Curves represent low (thick significantly higher GGT levels but normal ALT black), intermediate (thin black), and high GGT (gray). levels.24 Thus, a transient increase in GGT level

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______112

post-LT may reflect the host compensatory mechanism against oxidative stress and toxic metabolites generated by hypoxia, reperfusion, and surgical stress.21 Therefore, an increased GGT early after LT may reflect the ability of the host to initiate an appropriate systemic response.

Elevated serum GGT in the early post-LT period might also be a marker of better liver regeneration. Eisenbach and colleagues,25 showed that an early increase in serum GGT after LT was associated with a better outcome and the authors reasoned that this rise could be due to liver regeneration. Although the liver might regenerate to some extent after LT, there is no conclusive evidence to support this hypothesis. As we mentioned earlier, we observed a transient increase in serum GGT levels in patients who survived a surgical repair for a ruptured abdominal aortic aneurysm.11 In the latter group, it is less likely that significant liver regeneration occurs.

Furthermore, elevated serum GGT early post-LT may also be an indicator of better early secretory function of the liver. GGT in part comes from the surface of the bile ducts and is released from the anchor, which attaches this enzyme to the cell surface, by bile acids in bile. Bile acid/phospholipid ratio in bile has shown to be elevated after LT.26 Thus elevated serum GGT immediately after transplantation may be an indicator or an active bile flow. Hence, it may protect hepatocytes from cytotoxic bile acids.

Contrary to early elevated GGT, but in line with published literature,2-10 we observed that a late (i.e. six months post-LT) elevated GGT was significantly associated with decreased survival within the 5-years following LT. Although the finding that normal GGT levels in the late post-operative period is predictive of good outcomes is obvious and intuitive, the contrasting influences of GGT levels between early and late post-LT periods on survival may be compatible with the physiologic function of intracellular GGT. Notably, at three months, six months, and one year post-LT, the relative increase in serum GGT was two to four times higher in patients who did not survive for more than five years compared to those who did survive. This proportion was much higher than that of AST, ALT and TBI, which might imply that an elevated serum GGT is not only a marker of harm to the liver but it could be seen as a systemic response to harmful environmental factors. Indeed, in two studies, Lee and colleagues27,28 postulated that serum GGT in the general population might be a marker of increased exposure to environmental stress, internal xenobiotics, or other unknown factors that cause oxidative stress in the long run.

To avoid possible bias we performed our analysis after excluding obstructive mechanisms such as NAS, cholestatic disorders, and acute graft rejection early postoperatively. Exclusion of these cases did not change our results significantly, suggesting that the elevation in serum GGT early post-LT is independent of obstructive disorders. A practical clinical

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______113

consequence of our findings may be that care providers in hospitals should realize that an abnormally high GGT early post-LT is not a cause for alarm or specific diagnostic procedures. In fact, a GGT activity four to five times above the normal range during the second post- operative week might even be considered beneficial.

We acknowledge some considerable limitations in this study. First, due to the retrospective design of the study we can only identify association rather than causation. This could only be established by specific (intervention) studies that measure the interaction between serum GGT and ROS markers post-LT. However, there is a strong correlation between GGT and oxidative DNA damage in cirrhotic patients.24 Next, we cannot entirely exclude that liver regeneration plays a role in the early post-LT rise of GGT levels since GSH and to some extent GGT are mainly produced by the liver.1 Hence a high serum GGT in patients who survived more than 90 days can also be a reflection of a well-functioning graft in LT patients. It will be important to understand the relationship of serum GGT and cellular GGT in the period immediately after surgery. Besides, cumulative evidence suggests that there is a relationship between the induction and release of ROS and ischemia reperfusion injury after other types of abdominal surgery.18,21,22

In conclusion, an elevated GGT level early after LT was associated with a better short-term outcome. However, chronically elevated GGT was associated with poor long-term outcome in the outpatient setting after LT. This peculiar switch in the prognostic meaning of GGT may result from the superposition of several mechanisms. Apparently, higher expression of GGT reflects some protective process in the acute phase but reflects chronic damage over the long term.

Acknowledgments: We thank J. T. Bottema for providing us with additional data.

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______114

REFERENCES

1. Whitfield JB. Gamma glutamyl transferase. Crit Rev Clin Lab Sci 2001;38:263-355. 2. Fraser A, Harris R, Sattar N, Ebrahim S, Davey Smith G, Lawlor DA. Alanine aminotransferase, gamma-glutamyltransferase, and incident diabetes: the British Women's Heart and Health Study and meta-analysis. Diabetes Care 2009;32:741-50. 3. Kazemi-Shirazi L, Endler G, Winkler S, Schickbauer T, Wagner O, Marsik C. Gamma glutamyltransferase and long-term survival: is it just the liver? Clin Chem 2007;53:940-6. 4. Kristenson H, Ohrn J, Hood B. Convictions for drunkenness or drunken driving, sick absenteeism, and morbidity in middle-aged males with different levels of serum gamma- glutamyltransferase. Prev Med 1982;11:403-16. 5. Strasak AM, Kelleher CC, Klenk J, et al. Longitudinal change in serum gamma- glutamyltransferase and cardiovascular disease mortality: a prospective population-based study in 76,113 Austrian adults. Arterioscler Thromb Vasc Biol 2008;28:1857-65. 6. Peterson B, Trell E, Kristensson H, Fex G, Yettra M, Hood B. Comparison of gamma- glutamyltransferase and other health screening tests in average middle-aged males, heavy drinkers and alcohol non-users. Scand J Clin Lab Invest 1983;43:141-9. 7. Lee DS, Evans JC, Robins SJ, et al. Gamma glutamyl transferase and metabolic syndrome, cardiovascular disease, and mortality risk: the Framingham Heart Study. Arterioscler Thromb Vasc Biol 2007;27:127-33. 8. Ruttmann E, Brant LJ, Concin H, et al. Gamma-glutamyltransferase as a risk factor for cardiovascular disease mortality: an epidemiological investigation in a cohort of 163,944 Austrian adults. Circulation 2005;112:2130-7. 9. Breitling LP, Claessen H, Drath C, Arndt V, Brenner H. Gamma-glutamyltransferase, general and cause-specific mortality in 19,000 construction workers followed over 20 years. J Hepatol 2011;55:594-601. 10. Kengne AP, Czernichow S, Stamatakis E, Hamer M, Batty GD. Gamma-glutamyltransferase and risk of cardiovascular disease mortality in people with and without diabetes: Pooling of three British Health Surveys. J Hepatol 2012;57:1083-9. 11. Haveman JW, Zeebregts CJ, Verhoeven EL, et al. Changes in laboratory values and their relationship with time after rupture of an abdominal aortic aneurysm. Surg Today 2008;38:1091-101. 12. Alkozai EM, Nijsten MW, de Jong KP, et al. Immediate postoperative low platelet count is associated with delayed liver function recovery after partial liver resection. Ann Surg 2010;251:300-6. 13. Kobayashi H, Nonami T, Kurokawa T, et al. Changes in the glutathione redox system during ischemia and reperfusion in rat liver. Scand J Gastroenterol 1992;27:711-6. 14. Moriya S, Nagata S, Yokoyama H, et al. Expression of gamma-glutamyl transpeptidase mRNA after depletion of glutathione in rat liver. Alcohol Alcohol Suppl 1994;29:107-11. 15. Rajpert-De Meyts E, Shi M, Chang M, et al. Transfection with gamma-glutamyl transpeptidase enhances recovery from glutathione depletion using extracellular glutathione. Toxicol Appl Pharmacol 1992;114:56-62. 16. Stein HJ, Oosthuizen MM, Hinder RA, Lamprechts H. Oxygen free radicals and glutathione in hepatic ischemia/reperfusion injury. J Surg Res 1991;50:398-402.

G G T P O S T - L I V E R T R A N S P L A N T A T I O N

______115

17. Stenius U, Rubin K, Gullberg D, Hogberg J. gamma-Glutamyltranspeptidase-positive rat hepatocytes are protected from GSH depletion, oxidative stress and reversible alterations of collagen receptors. Carcinogenesis 1990;11:69-73. 18. Stastny F, Lisy V, Tomasova H, Trojan S. Effects of short-term and prolonged aerogenic hypoxia on gamma-glutamyl transpeptidase activity in the brain, liver, and biological fluids of young rats. Neurochem Res 1985;10:819-28. 19. Gupta S, Rajvanshi P, Malhi H, et al. Cell transplantation causes loss of gap junctions and activates GGT expression permanently in host liver. Am J Physiol Gastrointest Liver Physiol 2000;279:G815-26. 20. Schauer RJ, Gerbes AL, Vonier D, et al. Glutathione protects the rat liver against reperfusion injury after prolonged warm ischemia. Ann Surg 2004;239:220-31. 21. Zhang W, Wang M, Xie HY, et al. Role of reactive oxygen species in mediating hepatic ischemia-reperfusion injury and its therapeutic applications in liver transplantation. Transplant Proc 2007;39:1332-7. 22. Liu W, Schob O, Pugmire JE, et al. Glycohydrolases as markers of hepatic ischemia- reperfusion injury and recovery. Hepatology 1996;24:157-62. 23. Nonn L, Berggren M, Powis G. Increased expression of mitochondrial peroxiredoxin-3 (thioredoxin peroxidase-2) protects cancer cells against hypoxia and drug-induced hydrogen peroxide-dependent apoptosis. Mol Cancer Res 2003;1:682-9. 24. Cardin R, Saccoccio G, Masutti F, Bellentani S, Farinati F, Tiribelli C. DNA oxidative damage in leukocytes correlates with the severity of HCV-related liver disease: validation in an open population study. J Hepatol 2001;34:587-92. 25. Eisenbach C, Encke J, Merle U, et al. An early increase in gamma glutamyltranspeptidase and low aspartate aminotransferase peak values are associated with superior outcomes after orthotopic liver transplantation. Transplant Proc 2009;41:1727-30. 26. Op den Dries S, Sutton ME, Lisman T, Porte RJ. Protection of bile ducts in liver transplantation: looking beyond ischemia. Transplantation 2011;92:373-9. 27. Lee DH, Blomhoff R, Jacobs DR,Jr. Is serum gamma glutamyltransferase a marker of oxidative stress? Free Radic Res 2004;38:535-9. 28. Lee DH, Gross MD, Steffes MW, Jacobs DR,Jr. Is serum gamma-glutamyltransferase a biomarker of xenobiotics, which are conjugated by glutathione? Arterioscler Thromb Vasc Biol 2008;28:e26,8; author reply e29.

CHAPTER 9

SYSTEMATIC COMPARISON OF ROUTINE LABORATORY VALUES WITH OUTCOMES IDENTIFIES PARADOXICAL RELATION OF GAMMA-GLUTAMYL TRANSPEPTIDASE WITH MORTALITY: ICU- LABOME, A LARGE COHORT STUDY OF CRITICALLY ILL PATIENTS

EDRIS M. ALKOZAI BAKHTAWAR K. MAHMOODI JOHAN DECRUYENAERE ROBERT J. PORTE ANNEMIEKE OUDE LANSINK - HARTGRING TON LISMAN MAARTEN W. NIJSTEN

SUBMITTED

I C U - L A B O M E S T U D Y

______118

ABSTRACT

PURPOSE: In ICU patients the relation of a laboratory measurement with outcome may not primarily depend on its deviation from the standard reference interval (RI). The ICU-Labome study systematically evaluated the univariate association of routine laboratory measurements with outcome.

METHODS: We studied the 35 most frequent blood-based measurements in adults admitted ≥6h to the ICU of our university hospital in 1992-2013. Measurements were obtained from the first 14 ICU-days and directly before ICU-admission. Various metrics (including variability) of the laboratory measurements were related with hospital survival. ICU-based range were derived from laboratory measurements obtained just before ICU-discharge in patients with a good long-term outcome, defined as no ICU-readmission and 1-year survival.

RESULTS: In 49,464 patients we assessed >20·106 measurements. ICU-readmissions, in- hospital, and 1-year mortality were 13%, 14%, and 19%. On ICU-admission, lactate had the strongest relation with hospital mortality. Variability was independently related with hospital mortality in 30 of the 35 parameters and 13 of the 35 parameters displayed a U-shaped outcome-relation. The medians of 14 of 35 ICU-based ranges that we derived were outside the standard RI. Remarkably, gamma-glutamyl transpeptidase (GGT) had a paradoxical relation with hospital mortality in the second ICU week since more abnormal GGT-levels were observed in hospital survivors.

CONCLUSIONS: ICU-based ranges for blood-based measurements may be more useful than standard RIs in identifying ICU patients at risk. The association of variability with outcome for many laboratory tests, calls the importance of this metric into question. Late elevation of GGT may confer protection to ICU-patients.

I C U - L A B O M E S T U D Y

______119

INTRODUCTION

In critically ill patients laboratory measurements are frequently abnormal when compared with standard reference intervals (RI, also denoted by reference range or reference values). The use of standard RIs derived from healthy persons to assess the severity of disease or to identify complications may sometimes be inappropriate in ICU patients. In order to identify specific pathophysiological mechanisms or to develop multivariate predictive models for critically ill patients, many ICU studies have evaluated the relation between selected measurements and outcome.1-5 Depending on which measurements are selected or which clinical phase is considered of interest, measurements from patients who have a good outcome may be outside the standard RI. It has been proposed to define RIs for specific patient groups such as hospitalized patients.6 Moreover the implicit assumption that a more deranged measurement will be associated with a worse clinical situation may not always be correct.

The goal of the ICU-Labome study was to comprehensively evaluate the univariate relation of regular laboratory measurements with outcome in ICU patients, and identify ‘ICU-based’ reference ranges, derived from ICU patients with a good outcome. Parameter-specific a priori assumptions or multivariate models such as APACHE-IV or SAPS-3 2,3 were not the scope of this study. This study also assessed a potential U-shaped relation with outcome and whether, in addition to the measurement itself, measurement variability was also associated with outcome, as was previously observed for glucose and potassium.7-9

PATIENTS AND METHODS

[A] PATIENTS The observation period spanned 22 years, from January the 1st 1992 through December the 31th 2013 during which all laboratory values from all patients admitted to our regional tertiary 44-bed ICU were evaluated (Fig. 1). Patients aged 15 years or older were included and data were directly anonymized before further analysis. We recorded the type of ICU admission, and only one ICU admission per patient was evaluated. When patients were admitted multiple times to the ICU the first ICU admission during the last hospital admission was used as the reference ICU-admission. Patients who stayed shorter than 6 hours at the ICU were excluded.

This study was approved by our Medical Ethical Committee (METC 2014.264). Because only anonymized data that had been obtained during routine care were analysed and no additional sampling or interventions were performed for this retrospective study, the institutional review board considered patients’ informed consent not necessary.

I C U - L A B O M E S T U D Y

______120

[B] OUTCOME In-hospital mortality was used as main outcome measure. ICU mortality, ICU readmission, ICU length of stay, hospital length of stay and 1 year survival were also recorded. Good outcome was defined as no ICU-readmission and 1-year survival.

[C] SELECTION, CORRECTION AND PRIMARY REDUCTION OF VALUES We selected the 35 most frequently assessed laboratory measurements in the blood during the first 14 days of ICU stay admission, or directly preceding ICU-admission. This number was arbitrarily chosen to include the vast majority of measurements. Derived values such as were excluded. The standard RIs that we used were provided by our central laboratory in December 2013 (Table 1). Impossible values were searched for and deleted. Since the standard RI or some measurement techniques were repeatedly modified over the 22 year study period, for each of the 35 laboratory parameters we verified whether abrupt time-dependent long-term changes had occurred. For all laboratory measurements, we determined for each calendar day (day 1=01-01-1992, day 8036=31-12-2013) its overall median value. Then, a 100-day running mean was plotted and when evident structural changes were visually noted, we performed a linear correction to adjust the mean and median value over this period to the mean and median of the most recent period.

ICU-days (1 through 14) were determined in 24 hour blocks counting from the date and time of actual ICU-admission. When patients had multiple measurements of the same parameter within the same ICU-day, the mean of these values was calculated before further analysis, with the exception of the analysis of variability. When patients also had laboratory measurements directly preceding ICU-admission, the mean of all available values in the 5 days (120 hours) before ICU-admission was considered the baseline value.

[D] BIVARIATE CORRELATIONS. We performed bivariate analysis for the 35x(35-1)/2 parameter pairs to identify strongly overlapping or obviously redundant parameters, with a very high correlation (R) indicating that the same underlying signal is measured. All patients/ICU-days were pooled for this analysis.

[E] AREA UNDER THE RECEIVING OPERATING CHARACTERISTICS CURVE (AUROC) To assess the univariate monotonic association between the studied parameters on ICU day 1 with in-hospital mortality, we performed a AUROC. Likewise AUROCs were calculated for all parameters at baseline, ICU-day 2, and for the 12h window before ICU-discharge.

I C U - L A B O M E S T U D Y

______121

Figure 1. Flowchart depicting selection of patients, selection of the top 35 laboratory parameters and subsequent data reduction and analysis. *In case of multiple ICU-stays only the first ICU stay of the last hospital admission was used.

[F] VARIABILITY (SD) AND OUTCOME We also assessed the variability for all 35 parameters over the whole ICU stay (with a maximum of 14 days) in relation with in-hospital mortality. Similar to how variability has been determined for glucose 8 or potassium 9, variability for each parameter for each patient was simply defined as the standard deviation (SD) of all individual measurements, including multiple measurements on the same day, obtained during the ICU-admission. Whether the SD was relevantly associated with outcome was assessed by performing logistic regression analysis with in-hospital mortality as dependent and the parameter’s mean and SD as independent factors. The association of SD with outcome was then classified as: ‘ - ‘ SD was not associated with outcome. ‘ + ’ Both SD and mean were related with outcome, but mean had a stronger relation. ‘ ++ ’ Both SD and mean were related with outcome, but SD had a stronger relation, ‘+++’ Only SD was related with outcome.

I C U - L A B O M E S T U D Y

______122

[G] U-SHAPED RELATION WITH OUTCOME For some tests, both abnormally low and abnormally high levels may be associated with poor outcome, e.g. glucose. To assess the presence of such a U-shaped relation with outcome, logistic regression analysis was performed with the individual mean measurement value and squared mean measurement value on ICU-day 1 as independent parameters and in-hospital mortality as dependent parameter. A relevant U-shaped relation was considered present when both coefficients of the quadratic function were significant and the minimum of the parabola (i.e. lowest point of the U-curve) was situated within the 10%-90% range of all individual means.

[H] TIME-COURSE OF MEDIANS IN SURVIVORS AND NON-SURVIVORS AND SOCCER PLOTS. Over 14 ICU-days, we compared the time course of the medians (with IQR) of the 35 laboratory parameters between the patients who did or did not survive hospital admission. Likewise, baseline medians were compared. So-called soccer plots 10 were constructed to provide additional graphical information on the frequency distribution of values in relation with the standard RI, over the ICU-stay for in-hospital survivors and non-survivors. Values within the standard RI are green, whereas yellow, orange and red reflect values both below and above the standard RI, according to criteria detailed in supplementary material. The observed distribution of laboratory measurements may have structurally changed over the years either because changes in the ICU-treatment (e.g. glucose control) or because of changes in the laboratory (e.g. a new assay). For visual recognition and identification of these mechanisms, soccer plots were also made for the 1992-2002 and 2003-2013 periods.

[I] ‘ICU-BASED’- INTERQUARTILE RANGES The use of patient groups themselves to extract reference values has been described before.6 We defined ICU-patients who subsequently demonstrated an uncomplicated course (i.e. no ICU-readmission and survival>1 year) as the reference population with a good outcome to generate these ‘ICU-based‘ reference ranges. The clinical period that was used to extract these laboratory values from, was the last 12h of the ICU-admission before ICU- discharge. In order to obtain conservative estimates of ICU-based ranges, the P25% and P75% (i.e. IQR) were determined, and not the P2.5% and P97.5% as is usually done for standard RI derived from normal individuals.

[J] HEAT MAP TO SUMMARIZE RELATION OF DERANGEMENTS WITH OUTCOME One would expect that when laboratory measurements are deranged, i.e. below or above the standard reference range, that in the sickest patients such measurements would be furthest from the reference range. To verify if this was indeed the case, we plotted at which time the median levels were more deranged in patients who did not survive the in-hospital admission

I C U - L A B O M E S T U D Y

______123

compared to those who did survive. A heat-map was constructed to summarize into a single figure for all the 35 parameters at all time points whether they were more deranged (orange or red), similarly deranged (grey) or less deranged (green or dark green) in non-survivors than survivors on ICU day 1 through 14, at baseline and in the 12h window before ICU- discharge (fig. 4).

GENERAL STATISTICS AND SOFTWARE USED Frequencies between groups were compared using the chi-square test. Medians (interquartile range; IQR) were compared with the Mann-Whitney U-test. Analyses were performed with SPSS 22 (IBM SPSS, Chicago, IL) and graphical representations were generated with Excel (Microsoft, Redmond, WA)

RESULTS

DATA FLOW AND SUPPLEMENTARY MATERIAL Fig.1 depicts the study design and the major data reduction and data synthesis stages with the accompanying number of patients/laboratory measurements at the various stages. Given the scope of this study, the vast majority of the (intermediate) results of various analysis that were performed are reported in a large supplementary material file (supp. mat. 169 pages), to allow consistent reporting of the analyses for all the 35 parameters.

[A] PATIENTS Over the study-period there were 60,605 ICU admissions (Fig. 1). Excluded were 18% admissions since 5% were <15 years, 11% had multiple ICU admissions, and 2% had an ICU-stay <6h. The mean ±SD age of the selected population was 60 ±16 years, 37% were females and 18% were admitted from the emergency department. The largest admission category was cardiothoracic surgery (Table 2).

[B] OUTCOME Mean ICU and hospital length of stay were 4.3 ±9.6 and 18 ±21 days and the frequency distribution of the ICU lengths of stay did not markedly change from 1992-2002 to 2003-‘13 (supp. mat. p 6). 13% of the patients were readmitted to the ICU during the same hospital stay. ICU, in-hospital and one-year mortality were 11%, 14% and 19% respectively. A good outcome, as defined as no ICU-readmission and one-year survival, was reached by 72%.

[C] SELECTION, CORRECTION AND PRIMARY REDUCTION OF MEASUREMENTS Of all 23 million blood measurements recorded, 20 million (87%) were included (Fig. 1). Creatinine, glucose, hemoglobin and potassium were the four most frequently performed measurements and APTT, PT and troponin the least performed (supp. mat. P 4). For 10

I C U - L A B O M E S T U D Y

______124

Tabel 1. Key characteristics and results of 35 laboratory parameters evaluated SD related Median D=discharge Univariate relation Standard with in- ‘ICU-based’ value in patients with Abbr. Parameter Unit with in- hospital IR hospital IQR3 good outcome survival1 mortality2 relative to SIR4 alanine ALAT U/L <45 monotonic +++ 30 (20 - 48) in SIR aminotransferase Alb Albumin g/L 35-50 U-shaped + 28 (24- 32) < SIR Amy Amylase U/L <107 monotonic + 67 (38 - 115) in SIR AP U/L <115 monotonic + 55 (43 - 75) in SIR aPCO2 arterial pCO2 kPa 4.6-6.0 U-shaped ++ 5.0 (4.7 - 5.5) in SIR apH arterial ph 7.35-7.45 U-shaped ++ 7.40 (7.37 - 7.43) in SIR aPO2 arterial PO2 kPa 9.5-13.5 U-shaped ++ 12.4 (10.6 - 15.0) in SIR activated partial APTT Sec 23-33 monotonic + 27 (25 - 31) in SIR prothrombin time Aspartate ASAT U/L <35 monotonic + 40 (28 - 62) > SIR aminotransferase arterial oxygen aSatO2 0.96-0.99 monotonic 0.98 (0.96 - 0.99) in SIR saturation + Bic mmol/L 21-25 monotonic + 23 (21 - 25) in SIR Ca calcium (total) mmol/L 2.20-2.60 U-shaped ++ 1.96 (1.84 - 2.08) < SIR CK-MB creatine MB kinase U/L <5 monotonic + 14 (8 - 24) > SIR CKT total creatine kinase U/L <170 monotonic - 215 (121 - 415) > SIR Cl Chloride mmol/L 97-107 U-shaped +++ 107 (104 - 110) > SIR creat Creatinine umol/L 50-110 monotonic ++ 68 (57 - 83) in SIR CRP C-reactive protein Mg/L <5 monotonic - 57 (23 - 107) > SIR DBI direct bilirubin umol/L <5 monotonic + 4 (2 - 7) in SIR gamma-glutamyl GGT U/L <55 monotonic - 36 (18 - 86) in SIR transpeptidase Glu Glucose mmol/L 4.0-5.5 U-shaped ++ 7.4 (6.3 - 8.9) > SIR Hb hemoglobin mmol/L 7.7 – 10.6 U-shaped +++ 6.4 (5.8 - 7.2) < SIR Ht Hematocrit mmol/L 0.37-0.52 U-shaped ++ 0.30 (0.28 - 0.3) < SIR K Potassium mmol/L 3.5 – 5.0 U-shaped ++ 4.3 (4.0 - 4.6) in SIR Lac Lactate mmol/L 0.5 – 2.2 monotonic - 1.0 (0.8 - 1.4) in SIR lactate LDH U/L < 248 monotonic ++ 297 (213 - 408) > SIR dehydrogenase Leukos leukocyte count 109/L 4 – 10 monotonic ++ 12.2 (9.7 - 15.3) > SIR Mg Magnesium mmol/L 0.70 – 1.00 monotonic - 0.95 (0.82 - 1.11) in SIR Na Sodium mmol/L 135 – 145 U-shaped ++ 137 (135 - 139) in SIR P Phosphate mmol/L 0.70-1.50 monotonic ++ 1.02 (0.85 - 1.20) in SIR PLC platelet count 109/L 150-350 U-shaped ++ 180 (138 - 239) in SIR PT prothrombin time Sec 9.0 - 12.0 monotonic + 11.6 (10.8 - 12.7) in SIR TBI total bilirubin umol/L < 17 monotonic + 10 (7 - 16) in SIR TP total protein g/L 60-80 U-shaped + 53 (47 - 59) < SIR Trop Troponin ng/L <14 monotonic + 51 (15 - 124) > SIR Urea Urea mmol/L 2.2-7.5 monotonic + 6.4 (4.9 - 8.7) in SIR 1All parameters obtained showed a univariate relation with hospital mortality, and for 13 parameters relation was U-shaped. 2The relation of standard deviation (SD) as a measure of variability with in-hospital mortality was classified as detailed in the methods section: ‘-‘ : no relation of SD with mortality to ‘+++’ SD related with outcome, but mean is not. Thus an independent association of variability with outcome was present for the majority of parameters and that in 16 of the 35 parameters SD has stronger relation with outcome than the mean (i.e. ‘++‘And ‘+++’). 3the median (iqr) values obtained at discharge in patients with good outcome were be used as ‘icu-based’ interquartile ranges. A good outcome was defined as no readmission to the icu and long term survival (≥ 1 year). SIR: standard reference range , ‘in sir’ , “sir” respectively denote that median value provided in the previous column is within, below or above the standard reference range. For 14 of the 35 parameters, the median of the “icu- reference range” was found to be outside the standard reference range.

I C U - L A B O M E S T U D Y

______125

Table 2. Patient characteristics parameters linear adjustments were made (supp. mat. 130) due to evident structural Nr of patients 49,464 changes in reported laboratory values over Age (SD) 60 (16) Sex, male 63% time. The 35 day-to-day running means of the Admission from emergency 18% median laboratory values after these department corrections still indicated more several more Type of admission gradual changes as well as changes in spread Cardiothoracic surgery 42.5% Abdominal, vascular and 14.5 % of the medians (supp. mat. p 132 – 166). miscellaneous surgery Neurosurgery 11.2% Medical 5.8% [D] BIVARIATE CORRELATIONS Trauma 3.8% Many parameter pairs showed a positive R Transplantation 1.8% (supp. mat. p 8-10) and two pairs displayed Miscellaneous 20.4% an R>0.90. R was 0.96 for (hemoglobin, Mean (SD) ICU stay* (days) 4.3 (9.6) hematocrit) and 0.96 for (direct bilirubin, total Median (IQR) ICU stay* 1.0 (0.8 - 3.1) (days) bilirubin). Negative correlations were less Mean (SD) hospital stay 18 (21) often than positive correlation and were not (days) as marked as positive correlations. The most Median (IQR) hospital stay 12 (7- 21) (days) negative R was -0.50 as observed for (arterial ICU readmissions 12.8% pH, lactate) pair. ICU mortality 11.0% In-hospital mortality 13.5% 1 year mortality 19.1% [E] AUROC ANALYSIS In 32 out of 35 laboratory measurements we ICU: intensive care unit, IQR: interquartile range, SD: standard deviation. *ICU stay found an AUROC > 0.50. Lactate showed the includes ICU-stay for readmissions. strongest predictive power for in-hospital mortality on ICU-day 1 with an AUROC of 0.731 (95% CI: 0.722 to 0.740; Fig 2) as well as on ICU-day 2. Complete AUROC data values at baseline, ICU-day 1 and 2, the change from ICU- day 1 to 2, and at <12h before ICU-discharge were determined (supp. mat.11-14).

[F] RELATION OF PARAMETER VARIABILITY (SD) WITH OUTCOME For 30 of 35 parameters, SD had an independent relation with in-hospital mortality. Moreover in 16 of these parameters the SD had a stronger relation with outcome than the mean (ALAT, aPCO2, apH, aPO2, Ca, Cl, creat, glu, Hb, Ht, K, LDH, leukos, Na, P and PLC; Table 1).

[G] U-SHAPED RELATION WITH OUTCOME At ICU-day 1, all 35 parameters evaluated showed a univariate relation with in-hospital mortality and 13 of them (Table 1) this relation was U-shaped.

I C U - L A B O M E S T U D Y

______126

0.80

0.70

0.60

0.50

Figure 2: AUROC displayed with 95% confidence intervals. Univariate relation of laboratory parameters on ICU day 1 with in-hospital mortality as assessed by area under the receiver operating curve (AUROC). We assessed univariately the prognostic relevance of 35 studied laboratory variables on ICU day 1. All parameters had an AUC higher than 0.5, except for the total creatine kinase (CKT), hemoglobin (Hb), and potassium (K). Lactate (Lac), urea, and creatinine were the strongest predictors of in-hospital mortality at ICU day 1. Blue color reflects positive relation with mortality; red color reflects inverse relation with mortality. Note that the AUROC will do poorly for variables with a U-shaped outcome relation.

[H] TIME-COURSE OF MEDIANS IN SURVIVORS AND NON-SURVIVORS AND SOCCER PLOTS Time courses of the medians of all 35 measurements for hospital survivors and non- survivors are provided in the supplementary material (supp. mat p 15-50) including the significances of differences. Two of the most remarkable time courses, lactate and gamma- glutamyl transpeptidase (GGT), are shown in Fig. 4. Lactate was markedly higher at all time- points in non-survivors (P<10-21 to P<10-300) and GGT was initially significantly higher (P<3x10-156) and then significantly lower in non-survivors (P<5x10-7). The soccer plots (supp. mat. 58-128) indicate that for some parameters virtually all measurements were outside the standard RI (e.g. albumin), whereas for other parameters (e.g. potassium) virtually all measurements were inside the standard RI but extreme values predominantly occurred in non-survivors. Comparison of soccer plots between 1992-2002 and 2003-2013 show that many temporal trends during ICU-trends appear unchanged, but that some (i.e. Alb, apCO2, Cl, Hb, Ht, Na) become more deranged during ICU-stay in the 2003-2013 period whereas glucose becomes less deranged. All these effects might be accounted for by well-known changes in therapeutic strategies or changes in attitude regarding which laboratory deviations are deemed acceptable.10

I C U - L A B O M E S T U D Y

______127

Figure 3. Time courses of lactate and gamma-glutamyl transpeptidase and outcome. Changes in lactate and GGT during the first 14 ICU-days graphically represented in two different manners. Lactate showed the strongest discrimination between survivors and non-survivors compared to all other 34 parameters that were evaluated. At all time-points non-survivors had higher lactate levels. GGT was initially significantly lower for in-hospital survivors but it increased to levels that are higher than and more out of range than those of non-survivors from ICU day 6 onwards. Gray bars denote standard reference ranges; BL denotes the baseline values; * P<0.05; ** P<0.001.

A clear feature of some parameters was their sharp changes over time. Most extreme was the behavior of the leukocyte count with four phases during the 14 ICU-days, both in survivors and non-survivors (supp. mat. p 41,83,119). Although at each ICU-day the leukocyte count was significantly higher in non-survivors, it can be appreciated that the repeated fluctuations over time impede the association of leukocytosis with prognosis.11

[I] ‘ICU-BASED’ RANGES The median (IQR) laboratory values observed within 12h of ICU discharge in patients with a good outcome were considered as ICU-based IQRs. We found that the median value of ICU based IQRs for 14 parameters (Alb, ASAT, Ca, CK-MB, CKT, Cl, CRP, Glu, Hb, Ht, LDH, leukos,

I C U - L A B O M E S T U D Y

______128

Figure 4. Heat map for the 35 parameters studied, red and orange in this heat-map indicate that derangements (i.e. distance from official reference values) for in-hospital non-survivors were larger than derangements for survivors (red: P < 0.0005 ; orange: P < 0.01). Gray indicates no significant difference between non-survivors and survivors. Green indicates a ‘paradoxically’ larger derangement for survivors than non-survivors (light green: P < 0.0005; dark green: P<0.01). Arterial PCO2 (aPCO2) displays and a few other variables show an early paradoxical relation with outcome. During prolonged ICU stay, only GGT shows a consistently more deranged level in survivors.

TP and Trop) were outside the standard RI (Table 1). This underscores that many “abnormal” laboratory tests before the discharge from the ICU were not “abnormal” from the prognostic point of view. Time courses of all 35 parameters are shown in the supplementary material pages 15-50.

[J] HEAT-MAP TO SUMMARIZE RELATION OF DERANGEMENTS WITH OUTCOME The heat-map (Fig. 4) summarizes all parameters whether non-survivors had more deranged values than survivors at baseline, ICU-day 1 through 14 and before discharge (underlying P- values are provided in supp. mat p 167). As might be expected, some parameters are

I C U - L A B O M E S T U D Y

______129

significantly more deranged (i.e. red color) across all time points in non-survivors (Alb, apH, APTT, Bic, Ca, Creat, Lac, PT and Urea) with underlying P-values as low as <10-300 (supp. mat p 167).

Several parameters did not show differences at most time points, but few showed a ‘paradoxical’ difference (light and dark green) with a more deranged value. Between ICU-day 5 and 14 only GGT showed highly significantly more deranged values in survivors. This paradoxical relation persisted in subgroup analysis for the cardiothoracic surgery, miscellaneous surgery, trauma, and medical groups (supp. mat. 168.

DISCUSSION

In this comprehensive analysis of the most used blood-based laboratory measurements we found that 14 of the 35 parameters were out of the standard RI at ICU-discharge in patients who had a good outcome. Maybe unsurprisingly, our results underscore that early lactate has the strongest predictive value for in-hospital mortality compared to other laboratory variables. Furthermore, we found a U-shaped outcome relation in 13 variables.

Examination of the relation with outcome over the first 14 ICU-days showed a unique pattern of GGT compared to other laboratory variables. The levels of GGT were more deranged in the second week of ICU admission in survivors compared to non-survivors. This analysis of regularly performed lab measurements in our institution had no a priori assumptions about specific mechanisms. We hypothesized that the standard RI may not be a reliable tool in assessing critically ill patients since many patients have derangements upon the discharge from the ICU. In fact, it has been shown that adjusted RIs for the ICU population may decrease the false positive alerts up to 20% and increase the true negative results by 14%.12 Many laboratory derangements in ICU setting may not endanger the patient and do not need further intervention. Thus it should not prevent the patient’s discharge from the ICU. Obviously, if an observed value falls within the interquartile range in ICU-patients with a good outcome, it does not automatically imply that a specific value is associated with minimal risk. In order to assess the mortality risk associated with a specific laboratory parameter value, more advanced analyses are required.13,14

Regarding lactate in critically ill patients, it has become evident that stress and not hypoxia is the most important driver of lactate elevations and explanation of its unique association with outcome.15 An early lactate-guided therapy trial demonstrated improved outcome in critically ill patients.16 Although many other laboratory variables of lesser predictive power are used in common scoring systems 2-4 to predict mortality in ICU patients, lactate is not yet included. Yet the Sepsis-3 consensus recently incorporated lactate into the clinical definition

I C U - L A B O M E S T U D Y

______130

of septic shock.17 The ICU-based IQR for lactate of 0.8 to 1.4 mmol/L (Table 1) indicates that desirable lactate levels are in the lower part of the standard RI range (0.5 to 2.2 mmol/L; Table 1). In the past lactate was not always easy to measure routinely. With the current analytical equipment, we believe our results lend further support to both the routine use and the inclusion of lactate into future scoring systems.

The bivariate correlations showed that the (hemoglobin, hematocrit) and (total bilirubin, direct bilirubin) pairs were strongly correlated. This underscores that hematocrit is a redundant measure on top of hemoglobin, as has been in fact long been demonstrated.18,19 Likewise, the apparent unconditional performance of both the total bilirubin and direct bilirubin measurement does not provide additional information compared to the measurement of total bilirubin alone, at least for this overall patient group.

With regard to parameter variability we were surprised to find that variability was related with outcome for most parameters. Although there has been a strong focus on glucose variability as a therapeutic goal, it is remarkable that variability of, among others, several blood gas parameters or sodium or potassium had a stronger relation with outcome than the mean of these parameters (Table 1). To our judgment this strongly suggests that higher parameter variabilities in patients who do worse are a fundamental reflection of the clinical instability of such patients. Thus, unless specifically proven otherwise, a higher laboratory parameter instability should be considered a consequence and not a cause of a worse ICU outcome. With regard to the parameters that showed a U-shaped relation with outcome (Alb, aPCO2, apH, aPO2, Ca, Cl, glu, Hb, Ht, K, Na, PLC and TP; Table 1), all these parameters with the exception of albumin (Alb) and, to a lesser extent, total protein (TP) are known to manifest both pathological ‘hypo’ and ‘hyper’ states.

In our view the most surprising finding of our survey was the paradoxical relation of GGT with outcome. GGT is a key enzyme in modulating redox-sensitive (extra)cellular defenses 20-22 against toxins. GGT is constitutively expressed in several organs 18 and it breaks down extracellular glutathione (GSH), which generates cysteine for intracellular de novo synthesis of GSH. Higher serum GGT plausibly reflects increased cellular GGT activity and serum GGT increases with chronic exposure to toxic metabolites.20

Apart from its known association with the use of several drugs and ethanol consumption,20 chronically elevated GGT in otherwise healthy persons has emerged as strong risk factor for cardiovascular disease.23 Likewise in patients with liver disease elevated GGT is considered a marker of cholestasis together with indicators such as bilirubin and alkaline phosphatase. We observed earlier that secondarily elevated GGT was associated with increased survival rates after liver transplantation, liver resection and after surgical repair of a ruptured

I C U - L A B O M E S T U D Y

______131

abdominal aortic aneurysm.24-26 The current observation in a large cohort of critically ill patients and in subgroups without known primary liver disease (cardiovascular surgery, trauma and medical; supp. mat. p 168) supports the notion that transiently elevated GGT is crucially involved in some protective mechanisms leading to better outcome of patients admitted to the hospital.

One may also speculate interventions that increase GGT levels, such as GGT-inducing drugs 20 might also confer protection since the restoration of GSH levels with acetylcysteine is key in the management of patients with acetaminophen intoxication.27,28 Acetylcysteine provides cysteine for intracellular de novo synthesis of GSH. Apart from this indication, the benefit of acetylcysteine for related conditions is still unproven.29 We believe this paradoxical relation of GGT might be observed in many existing ICU databases, but it has probably been overlooked because most studies focused on the first few ICU-days. At the very least, caregivers should realize that the development of an elevated GGT in the second ICU-week is not worrisome and should not automatically lead to investigations into its cause.

Another remarkable point about the heat map are the grey zones that reflect phases when abnormalities may be present but that do not discriminate outcome. With regard to the other parameters that are green on the heat-map, these occur all early during ICU-stay and reflect very small absolute differences (i.e. hemoglobin, hematocrit, chloride, total protein and troponin; supp. mat. p 15-50). The heat map shows that at discharge total creatine kinase and glucose were higher (dark green) in survivors than in non-survivors. For total creatine kinase activity it has been demonstrated earlier that lower enzyme activity levels reflect worse outcome. A confounder may also have been the large number of elective cardiothoracical surgery patients that have elevated post-operative total creatine kinase activity but have a lower mortality compared to other patients.

As already noted, a large body of ICU-literature concerns the association of deviating laboratory values and outcome.1-5,13,14 Our observations fully corroborate those observations since patients who did not survive the hospital admission, had more derangements of their laboratory variables. However, our results may also have some practical consequences as it clearly shows that the association of laboratory derangements with poor outcome may sometimes be time dependent and that some derangements might even predict good outcome such as the levels of GGT in the second week of ICU-admission. Thus, ICU care providers should realize that an abnormally high GGT may not necessarily be a cause for alarm or specific diagnostic procedures. In addition, our data show that some laboratory derangements in critically ill patients might not be considered “abnormal” from the prognostic point of view since derangements of 14 out of 35 laboratory variables were not associated with poor outcome (re-admission to the ICU and diseased within 1 year).

I C U - L A B O M E S T U D Y

______132

Our study has several limitations including the retrospective design and that we could not obtain the commonly used ICU severity scores such as APACHE or SOFA 2,5 since these were not available for most patients. Furthermore, we only performed univariate analysis of laboratory variables with outcome as it was not our aim to create predictive models. The study covered more than two decades, with many potentially influential changes in measurement or therapy. However, we surveyed the data exclusively for time effects and report these changes. The 1992-’2002 vs. 2003-2013 soccer plots (supp. mat. p 94-128) indicate that apart from the exceptions noted earlier, the behavior of most parameters has not dramatically changed. We also believe the extensive study period increases the robustness and external validity and reproducibility of many of our observations. Thus we think that our results represent clinically relevant information for the assessment of critically ill patients. Thus we are confident given that our observations can also be reproduced in other ICU cohorts. ICUs possessing large cohorts stored in integrated patient database management systems may be able to elucidate relations between medication or other therapies administered and later laboratory abnormalities, such as elevated GGT.

In conclusion, we believe that using ICU-based interquartile ranges instead of standard reference ranges will decrease the level of uncertainty in clinical decision making. The widespread association of parameter variability with outcome, makes it doubtful whether reducing variability of specific parameters is a useful therapeutic goal. Furthermore, even some derangements such as a late elevation in GGT apparently confer a good outcome.

Supplementary material can be viewed on: http://www.ivcompatibility.org/documents/download_appendix.php?userid=nijstenmwn

ACKNOWLEDGEMENT We thank Frank Doesburg for computer support.

I C U - L A B O M E S T U D Y

______133

REFERENCES 1. Pickering BW, Gajic O, Ahmed A, Herasevich V, Keegan MT. Data utilization for medical decision making at the time of patient admission to ICU. Crit Care Med 2013;41:1502-10. 2. Zimmerman JE, Kramer AA, McNair DS, Malila FM. Acute Physiology and Chronic Health Evaluation (APACHE) IV: hospital mortality assessment for today's critically ill patients. Crit Care Med 2006;34:1297-310. 3. Moreno RP, Metnitz PG, Almeida E, et al. SAPS 3--From evaluation of the patient to evaluation of the intensive care unit. Part 2: Development of a prognostic model for hospital mortality at ICU admission. Intensive Care Med 2005;31:1345-55. 4. Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ. Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med 1995;23:1638-52. 5. Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996;22:707- 10. 6. Kouri T, Kairisto V, Virtanen A, et al. Reference intervals developed from data for hospitalized patients: computerized method based on combination of laboratory and diagnostic data. Clin Chem 1994;40:2209-15. 7. Finfer S, Wernerman J, Preiser JC, et al. Clinical review: Consensus recommendations on measurement of blood glucose and reporting glycemic control in critically ill adults. Crit Care 2013;17:229. 8. Krinsley JS. Glycemic variability: a strong independent predictor of mortality in critically ill patients. Crit Care Med 2008;36:3008-13. 9. Hessels L, Hoekstra M, Mijzen LJ, et al. The relationship between serum potassium, potassium variability and in-hospital mortality in critically ill patients and a before-after analysis on the impact of computer-assisted potassium control. Crit Care 2015;19:4,014-0720-9. 10. Oude Lansink-Hartgring A, Hessels L, Weigel J, de Smet AM, Gommers D, Nannan Panday PV, Hoorn E, Nijsten MW. Long term changes in dysnatremia incidence in the ICU: A shift from hyponatremia to hypernatremi; Ann Intensive Care. 2016 Dec;6(1):22 11. Hansen JB, Wilsgard L, Osterud B. Biphasic changes in leukocytes induced by strenuous exercise. Eur J Appl Physiol Occup Physiol 1991;62:157-61. 12. Kilickaya O, Schmickl C, Ahmed A, et al. Customized reference ranges for laboratory values decrease false positive alerts in intensive care unit patients. PLoS One 2014;9:e107930. 13. Stachon A, Segbers E, Hering S, Kempf R, Holland-Letz T, Krieg M. A laboratory-based risk score for medical intensive care patients. Clin Chem Lab Med 2008;46:855-62. 14. Solinger AB, Rothman SI. Risks of mortality associated with common laboratory tests: a novel, simple and meaningful way to set decision limits from data available in the Electronic Medical Record. Clin Chem Lab Med 2013;51:1803-13. 15. Bakker J, Nijsten MW, Jansen TC. Clinical use of lactate monitoring in critically ill patients. Ann Intensive Care 2013;3:12,5820-3-12. 16. Jansen TC, van Bommel J, Schoonderbeek FJ, et al. Early lactate-guided therapy in intensive care unit patients: a multicenter, open-label, randomized controlled trial. Am J Respir Crit Care Med 2010;182:752-61.

I C U - L A B O M E S T U D Y

______134

17. Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA 2016;315:801-10. 18. Addison DJ. Is routine ordering of both hemoglobin and hematocrit justifiable? Can Med Assoc J 1966;95:974-5. 19. Nijboer JM, van der Horst IC, Hendriks HG, ten Duis HJ, Nijsten MW. Myth or reality: hematocrit and hemoglobin differ in trauma. J Trauma 2007;62:1310-2. 20. Whitfield JB. Gamma glutamyl transferase. Crit Rev Clin Lab Sci 2001;38:263-355. 21. Stenius U, Rubin K, Gullberg D, Hogberg J. gamma-Glutamyltranspeptidase-positive rat hepatocytes are protected from GSH depletion, oxidative stress and reversible alterations of collagen receptors. Carcinogenesis 1990;11:69-73. 22. Rajpert-De Meyts E, Shi M, Chang M, et al. Transfection with gamma-glutamyl transpeptidase enhances recovery from glutathione depletion using extracellular glutathione. Toxicol Appl Pharmacol 1992;114:56-62. 23. Lee DS, Evans JC, Robins SJ, et al. Gamma glutamyl transferase and metabolic syndrome, cardiovascular disease, and mortality risk: the Framingham Heart Study. Arterioscler Thromb Vasc Biol 2007;27:127-33. 24. Alkozai EM, Lisman T, Porte RJ, Nijsten MW. Early elevated serum gamma glutamyl transpeptidase after liver transplantation is associated with better survival. F1000Res 2014;3:85. 25. Alkozai EM, Nijsten MW, de Jong KP, et al. Immediate postoperative low platelet count is associated with delayed liver function recovery after partial liver resection. Ann Surg 2010;251:300-6. 26. Haveman JW, Zeebregts CJ, Verhoeven EL, et al. Changes in laboratory values and their relationship with time after rupture of an abdominal aortic aneurysm. Surg Today 2008;38:1091-101. 27. Keays R, Harrison PM, Wendon JA, et al. Intravenous acetylcysteine in paracetamol induced fulminant hepatic failure: a prospective controlled trial. BMJ 1991;303:1026-9. 28. Heard KJ. Acetylcysteine for acetaminophen poisoning. N Engl J Med 2008;359:285-92. 29. Sklar GE, Subramaniam M. Acetylcysteine treatment for non-acetaminophen-induced acute liver failure. Ann Pharmacother 2004;38:498-500.

CHAPTER 10

GENERAL DISCUSSION AND PERSPECTIVES

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______138

Chronic liver disease resulting in fibrosis and cirrhosis is associated with morbidity and mortality, irrespective of the etiology.1,2 Although major progress has been made in the medical management of patients with liver diseases, liver surgery remains the only curative treatment strategy for many conditions. Yet, factors such as perioperative bleeding, and the capacity of remnant liver to regenerate may affect the outcome following the liver surgery.3-5 In this thesis we conducted clinical and pre-clinical studies that aimed to gain better understanding of factors influencing the outcome of liver surgery. In addition, we evaluate if the development of HCC in patients with cirrhosis is associated with activation of primary hemostasis and whether or not the bioactive molecules within the platelets are altered in cirrhotic patients with hepatocellular carcinoma (HCC). In addition, we studied the paradoxical relation of gamma glutamyl-transferase (GGT) with outcome in liver transplantation patients and in patients admitted to the intensive care unit.

In chapter two, we discussed the surgical, anesthetic, and pharmacological strategies to reduce blood loss in patients undergoing a major liver surgery. Bleeding in liver surgery cannot be completely avoided because of the dual and large blood supply to the liver and because of retrograde bleeding from the sinusoids and hepatic veins during the parenchymal transection. Major blood loss requiring transfusion of blood or blood products is an independent risk factor for postoperative morbidity and mortality.4 Generally, refinements in surgical techniques, anesthesiologic care, and better understanding of coagulation in patients with underlying liver disease have contributed to minimizing blood loss in liver surgery.

The introduction of maintaining a low central venous pressure (CVP < 5mm Hg) during the parenchymal transection, has been one of the key contributors of anesthesiologic techniques in the reduction of bleeding complication in liver surgery. A CVP < 5 mm Hg has shown to result in a four-fold reduction in bleeding and transfusion requirements compared to an anesthesiological management resulting in higher CVPs 6. This mechanism is based on the pressure gradient between the CVP and liver sinusoids. The pressure within the liver sinusoids is related to the pressure in the hepatic veins, which subsequently, is directly related to CVP. The pressure within the sinusoids drops as consequence of clamping the inflow of blood to the liver through the hepatic artery and portal vein during parenchyma transection. A major disadvantage of lower CVP is the increased risk of complications such as air embolism, systemic tissue hypoperfusion, and renal insufficiency. However, most of these complications are transient and not (always) clinically relevant.7,8

Furthermore, better understanding of hemostasis especially in patients with cirrhosis has contributed significantly to a reduction of bleeding complications in liver surgery.9-13 Many patients with advanced cirrhosis may have abnormal coagulation tests including an increased bleeding time, APTT or PT/INR. Historically, it was believed that abnormal coagulation assays are associated with an increased risk of bleeding complications. Hence, many centers

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______139 attempted to decrease blood loss and transfusion requirements by correcting the coagulation assays using whole blood or blood products.14 Frequently, patients were fluid overloaded during the operation. Yet, no reduction in blood loss or transfusion requirements was achieved.15 Paradoxically, the correction of coagulation assays resulted in significantly increased bleeding and transfusion requirements.16 A common misconception has been the assumption that cirrhotic patients have a hemostasis-related bleeding tendency. This was based on the clinical observations that spontaneous bleeding occurs frequently in patients with cirrhosis, liver transplant recipients require considerable transfusion of blood products, patients have thrombocytopenia with functional platelet defects, and that conventional coagulation tests such as PT and APTT are frequently abnormal. It was believed that the increased risk of bleeding is resulted from an inherent hemostatic defect evidenced by both abnormalities in coagulation tests and clinical bleeding. However, bleeding complications are not only limited to cirrhotic patients. Non-cirrhotic patients may also develop bleeding during liver surgery. Moreover, cirrhotic patients have both increased risk of spontaneous bleeding as well as increased risk of developing thrombosis. In fact, the risk of thrombosis may be higher in these patients than the risk of a bleeding complication.17-19 In addition, multiple compensatory changes for hemostatic defects may occur in patients with cirrhosis such as elevation of von Willebrand factor (VWF) and decreased levels of ADAMTS-13 that rebalance the thrombocytopenia and functional platelet defects.14 In addition, defects in procoagulant proteins are compensated for by defects in anticoagulant proteins. Importantly, the conventional coagulation assays such as INR and APTT reflect only functionality of some of the pro-coagulant factors and are fully insensitive to anticoagulant proteins. Therefore conventional coagulation assays in patients with complex hemostatic abnormalities such as patients with cirrhosis are not reliable in predicting bleeding risk.

In chapter three, we evaluated the coagulation status of patients undergoing a (extended) right hemi-hepatectomy using both conventional coagulation tests and thrombin generation assays. Using the thrombin generation assay, we found that partial liver resection was associated with postoperative hypercoagulability, while substantially prolonged PT levels were detected in same patients, reflecting hypocoagulability. In samples taken from hepatectomy patients, the thrombin generation potential remained consistently and significantly high after induction of thromobomodulin, the physiological activator of the natural anti-coagulant protein C. This reflects resistance to the anti-coagulant action of thrombomodulin that may be explained by the sustained post-operative deficiency of the natural anti-coagulant protein C and high levels of factor VIII. Compared to conventional coagulation assays that only reflect functionality of some of the pro-coagulant factors, the thrombin generation assay measures the total amount of thrombin generated during in vitro coagulation taking plasma concentrations of both pro- and anti-coagulants into account. Therefore, the thrombin generation test likely better reflects the in vivo hemostatic balance

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______140 compared to conventional coagulation tests. Our data suggests that a restrictive policy of plasma transfusion during liver resection and that even a more extensive anticoagulation prophylaxis might be warranted since the risk of developing thromboembolic events outweighs the risk of bleeding.20-22 Clinical studies evaluating the policy of more extensive pharmacological anticoagulation prophylaxis in patients undergoing liver resection might be warranted.

Another key factor contributing to the outcome following liver surgery is the ability of the remnant liver to sustain the metabolic needs of the body and to regenerate.23 Despite significant improvements in preoperative screening, a significant proportion of patients develop postoperative complications because the remnant liver or graft is too small or of poor quality to sustain sufficient organ function. This condition is defined as “small for size syndrome (SFSS)” that is characterized by coagulopathy, , ascites, hyperbilirubinemia, encephalopathy, multi organ failure, and ultimately death.24 Defective liver regeneration has been shown to be a key contributor for the development of SFSS.23-26 Therefore, strategies to accelerate posthepatectomy liver regeneration and identification of patients at risk are needed.

Liver regeneration has been studied extensively, and the pathways involving liver regeneration are largely recognized.27-29 Yet, clinically available strategies to enhance posthepatectomy liver regeneration are still lacking. Emerging evidence from in vitro models and small animal models suggest that platelets are an attractive target to accelerate posthepatectomy liver regeneration.30-36 Platelets contain a wide range of biologically active molecules 37,38 that might contribute to the proliferation of hepatocytes following liver resection. In addition, recent innovative studies have shown that platelets are capable of de novo synthesis of a variety of proteins, even though platelets lack a nucleus. Platelets contain both pre-mRNA and mature mRNA and possess the machinery to splice pre-mRNA and to translate this into protein.39,40 Given the recent medical advances in targeting platelets selectively,41 platelets might be a promising target to enhance posthepatectomy liver regeneration and to reduce the risk of SFSS. Yet, it should be emphasized that the beneficial effect of endogenous platelets on liver regeneration and SFSS do not advocate transfusion of exogenous platelet concentrate since transfusion of platelets is associated with an increased risk of postoperative morbidity and mortality.3,42

The role of platelets in mediating liver regeneration has been subject of several small and large animal studies, and few clinical studies in humans.33-36,43 Based on in vitro and animal studies, we evaluated the role of platelets in liver regeneration in a retrospective study of patients who underwent a partial liver resection for colorectal liver metastasis. We provided the first evidence that platelets may also contribute to liver regeneration in humans. Specifically, we showed that a low postoperative platelet count (< 100x 109/L) after partial

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______141 liver resection was associated with an increased risk of mortality and delayed recovery of liver function. The platelet count was a strong and independent predictor of delayed recovery of liver function besides other established risk factor such as RBC transfusion (i.e., extent of bleeding) and the percentage of liver volume removed. Our findings suggest that platelets are likely target to accelerate posthepatectomy liver regeneration. Yet, because of the retrospective design of our study we cannot draw firm conclusions. However, if it is confirmed that platelets are directly related to liver regeneration, selective platelet targeting therapies or enhancement of platelet counts in patients undergoing a major liver resections or (partial) liver transplantation might be an option to prevent posthepatectomy liver dysfunction. Moreover, the clinical implementation of strategies to stimulate platelet-mediated liver regeneration may go beyond the context of partial liver resection or (partial) liver transplantation. It may also be applied in other clinical scenarios in which there is a demand for liver regeneration such as acute liver failure and liver fibrosis. Nevertheless, the precise mechanism by which platelets mediate liver regeneration have yet to be identified. A study from our group 44 showed that platelet internalization followed by platelet RNA transfer to the hepatocytes was in part responsible for platelet-induced hepatocyte proliferation in vitro. Inhibition of platelet uptake as well as treatment of platelets with an RNA-degrading enzyme significantly decreased in vitro proliferation of hepatocytes. It was hypothesized that functional transfer of either or both coding and regulatory RNA species from platelets to hepatocytes may be the key drivers of platelet-mediated liver regeneration. Meyer and colleagues 45 suggested that platelets are internalized either by liver endothelial cells or by hepatocytes from the liver sinusoids and the space of Disse following hemi-hepatectomy. Subsequently, they may release their bioactive contents such as serotonin and other growth factors that directly or indirectly stimulate liver regeneration.

Starlinger and colleagues43 reported recently that a low preoperative serotonin level was associated with postoperative liver dysfunction and morbidity. They concluded to have found evidence that the platelet serotonin levels correlated with liver regeneration in humans. In chapter five we provide evidence against the notion that the decreased postoperative levels of serotonin were associated with liver regeneration. In fact, we prospectively studied the levels of circulating serotonin in patients undergoing a (extended) right hemi-hepatectomy, a pylorus preserving pancreaticoduodenectomy (PPPD), and healthy controls. In addition, we studied the levels of serotonin in afferent and efferent hepatic veins just prior to the start and just after the completion of parenchymal transection. Serotonin levels decreased postoperatively in both patient groups, but the decline was comparable between the hepatectomy and PPPD groups. We also found no significant change in serotonin levels between samples taken in the afferent and efferent liver veins prior to and after hemi- hepatectomy, suggesting that serotonin is not consumed within the regenerative liver.

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______142

Platelets are also increasingly recognized as important players in cancer growth and metastasis.46-51 Elevation of the circulating platelet count (i.e., thrombocytosis),52-55 increased platelet activation, hypercoagulability, and an increased risk of developing venous thromboembolism (VTE) have been reported in different types of cancer 56-61. Elevation of the platelet count has also been reported in patients with HCC.54,55 The elevation of platelet count in HCC is interesting for many reasons. First, almost 90% of patients who are diagnosed with HCC have underlying cirrhosis, which is associated with thrombocytopenia.62 Second, cirrhotic patients have a relatively unstable hemostatic balance that is evidenced by the occurrence of bleeding and thrombotic complications in a significant proportion of patients.14 If HCC, like other types of cancer, leads to a hyperactive primary hemostatic system, the development of HCC in a patient with cirrhosis could shift the balance towards thrombotic complications. Third, given the clinical benefit of aspirin and other platelet inhibitors in metastasis and death from colorectal cancer, the concept of activated platelet guided anti-cancer drug delivery has been suggested that may increase drug concentrations at the cancer sites, while reducing the systemic toxicities 61,63,64 Based on these assumptions, we evaluated whether the development of HCC in patients with cirrhosis results in hyperactivity of primary hemostasis including elevation of platelet activation, altered VWF/ADMTS13 ratio, and platelet activatability. We found no difference in the activity of the hemostatic system in patients with and without HCC. We also found no significant difference in basal platelet activation between the patients and healthy controls. We, however, found decreased activatability in patients compared to controls. Our findings are in sharp contrast with studies in other types of cancer in which increased platelet activation was reported. Differences in methodology such as using soluble P-selectin or CD40 ligands as markers of in-vivo platelet activation instead of a more direct measure of platelet activation as used in our study may be responsible for these differences.65 Nevertheless, our in vitro activation assay may not fully reflect physiology since platelet activation when studied under conditions of flow is similar in patients with cirrhosis compared to controls, despite decreased platelet activation capacity in assays performed under static conditions.66 Based on our findings we suggest that the concept of activated platelet guided anti-cancer drug delivery may be less effective in cirrhotic patients with HCC compared to patients with other types of cancer. The changes in platelet function in patients with HCC appear fully driven by cirrhosis rather than cancer. This may have consequences for antihemostatic therapies, which need to be tailored to the cirrhosis rather than the presence of cancer. Whether the development of HCC results in hypercoagulability in non-cirrhotic patients needs further studies.

Elevation of angiogenic proteins has been documented in platelets of mice in presence of microscopic tumors (< 1 or 2 mm), and in platelets of patients with colorectal cancer 67,68. As platelets store and release a number of angiogenic regulators, these molecules might function as a biomarker for the presence of cancer.47,69,70 We studied if the levels of angiogenic proteins are altered in platelets of patients with hepatitis B- or C related cirrhosis

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______143 with HCC. We studied seven angiogenic proteins including vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF) and platelet-derived growth factor (PDGF). We found no significant differences in platelet levels of any of these proteins between patients who did or did not have HCC. In contrast, we found significant difference in platelet levels of many of these proteins between the patients to healthy controls. This finding may again suggest that the cirrhosis rather than the cancer drives changes in platelet-derived angiogenic proteins. The clear absence of a difference between patients with or without HCC indicates that intraplatelet levels of angiogenic proteins may not be used as a biomarker for the presence of cancer in cirrhosis. Nevertheless, we cannot draw firm conclusions due to the small cohort size, heterogeneity of the patient population, relatively small cancer size and non-advance stage of HCC, and because the majority of HCC patients were not “treatment- naïve”. Whether the levels of angiogenic proteins in platelets of patients with HCC without cirrhosis alters, needs further evaluation.

PREDICTORS OF OUTCOME FOLLOWING LIVER SURGERY

While studying the outcome following liver surgery, we observed a paradoxical inverse behavior of immediate postoperative gamma glutamyl transferase (GGT) compared to other liver function parameters such as total bilirubin, AST, and ALT following liver resection and patients who underwent an operation for a ruptured abdominal aortic aneurysm 71,72. The clinical relevance of GGT is currently limited to patients with cholestasis, as a liver damage maker after excessive alcohol intake, and as a predictor of usage of certain drugs.73 Epidemiological studies in the general population correlate an elevated GGT with cancer development, overall cardiovascular mortality, hypertension, hypertriglyceridemia, obesity, type 2 diabetes mellitus, and stroke.74-82 Yet, GGT is usually elevated in liver transplant recipients that experience good outcomes. Based on these observations, we studied the clinical relevance of GGT in liver transplant patients early- and late postoperatively. We found that the elevation of GGT during the first postoperative week was significantly associated with increased 90-day and 5-year survival. In contrast, an elevated GGT 6 month after transplantation was associated with significantly decreased 5-year survival. Subsequently, we studied the levels of GGT in critically ill patients admitted to the intensive care unit. We found that elevated GGT levels were inversely associated with in hospital mortality; the more deviated the levels of GGT the better the hospital survival and vice versa. GGT was not correlated with most of the liver function tests such as AST (r=0.11), ALT (r=0.11) and total bilirubin (r=0.17). However, it did correlate with alkaline phosphatase (r=0.55). Based on these observations, we suggest that postoperative increase in GGT may not simply be due to the release of GGT from damaged cells because GGT levels do not correlate with other liver damage makers. Besides, the rise in postoperative GGT occurs gradually within a week, reaching the maximum between day 7 and 9 postoperatively. It appears unlikely that GGT is

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______144 leaking from damaged cholangiocytes for a week and that patients who survive the hospital admission would leak more than those who did not survive the hospital admission.

Given the critical role of GGT in glutathione (GSH) metabolism, it seems plausible that GGT plays a key role in oxidative stress.73,83,84 Studies on oxidative stress in patients with chronic hepatitis C virus infection showed an association between the levels of GGT and 8- hydroxydeoxyguanosine (8-OHdG), a marker of oxidative DNA damage. Patients who had a high level of 8-OHdG had significantly higher GGT levels but normal ALT levels.85 Moreover, GGT-deficient mice showed an increase in oxidative stress in the kidney, accumulation of DNA damage in the organs, depletion of mitochondrial glutathione, and impaired production of ATP suggesting the lack of normal antioxidant defenses 86-89. Administration of N-acetylcysteine was shown to reverse many diseases caused by GGT-depletion 86-89. In humans, N- acetylcysteine is the treatment of choice in acetaminophen induced acute liver failure acting as a hepatoprotective agent by restoring hepatic glutathione.

These findings highlights the nature and significance of the normal physiological role of GGT in humans defense mechanism against the oxidative stress induced by reactive oxygen species (ROS). If it is true that GGT protects the hepatocytes against the oxidative stress, this might have clinical implications not only for the postoperative patients but also in patients with chronic and sustained hepatitis- or alcohol-related hepatotoxicity or patients with acute liver failure. Since elevated GGT, almost 100 fold that of normal levels, was found to have no cytotoxic or negative impact on humans health and survival,90 GGT upregulation or administration might be an option. Yet, more research before this hypothesis can be tested in clinical practice is needed.

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______145

REFERENCES

1. Mokdad AA, Lopez AD, Shahraz S, et al. Liver cirrhosis mortality in 187 countries between 1980 and 2010: a systematic analysis. BMC Med 2014;12:145. 2. Blachier M, Leleu H, Peck-Radosavljevic M, Valla DC, Roudot-Thoraval F. The burden of liver disease in Europe: a review of available epidemiological data. J Hepatol 2013;58:593-608. 3. de Boer MT, Christensen MC, Asmussen M, et al. The impact of intraoperative transfusion of platelets and red blood cells on survival after liver transplantation. Anesth Analg 2008;106:32,44. 4. Hendriks HG, van der Meer J, de Wolf JT, et al. Intraoperative blood transfusion requirement is the main determinant of early surgical re-intervention after orthotopic liver transplantation. Transpl Int 2005;17:673-9. 5. Poon RT, Fan ST, Lo CM, et al. Improving perioperative outcome expands the role of hepatectomy in management of benign and malignant hepatobiliary diseases: analysis of 1222 consecutive patients from a prospective database. Ann Surg 2004;240:698,708. 6. Jones RM, Moulton CE, Hardy KJ. Central venous pressure and its effect on blood loss during liver resection. Br J Surg 1998;85:1058-60. 7. Zhang XL, Wang WJ, Wang WJ, Cao N. Effectiveness and safety of controlled venous pressure in liver surgery: a systematic review and network meta-analysis. Biomed Res Int 2015:2015:290234. 8. Correa-Gallego C, Berman A, Denis SC, et al. Renal function after low central venous pressure- assisted liver resection: assessment of 2116 cases. HPB (Oxford) 2015;17:258-64. 9. Habib M, Roberts LN, Patel RK, Wendon J, Bernal W, Arya R. Evidence of rebalanced coagulation in acute liver injury and acute liver failure as measured by thrombin generation. Liver Int 2014;34:672-8. 10. Lisman T, Leebeek FW, de Groot PG. Haemostatic abnormalities in patients with liver disease. J Hepatol 2002;37:280-7. 11. Lisman T, Caldwell SH, Burroughs AK, et al. Hemostasis and thrombosis in patients with liver disease: the ups and downs. J Hepatol 2010;53:362-71. 12. Violi F, Basili S, Raparelli V, Chowdary P, Gatt A, Burroughs AK. Patients with liver cirrhosis suffer from primary haemostatic defects? Fact or fiction? J Hepatol 2011;55:1415-27. 13. Lisman T, Porte RJ. Rebalanced hemostasis in patients with liver disease: evidence and clinical consequences. Blood 2010;116:878-85. 14. Caldwell SH, Hoffman M, Lisman T, et al. Coagulation disorders and hemostasis in liver disease: pathophysiology and critical assessment of current management. Hepatology 2006;44:1039-46. 15. de Franchis R, Arcidiacono PG, Carpinelli L, et al. Randomized controlled trial of desmopressin plus terlipressin vs. terlipressin alone for the treatment of acute variceal hemorrhage in cirrhotic patients: a multicenter, double-blind study. New Italian Endoscopic Club. Hepatology 1993;18:1102-7.

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______146

16. Smyrniotis V, Kostopanagiotou G, Theodoraki K, Tsantoulas D, Contis JC. The role of central venous pressure and type of vascular control in blood loss during major liver resections. Am J Surg 2004;187:398-402. 17. Sogaard KK, Horvath-Puho E, Gronbaek H, Jepsen P, Vilstrup H, Sorensen HT. Risk of venous thromboembolism in patients with liver disease: a nationwide population-based case-control study. Am J Gastroenterol 2009;104:96-101. 18. Northup PG, McMahon MM, Ruhl AP, et al. Coagulopathy does not fully protect hospitalized cirrhosis patients from peripheral venous thromboembolism. Am J Gastroenterol 2006;101:1524. 19. Senzolo M, Sartori MT, Lisman T. Should we give thromboprophylaxis to patients with liver cirrhosis and coagulopathy? HPB (Oxford) 2009;11:459-64. 20. Tzeng CW, Katz MH, Fleming JB, et al. Risk of venous thromboembolism outweighs post- hepatectomy bleeding complications: analysis of 5651 National Surgical Quality Improvement Program patients. HPB (Oxford) 2012;14:506-13. 21. Turley RS, Reddy SK, Shortell CK, Clary BM, Scarborough JE. Venous thromboembolism after hepatic resection: analysis of 5,706 patients. J Gastrointest Surg 2012;16:1705-14. 22. Dondero F, Taille C, Mal H, et al. Respiratory complications: a major concern after right hepatectomy in living liver donors. Transplantation 2006;81:181-6. 23. Dahm F, Georgiev P, Clavien PA. Small-for-size syndrome after partial liver transplantation: definition, mechanisms of disease and clinical implications. Am J Transplant 2005;5:2605-10. 24. Clavien PA, Oberkofler CE, Raptis DA, Lehmann K, Rickenbacher A, El-Badry AM. What is critical for liver surgery and partial liver transplantation: size or quality? Hepatology 2010;52:715-29. 25. Cataldegirmen G, Zeng S, Feirt N, et al. RAGE limits regeneration after massive liver injury by coordinated suppression of TNF-alpha and NF-kappaB. J Exp Med 2005;201:473-84. 26. Longo CR, Patel VI, Shrikhande GV, et al. A20 protects mice from lethal radical hepatectomy by promoting hepatocyte proliferation via a p21waf1-dependent mechanism. Hepatology 2005;42:156-64. 27. Kurinna S, Barton MC. Cascades of transcription regulation during liver regeneration. Int J Biochem Cell Biol 2011;43:189-97. 28. Fausto N, Campbell JS, Riehle KJ. Liver regeneration. Hepatology 2006;43:S45-53. 29. Michalopoulos GK. Liver regeneration after partial hepatectomy: critical analysis of mechanistic dilemmas. Am J Pathol 2010;176:2-13. 30. Clavien PA, Graf R. Liver regeneration and platelets. Br J Surg 2009;96:965-6. 31. Clavien PA. Liver regeneration: a spotlight on the novel role of platelets and serotonin. Swiss Med Wkly 2008;138:361-70. 32. Murata S, Ohkohchi N, Matsuo R, Ikeda O, Myronovych A, Hoshi R. Platelets promote liver regeneration in early period after hepatectomy in mice. World J Surg 2007;31:808-16.

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______147

33. Murata S, Matsuo R, Ikeda O, et al. Platelets promote liver regeneration under conditions of Kupffer cell depletion after hepatectomy in mice. World J Surg 2008;32:1088-96. 34. Matsuo R, Ohkohchi N, Murata S, et al. Platelets Strongly Induce Hepatocyte Proliferation with IGF-1 and HGF In Vitro. J Surg Res 2008;145:279-86. 35. Myronovych A, Murata S, Chiba M, et al. Role of platelets on liver regeneration after 90% hepatectomy in mice. J Hepatol 2008;49:363-72. 36. Lesurtel M, Graf R, Aleil B, et al. Platelet-derived serotonin mediates liver regeneration. Science 2006;312:104-7. 37. Blair P, Flaumenhaft R. Platelet alpha-granules: basic biology and clinical correlates. Blood Rev 2009;23:177-89. 38. Rendu F, Brohard-Bohn B. The platelet release reaction: granules' constituents, secretion and functions. Platelets 2001;12:261-73. 39. Smyth SS, McEver RP, Weyrich AS, et al. Platelet functions beyond hemostasis. J Thromb Haemost 2009;7:1759-66. 40. Schwertz H, Tolley ND, Foulks JM, et al. Signal-dependent splicing of tissue factor pre-mRNA modulates the thrombogenicity of human platelets. J Exp Med 2006;203:2433-40. 41. Franchi F, Rollini F, Park Y, Angiolillo DJ. Novel Antiplatelet Agents: The Current State and What Is Coming Down the Pike. Prog Cardiovasc Dis 2015;58:267-77. 42. Pereboom IT, de Boer MT, Haagsma EB, Hendriks HG, Lisman T, Porte RJ. Platelet transfusion during liver transplantation is associated with increased postoperative mortality due to acute lung injury. Anesth Analg 2009;108:1083-91. 43. Starlinger P, Assinger A, Haegele S, et al. Evidence for serotonin as a relevant inducer of liver regeneration after liver resection in humans. Hepatology 2014;60:257-66. 44. Kirschbaum M, Karimian G, Adelmeijer J, Giepmans BN, Porte RJ, Lisman T. Horizontal RNA transfer mediates platelet-induced hepatocyte proliferation. Blood 2015;126:798-806. 45. Meyer J, Lejmi E, Fontana P, Morel P, Gonelle-Gispert C, Buhler L. A focus on the role of platelets in liver regeneration: Do platelet-endothelial cell interactions initiate the regenerative process? J Hepatol 2015;. 46. Gay LJ, Felding-Habermann B. Contribution of platelets to tumour metastasis. Nat Rev Cancer 2011;11:123-34. 47. Rothwell PM, Wilson M, Price JF, Belch JF, Meade TW, Mehta Z. Effect of daily aspirin on risk of cancer metastasis: a study of incident cancers during randomised controlled trials. Lancet 2012;397:1591-601. 48. Gastpar H. Platelet-cancer cell interaction in metastasis formation: a possible therapeutic approcach to metastasis prophylaxis. J Med 1977;8:103-14. 49. Nash GF, Turner LF, Scully MF, Kakkar AK. Platelets and cancer. Lancet Oncol 2002;3:425- 30.

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______148

50. Honn KV, Tang DG, Crissman JD. Platelets and cancer metastasis: a causal relationship? Cancer Metastasis Rev 1992;11:325-51. 51. Sabrkhany S, Griffioen AW, Oude Egbrink MG. The role of blood platelets in tumor angiogenesis. Biochim Biophys Acta 2011;1815:189-96. 52. Stone RL, Nick AM, McNeish IA, et al. Paraneoplastic thrombocytosis in ovarian cancer. N Engl J Med 2012;366:610-8. 53. Salgado R, Vermeulen PB, Benoy I, et al. Platelet number and interleukin-6 correlate with VEGF but not with bFGF serum levels of advanced cancer patients. Br J Cancer 1999;80:892-7. 54. Carr BI, Guerra V. Thrombocytosis and hepatocellular carcinoma. Dig Dis Sci 2013;58:1790- 6. 55. Hwang SJ, Luo JC, Li CP, et al. Thrombocytosis: a paraneoplastic syndrome in patients with hepatocellular carcinoma. World J Gastroenterol 2004;10:2472-7. 56. Samonakis DN, Koutroubakis IE, Sfiridaki A, et al. Hypercoagulable states in patients with hepatocellular carcinoma. Dig Dis Sci 2004;49:854-8. 57. Chiang HS, Swaim MW, Huang TF. Characterization of platelet aggregation induced by human colon adenocarcinoma cells and its inhibition by snake venom peptides, trigramin and rhodostomin. Br J Haematol 1994;87:325-31. 58. Egan K, Crowley D, Smyth P, et al. Platelet adhesion and degranulation induce pro-survival and pro-angiogenic signalling in ovarian cancer cells. PLoS One 2011;6:e26125. 59. Zucchella M, Dezza L, Pacchiarini L, et al. Human tumor cells cultured "in vitro" activate platelet function by producing ADP or thrombin. Haematologica 1989;74:541-5. 60. Grignani G, Pacchiarini L, Ricetti MM, et al. Mechanisms of platelet activation by cultured human cancer cells and cells freshly isolated from tumor tissues. Invasion Metastasis 1989;9:298-309. 61. Li N. Platelets in cancer metastasis: To help the "villain" to do evil. Int J Cancer 2015;138:2078-87. 62. El-Serag HB. Hepatocellular carcinoma. N Engl J Med 2011;365:1118-27. 63. Demers M, Ho-Tin-Noe B, Schatzberg D, Yang JJ, Wagner DD. Increased efficacy of breast cancer chemotherapy in thrombocytopenic mice. Cancer Res 2011;71:1540-9. 64. Sarkar S, Alam MA, Shaw J, Dasgupta AK. Drug delivery using platelet cancer cell interaction. Pharm Res 2013;30:2785-94. 65. Lisman T, Porte RJ. Pitfalls in assessing platelet activation status in patients with liver disease. Liver Int 2012;32:1027; author reply 1028. 66. Lisman T, Adelmeijer J, de Groot PG, Janssen HL, Leebeek FW. No evidence for an intrinsic platelet defect in patients with liver cirrhosis--studies under flow conditions. J Thromb Haemost 2006;4:2070-2. 67. Klement GL, Yip TT, Cassiola F, et al. Platelets actively sequester angiogenesis regulators. Blood 2009;113:2835-42.

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______149

68. Peterson JE, Zurakowski D, Italiano JE,Jr, et al. VEGF, PF4 and PDGF are elevated in platelets of colorectal cancer patients. Angiogenesis 2012;15:265-73. 69. Italiano JE,Jr, Richardson JL, Patel-Hett S, et al. Angiogenesis is regulated by a novel mechanism: pro- and antiangiogenic proteins are organized into separate platelet alpha granules and differentially released. Blood 2008;111:1227-33. 70. Almog N, Klement GL. Platelet proteome and tumor dormancy: can platelets content serve as predictive biomarkers for exit of tumors from dormancy? Cancers (Basel) 2010;2:842-58. 71. Alkozai EM, Nijsten MW, de Jong KP, et al. Immediate postoperative low platelet count is associated with delayed liver function recovery after partial liver resection. Ann Surg 2010;251:300-6. 72. Haveman JW, Zeebregts CJ, Verhoeven EL, et al. Changes in laboratory values and their relationship with time after rupture of an abdominal aortic aneurysm. Surg Today 2008;38:1091- 101. 73. Whitfield JB. Gamma glutamyl transferase. Crit Rev Clin Lab Sci 2001;38:263-355. 74. Lee DS, Evans JC, Robins SJ, et al. Gamma glutamyl transferase and metabolic syndrome, cardiovascular disease, and mortality risk: the Framingham Heart Study. Arterioscler Thromb Vasc Biol 2007;27:127-33. 75. Fraser A, Harris R, Sattar N, Ebrahim S, Davey Smith G, Lawlor DA. Alanine aminotransferase, gamma-glutamyltransferase, and incident diabetes: the British Women's Heart and Health Study and meta-analysis. Diabetes Care 2009;32:741-50. 76. Brenner H, Rothenbacher D, Arndt V, Schuberth S, Fraisse E, Fliedner TM. Distribution, determinants, and prognostic value of gamma-glutamyltransferase for all-cause mortality in a cohort of construction workers from southern Germany. Prev Med 1997;26:305-10. 77. Kengne AP, Czernichow S, Stamatakis E, Hamer M, Batty GD. Gamma-glutamyltransferase and risk of cardiovascular disease mortality in people with and without diabetes: Pooling of three British Health Surveys. J Hepatol 2012;57:1083-9. 78. Breitling LP, Claessen H, Drath C, Arndt V, Brenner H. Gamma-glutamyltransferase, general and cause-specific mortality in 19,000 construction workers followed over 20 years. J Hepatol 2011;55:594-601. 79. Strasak AM, Kelleher CC, Klenk J, et al. Longitudinal change in serum gamma- glutamyltransferase and cardiovascular disease mortality: a prospective population-based study in 76,113 Austrian adults. Arterioscler Thromb Vasc Biol 2008;28:1857-65. 80. Strasak AM, Pfeiffer RM, Klenk J, et al. Prospective study of the association of gamma- glutamyltransferase with cancer incidence in women. Int J Cancer 2008;123:1902-6. 81. Strasak AM, Goebel G, Concin H, et al. Prospective study of the association of serum gamma- glutamyltransferase with cervical intraepithelial neoplasia III and invasive cervical cancer. Cancer Res 2010;70:3586-93.

G E N E R A L D I S C U S S I O N A N D P E R S P E C T I V E S

______150

82. Li M, McDermott R. Obesity, albuminuria, and gamma-glutamyl transferase predict incidence of hypertension in indigenous Australians in rural and remote communities in northern Australia. J Hypertens 2015;33:704,9; discussion 709-10. 83. Stenius U, Rubin K, Gullberg D, Hogberg J. gamma-Glutamyltranspeptidase-positive rat hepatocytes are protected from GSH depletion, oxidative stress and reversible alterations of collagen receptors. Carcinogenesis 1990;11:69-73. 84. Lee DH, Blomhoff R, Jacobs DR,Jr. Is serum gamma glutamyltransferase a marker of oxidative stress? Free Radic Res 2004;38:535-9. 85. Cardin R, Saccoccio G, Masutti F, Bellentani S, Farinati F, Tiribelli C. DNA oxidative damage in leukocytes correlates with the severity of HCV-related liver disease: validation in an open population study. J Hepatol 2001;34:587-92. 86. Lieberman MW, Wiseman AL, Shi ZZ, et al. Growth retardation and cysteine deficiency in gamma-glutamyl transpeptidase-deficient mice. Proc Natl Acad Sci U S A 1996;93:7923-6. 87. Jean JC, Harding CO, Oakes SM, Yu Q, Held PK, Joyce-Brady M. gamma-Glutamyl transferase (GGT) deficiency in the GGTenu1 mouse results from a single point mutation that leads to a stop codon in the first coding exon of GGT mRNA. Mutagenesis 1999;14:31-6. 88. Rojas E, Valverde M, Kala SV, Kala G, Lieberman MW. Accumulation of DNA damage in the organs of mice deficient in gamma-glutamyltranspeptidase. Mutat Res 2000;447:305-16. 89. Will Y, Fischer KA, Horton RA, et al. Gamma-Glutamyltranspeptidase-Deficient Knockout Mice as a Model to Study the Relationship between Glutathione Status, Mitochondrial Function, and Cellular Function. Hepatology 2000;32:740-9. 90. Bibas M, Zampa G, Procopio A, Guaitolini R. High serum gamma-glutamyltransferase concentrations in a family. N Engl J Med 1994;330:1832-3.

N E D E R L A N D S E S A M E N V A T T I N G

______153

NEDERLANDSE SAMENVATTING

Chronische leverziekte die resulteert in fibrose- en of cirrose is, ongeacht de onderliggende etiologie, geassocieerd met significante morbiditeit en mortaliteit. Hoewel de medicamenteuze behandeling een belangrijk onderdeel is van de specialistische zorg voor deze patiënten, is levertransplantatie de enige potentieel curatieve behandeling. De uitkomst van de leverchirurgie wordt beïnvloed door verschillende perioperatieve factoren, met name door de intra-operatieve bloedverlies, en door de regeneratiecapaciteit van de restlever. Nieuwe therapeutische strategieën om deze factoren te beïnvloeden zullen van groot belang zijn voor verdere verbetering van de prognose. Dit proefschrift beschrijft verschillende klinische- en preklinische studies en heeft als doel om de perioperatieve factoren die de uitkomsten van een leveroperatie beïnvloeden beter in kaart te brengen. Daarnaast richt dit proefschrift zich op de vraag of het ontwikkelen van levercelkanker of hepatocellulair carcinoom (HCC) bij patiënten met cirrose, geassocieerd is met activatie van primaire hemostase en een veranderde samenstelling van de in bloedplaatjes opgeslagen bioactieve moleculen. Verder wordt in twee hoofdstukken de paradoxale rol van gamma glutamyl transferase (GGT) bestudeerd.

Hoofdstuk 1 geeft een beknopte algemene inleiding en de doelstellingen van dit proefschrift weer.

Hoofdstuk 2 is een overzichtsartikel dat zich wijdt aan de risico’s, preventie en de behandeling van de bloedingscomplicaties bij patiënten met of zonder cirrose die een electieve chirurgische ingreep aan de lever ondergaan. Dit artikel beschrijft het mechanisme van bloedverlies bij leverchirurgie en behandelt verschillende chirurgische, anesthesiologische en farmacologische strategieën in het voorkomen, verminderen en behandelen van bloedingscomplicaties.

Hoofdstuk 3 beschrijft een prospectieve observationele studie waarin de perioperatieve stollingsstatus van patiënten die een partiële leverresectie ondergingen werd bestudeerd en vergeleken met de perioperatieve stollingsstatus in patiënten die een alvleesklierresectie ondergingen. Ook werd een groep gezonde vrijwilligers geïncludeerd in de studie als controle groep. Aangezien conventionele stollingstesten zoals de PT/INR en APTT de plasma spiegels van procoagulante stollingsfactoren meten, en de anticoagulante stollingsfactoren buiten beschouwing laten, reflecteren deze testen niet adequaat de stollingsstatus in patiënten met verminderde aanmaak van anticoagulante factoren. De trombine generatie test daarentegen meet zowel de pro- als anticoagulante stollingsfactoren. Het reflecteert derhalve de fysiologische stollingsstatus. Om rekening te houden met het anticoagulante en thrombomoduline afhankelijke proteïne C systeem, werd de trombinegeneratie test N E D E R L A N D S E S A M E N V A T T I N G

______154

uitgevoerd met en zonder trombomoduline. In afwezigheid van thrombomoduline daalde endogene thrombine-potentiaal (ETP) in patiënten in de eerste 7 dagen post leverresectie in vergelijking met de baseline. Echter, in aanwezigheid van thrombomoduline steeg de ETP tot dag 3 post-leverresectie die ook nog verhoogd was dan de ETP waarde van de gezonde vrijwilligers. Deze bevinding stond in scherp contrast met de verhoogde PT in de patiënten post leverresectie, welke juist wees op hypocoagulabiliteit. In de patiënten die een alvleesklierresectie ondergingen werd een normaal tot licht verlaagde ETP gevonden in vergelijking met de gezonde vrijwilligers. Opvallend was dat de trombine generatie test in de patiënten postleverresectie minder gevoelig was voor trombomoduline, terwijl het bij de gezonde vrijwilligers een sterke daling van de ETP veroorzaakte. Wij concludeerden dat de verminderde gevoeligheid voor de trombomoduline post-leverresectie werd veroorzaakt door verlaagde proteïne C en S, in combinatie met het verhoogde factor VIII. Hoewel de conventionele stollingstest hypocoagulabiliteit indiceerde, wezen de trombine generatie testen van patiënten post leverresectie juist op hypercoagulabiliteit. Gezien de complicaties geassocieerd met plasmatransfusie onderstrepen deze resultaten de noodzaak voor een restrictief transfusiebeleid en pleiten ze voor exploratie van meer extensieve toepassing van postoperatieve tromboseprofylaxe.

Hoofdstuk 4 is een retrospectieve studie waarin we aantonen dat bloedplaatjes belangrijk zijn voor de leverfunctieherstel na een partiële leverresectie. Studies in dieren lieten zien dat bloedplaatjes een cruciale rol speelden bij de leverregeneratie in muis. Echter de rol van bloedplaatsje in leverregeneratie bij de mens was onbekend. Wij bestudeerden dit in een relatief grote homogene populatie van patiënten met gemetastaseerd darmkanker naar de lever (n = 216). Een laag aantal bloedplaatjes (< 100 x 109/L) direct na de operatie was een onafhankelijke voorspeller van vertraagd leverfunctieherstel. Verder was een laag aantal bloedpaatjes geassocieerd met een viermaal hoger risico op postoperatieve mortaliteit, hoewel statistisch was het net niet significant (P = 0,06). Concluderend waren onze resultaten in overeenstemming met de eerdere dierenstudies en wezen ze op een cruciale rol van bloedplaatjes in het leverfunctieherstel bij de mens.

Hoofdstuk 5 beschrijft de bevindingen van een prospectief observationeel onderzoek waarin de samenhang tussen de serotoninegehalte en leverregeneratie post-leverresectie werd bestudeerd. In muizen is het serotonine gehalte in bloedplaatjes geassocieerd is met leverregeneratie. De samenhang tussen het serotonine gehalte en leverregeneratie bij de mens is echter nog onduidelijk. In patiënten die een leverresectie ondergingen werd op verschillende tijdstippen (postoperatieve dag 1, 3, 5, 7 en 30) het serotonine gehalte in het plaatjes rijk plasma (PRP) gemeten. Ook werden er vlak voor en vlak na de leverparenchym transsectie bloedmonsters verkregen uit de aanvoerende (v. Porta) en de afvoerende (v. Hepatica) levervenen. Als controlegroep werden patiënten die een alvleesklierresectie N E D E R L A N D S E S A M E N V A T T I N G

______155

ondergingen geïncludeerd. Bovendien werd een groep gezonde vrijwilligers geïncludeerd in de studie. De serotonine spiegel verliep parallel en daalde geleidelijk in beide patiëntengroepen. Vergeleken met de preoperatieve spiegels waren de spiegels op dag 5 post-alvleesklierresectie en op dag 7 post-leverresectie significant verlaagd. Verder werd er geen significant verschil gevonden in de serotoninespiegel tussen de aanvoerende- en de afvoerende levervenen zowel voor als na de leverparenchymtranssectie. We concludeerden dat er geen aanwijzingen werden gevonden dat serotonine betrokken is bij leverregeneratie in de mens.

Hoofdstuk 6 geeft de resultaten van een prospectieve, cross-sectionele studie weer waarin werd onderzocht of het ontwikkelen van HCC bij patiënten met een reeds bestaande hepatitis B- of C- geïnduceerde cirrose een hyperreactiviteit van de primaire hemostase veroorzaakt. Patiënten met hepatitis B/C- geïnduceerde cirrose met en zonder HCC werden geïncludeerd. Ook werd er een groep gezonde vrijwilligers geïncludeerd. De studie werd opgezet op basis van de reeds bestaande literatuur dat het ontwikkelen van diverse soorten kanker in verband brengen met een verhoogde risico op veneuze trombo-embolieën (VTE). De basale- en de agonist (ATP en TRAP) geïnduceerde bloedplaatjesactivatie en bloedplaatjes activeerbaarheid, de plasmaspiegel van von Willebrand factor (VWF) en VWF-cleaving protease ADAMTS13 werden bestudeerd. Er werd geen significant verschil gevonden in de basale bloedplaatjesactivatie en plaatjes activeerbaarheid tussen de patiëntengroepen. De bloedplaatjesactivatie was ook verglijkbaar tussen de patiënten en gezonde vrijwilligers. Echter, de activeerbaarheid van de bloedplaatjes afkomstig van de patiëntengroepen was significant lager vergeleken met die van de gezonde controlegroep. Er werd verder een verhoogde plasmaspiegel van VWF en een verminderde ADAMTS13 activiteit gevonden in patiënten vergeleken met controles. Er werd echter geen verschil gevonden in de spiegels van VWF en ADAMTS13 tussen de patiënten. Wij concludeerden dat het ontwikkelen van de HCC bij patiënten met een reeds bestaande cirrose niet geassocieerd was met relatieve hyperreactiviteit van de primaire hemostase. De complexiteit van de veranderingen in de hemostase van patiënten met cirrose overstijgt waarschijnlijk de verandering die de HCC in hemostase van de cirrotische patiënten induceert.

In hoofdstuk 7 werd onderzocht of het ontwikkelen van de HCC in patiënten met hepatitis B en/of C gerelateerde cirrose geassocieerd is met een veranderde plaatjesspiegel van de pro- en antiangiogene groeifactoren. Er werd recent aangetoond dat de aanwezigheid van microscopische kanker bij muizen de spiegels van pro- en antiangiogene factoren in bloedplaatjes deed stijgen. Aangezien patiënten met hepatitis B en/of C gerelateerde cirrose een verhoogde kans hebben op het ontwikkelen van HCC, zal een vroege diagnose de kans op gunstig beloop doen verhogen. Bestudeerd werden de spiegels van vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), platelet-derived growth factor N E D E R L A N D S E S A M E N V A T T I N G

______156

(PDGF), hepatocyte growth factor (HGF), endostatin, platelet factor 4 (PF4) en thrombospondin type 1 (TSP-1) in plasma en plaatjes. De resultaten werden vergeleken met die van de patiënten met hepatitis B/C- geïnduceerde cirrose die nog geen HCC hadden ontwikkeld of eerder een succesvolle behandeling voor HCC hadden ondergaan en er op het moment van de studie geen aanwijzingen voor HCC waren. Ook werd een groep gezonde vrijwilligers bestudeerd. Hoewel de spiegels van VEGF, bFGF, HGF en endostatin in plaatjes significant hoger waren in patiënten vergeleken met de controle groep, werd er geen significant verschil tussen patiënten en controles gevonden in de spiegels van PDGF, PF4 en TSP-1 in plaatjes. De plasma spiegels van VEGF, bFGF and endostatin waren verglijkbaar in patiënten en controle groep. Verder werd er een verlaagde plasmaspiegel van PDGF, PF4 en TSP-1 in patiënten gevonden. Echter na de correctie voor de plaatjesaantallen verdween dit verschil. Er werd echter geen verschil gevonden in de plasma- en intraplaatjes spiegel van alle onderzochte angiogene eiwitten tussen de patiënten die wel of geen HCC hadden. Wij concludeerden dat plaatjesspiegel van enkele angiogene eiwitten zijn verhoogd bij patiënten met cirrose maar dat er geen verschil was tussen patiënten met en zonder HCC. Derhalve kunnen deze eiwitten niet gebruikt worden als diagnostische en of prognostische marker voor het ontwikkelen van HCC in cirrotische patiënten.

Hoofdstuk 8 beschrijft een retrospectieve studie waarin de klinische relevantie van een verhoogde gamma glutamyltranspeptidase (GGT) vroeg en laat na een levertransplantatie werd onderzocht. Wij bestudeerden 521 patiënten die een eerste levertransplantatie hadden ondergaan. Wij vonden dat een vroeg verhoogd GGT (postoperatieve dag 7) na een levertransplantatie geassocieerd was met verbeterde 90-dagen overleving. Echter, een chronisch verhoogd GGT (zes maanden na de operatie) was geassocieerd met verlaagde 5- jaarsoverleving. Wij concludeerden dat deze opmerkelijke wijziging in de prognostische relevantie van GGT waarschijnlijk verklaard kan worden door een tijd-afhankelijke rol van de GGT in glutathion metabolisme. Postoperatief is een vroeg verhoogde GGT is waarschijnlijk een reflectie van een adequate (systemische) fysiologische reactie op stress terwijl een laat verhoogde GGT een pathologisch proces reflecteert.

Hoofdstuk 9 rapporteert de resultaten van de ICU LABOME studie. Alle patiënten die tussen 1992-2013 voor ≥ 6 uur werden opgenomen op de afdeling intensieve care unit (ICU) van het UMC Groningen werden geïncludeerd. Wij bestudeerden de univariate relatie van de 35 laboratorium variabelen met de uitkomst (ziekenhuismortaliteit, ICU-heropname en eenjaarsoverleving). Op basis van lab-bepalingen van de ICU patiënten die een goede uitkomst hadden (overlevingsduur van meer dan 1 jaar zonder ICU heropname) werd de zogenaamde ICU-based interval ontwikkeld. Ook werd onderzocht welke laboratorium variabelen een U-shape relatie hadden met de ziekenhuismortaliteit.

N E D E R L A N D S E S A M E N V A T T I N G

______157

Meer dan 20x106 laboratorium bepalingen, afkomstig van 49.464 patiënten, werden systematisch geanalyseerd. ICU-heropname, ziekenhuismortaliteit en eenjaarsmortaliteit waren 13%, 14% en 19%. Lactaat had de sterkste relatie met ziekenhuismortaliteit. Bij patiënten met een goede uitkomst lagen waarden van 14 van de 35 bepalingen buiten de standaard referentie intervallen (RIs). Verder werd voor 30 lab-bepalingen een onafhankelijke relatie gevonden tussen de variabiliteit en ziekenhuismortaliteit. 13 variabelen lieten een U- vormige relatie zien met ziekenhuismortaliteit. Opmerkelijk was de paradoxale relatie van Gamma-glutamyl transpeptidase (GGT) in de tweede week van de ICU-opname met ziekenhuismortaliteit. Een verhoogd GGT in de tweede week van de ICU-opname was geassocieerd met verhoogde ziekenhuisoverleving.

Wij concludeerden dat referentie waarden samengesteld uit ICU patiënten die in goede conditie de ICU verlaten geschikter zijn dan de standaard referentie waarden om ICU patiënten met een hoog risico op een slechte uitkomst te identificeren. De associatie tussen variabiliteit en uitkomst voor het merendeel van de laboratorium variabelen brengt het belang van deze bepaling in twijfel.

Hoofdstuk 10 is het huidige hoofdstuk dat bevat een beknopte samenvatting van dit proefschrift gevolgd door discussie en toekomstperspectieven.

A C K N O W L E D G E M E N T S [ D A N K W O O R D ]

______159

ACKNOWLEDGEMENTS (DANKWOORD)

Dit proefschrift is tot stand gekomen door de inzet en medewerking van meerdere personen die op diverse wijze een bijdrage hebben geleverd. Graag zou ik de volgende personen willen bedanken.

Prof. dr. Lisman. Beste Ton, door de jaren heen mocht ik genieten van je expertise en ongekende kunde met betrekking tot alle facetten van het wetenschappelijk onderzoek. Je was altijd constructief in je feedback en gaf mij veel ruimte voor eigen invulling en interpretaties. Ondanks al je drukte heb je mijn artikelen altijd in een rap tempo van kritische commentaar voorzien. Wij hebben in korte tijd veel artikelen gepubliceerd. Door mijn keuzes werd echter de promotiedatum uitgesteld. Daar ben je erg coulant mee omgegaan. Daar ben ik je in het bijzonder enorm dankbaar voor. Je bent de beste promotor die ik kon hebben. Ik wil je ook bedanken voor je tomeloze inzet tijdens mijn gehele promotietraject.

Prof. Dr. R.J. Porte, Beste Robert, tijdens de JSM cursus op Schiermonnikoog werd mijn interesse gewekt voor medisch wetenschappelijk onderzoek. Jij inspireerde mij om mij aan te sluiten bij je onderzoeksgroep. Dat resulteerde in 2010 in de prestigieuze Mozaïek-beurs van de Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO) waarmee ik mijn eigen promotietraject financierde. Ik leerde veel van je. Je bent mijn rolmodel vooral in kennisoverdracht. Je enthousiasme, uitgebreide deskundigheid en interpersoonlijke eigenschappen maken je de beste tweede promotor die ik kon hebben. Bedankt voor je steun en begeleiding.

Dr. M.W. Nijsten, Beste Maarten, wat heb ik mazzel dat je mijn co-promotor werd. Ik bewonder je expertise in het verzamelen en verwerken van de data. Je accurate en zorgvuldige manier van data verzamelen, ruime kennis in wetenschappelijk onderzoek maken je heel bijzonder. Wij hebben soms in de avonden en weekenden gewerkt om de grote LABOM studie af te ronden. Ik heb met veel plezier met je gewerkt en heb enorm veel van je geleerd. Bedankt voor je laagdrempeligheid, leuke en leerzame gesprekken en je begeleiding.

Graag wil ik de leden van de leescommissie, Prof. dr. K.N. Faber, Prof. dr. J.E. Tulleken en Prof. dr. S.W.M. Olde Damink bedanken voor het kritisch lezen en beoordelen van mijn proefschrift.

Dr. Marco Senzolo, dear Marco, my stay in Padua was one of the key moments of my PhD! We were able to collect enough data for our collaborative study in short period of time, which resulted in two great publications. It was amazing to be part of your research group and to be welcomed by all team members. I had a wonderful time at you ward and at the laboratory. Thank you for everything.

A C K N O W L E D G E M E N T S [ D A N K W O O R D ]

______160

Prof. dr. Paolo Simioni, dear prof. Simioni, You welcomed me in Padua by offering me an Italian lunch and the first espresso in Padua! You introduced me to your team and asked everybody to cooperate. At your laboratory, where I spent most of my time in Padua, you facilitated my research with all instruments without any restrictions. I could walk into your office without any appointment for any matter. Your hospitality is unique! You are a true gentleman. Thank you for those great moments.

Dr. Claudia Radu, dear Claudia, I believe my project in Padua would not have been completed without your help in such a short time. It took you hours to adjust the flow cytometer for my analysis. In the days that followed, you kept helping me despite your own scientific projects. I really appreciate your help and thank you for the nice moments in Padua. You are a hardworking and very likable woman. I wish you all the best with your scientific work and personal life. Thank you!

Dr. Crysia Rodrigeuz, dear Crysia, Thanks to my project in Padua, I got to know you. You were my contact person in Padua and assessed me with almost all arrangements. By the time I was there, you had such a busy life. You were in your final months of PhD, you worked as a gastroenterologist, and your parents were there to visit you. Yet, you were always there to help me. I admire your personality and dealing with challenges in life. I got to know you as a friendly, lovable, and hardworking person. Thanks for your friendship and thanks for the nice memories of Padua.

Marc Kirschbaum, Beste Marc, met jou heb ik het meest intensief contact gehad, zowel binnen als buiten het ziekenhuis. Samen zijn we naar symposium in Leiden gegaan waar jij de prijs, een luxe reis naar Bosten, in ontvangst nam. Een reis waaraan veel waardevolle herinneringen zijn gekoppeld. Je stond altijd klaar om mij te vervangen als ik er niet kon zijn. Wij hebben veel gelachen, vaak gebiljart en samen voetbal gekeken. Je bent een intelligente en leuke collega met heel veel humor. Ik wens je nog heel veel succes met je eigen promotieonderzoek en je toekomstige carrière. Bedankt voor alles!

Bakhtawar Khan Mahmoodi, Beste Bakhtawar, Wij kennen elkaar vanaf de eerste dagen dat wij naar Nederland kwamen. Samen hebben we het schakeljaar gedaan en uiteindelijk zijn we beide beland in de medische- en onderzoekswereld. Ik heb enorm veel bewondering voor jouw intelligentie en doorzettingsvermogen. Onze vriendschap is uniek en ik hoop dat het ook zo blijft. Bedankt daarvoor en veel dank dat je mijn paranimf wilt zijn. Als beste vriend, iemand die ontzettend veel kennis heeft in stolling, voel ik me enorm gesteund.

Freeha Arshad, Beste Freeha, Iedere keer dat we iets gingen ondernemen was het veel lachen. Onze achtergrond en gemeenschappelijke interesses, met name voor de situatie van de minderheden en mensenrechten in Nederland, maakte dat we een hechte band kregen met elkaar. Afgelopen zomer heb je je ingezet voor de vluchtelingen die aankwamen bij de kust van Griekenland. Daarvoor heb ik veel respect. Je bent drie maanden geleden

A C K N O W L E D G E M E N T S [ D A N K W O O R D ]

______161

gepromoveerd en je bent recent aangenomen in opleiding tot Spoedeisende Hulp Arts. Wat ben ik trots op je! Dank voor je vriendschap en dank dat je mijn paranimf wilt zijn. Lief van je!

Greg Hugenholtz, Beste Greg, Met jou heb ik vaak leuke en interessante discussies gehad over ons werk maar ook over sociaal maatschappelijke thema’s zoals de Islam, armoede etc. Bedankt voor al die leuke en interessante momenten en veel succes met je promotieonderzoek.

Golnar and Negin Karimian, and Paria Mahboob, It was so amazing to have you ladies in the group. We laughed, we partied, we could speak to each other in our mother tong, we made “Barareh” jokes, and we worked together with lots of pleasure. Thank you for those awesome moments! Beste Golnar, jij hielp mij bovendien met het verzamelen van bloedsamples als ik er niet was. Bedankt daarvoor en succes met je opleiding als MDL-arts.

Ik wil de HPB-chirurgen, Ruben de Kleine, Paul Peeters, Marieke de Boer, Ger Sieders, dr. Koert P. de Jong, Sander IJtsma (toen nog fellow HPB) bedanken voor de medewerking en het beschikbaar maken van bloedsamples tijden de operaties. Ook wil ik graag de collega’s van de anesthesiologie, de intensive care unit, de operatieassistenten en arts-assistenten chirurgie en verpleegkundigen van de afdeling HPB en levertransplantatie ten tijde van mijn onderzoek bedanken voor hun medewerking.

Dott. Alberto Zaneto, dear Alberto, thank you for including participants into my studies. You came very often early in the morning to check if there were patients in the ward who were eligible for my study. It was great to work with you. Thank you!

Thanks to Fabrizio and Titiana Semeraro who were always co-operative and helped me during my experiments in Padua. You both are very loveable persons with great personality. I really admire both of you for your stamina! I wish you both best of luck!

I would like to thank prof. Patrizia Burra, Prof. Paulo Farinati, dr. Francisco Russo, dr. Giacomo Germani and dr. Romilda Cardin for being co-operative during my project in Padua. I would also like to thank all residents of department of hepatology including Elena Nadal, Gemma Maddalo, Ilaria Bortoluzzi, Caterina Pozzan, and Irene Franceschet for being so nice and cooperative. Also special thanks to the employees of hepatology laboratory for helping me throughout my project.

Verder wil ik de zeer gewaardeerde collega’s van het chirurgisch onderzoekslab, Jacco Zwagstra, Jelle Adelmeijer, Susanne Veldhuis, dr. Henri Leuvenink, Janneke Wiersema-Buist, Petra Ottens, Rozemarijn Kox, Douwe Samplonius en Renée Gras bedanken voor de ondersteuning, geduld, hulp en uitleg. Jelle, je hebt vaak mijn werkzaamheden overgenomen terwijl ik in het buitenland op een congres was. Bedankt daarvoor. Jacco, je was altijd bereid om mij te helpen. Ook hadden wij het erg gezellig tijdens feesten en pauzes. Je bent een fijne collega om mee te werken. Veel dank.

A C K N O W L E D G E M E N T S [ D A N K W O O R D ]

______162

Thea Dijkstra-Jansma, Beste Thea, wij kennen elkaar van de goede - en slechte tijden. Een paar jaar geleden verbleef ik in een asielzoekerscentrum in Groningen waar jij bij het Bureau Medische Opvang Asielzoekers werkte. Jaren later werden wij collega’s in het UMCG. Twee verschillende werelden. Maar toch heel bijzonder. Dit gebeurt niet vaak. Veel dank voor de leuke en boeiende gesprekken en je bereidwilligheid om mijn Nederlandse teksten na te kijken.

Linda Albronda, Ellen F. Timmermans, secretaresse HPB en levertransplantatie, bedankt voor je hulp bij de administratieve kant van de promotie. Jan. T. Bottema, Beste Jan, Ik wil je in het bijzonder bedanken voor je bijdrage voor mijn onderzoek.

Mede-promovendi/(oud) studenten in het Chirurgisch Onderzoekslaboratorium, Sanna op den Dries, Ilsalien Bakker, Michiel Kuipers, Michael Sutton, Dane Hoeksma, Leon van Dullemen, Welmoet Westendorp, Pepijn Weeder, Andrie Westerkamp, Anne Marieke Schut, Alix Matton, Laura Burlage, Rolando Rebolledo, Marleen van Oosten, Sanne Nieveld, Lucy Crane, Annelien Morks, Jeffrey Damman, Maximilia Hottenrott, Deborah van Dijk, Astrid Klooster, Valerie Wiersma, Marco Verkaik, samen hebben we bijzondere, leuke en leerzame momenten gehad. Veel dank daarvoor.

Graag bedank ik Dr. ing. Michiel Hooiveld, prof. dr. J.C. Kluin-Nelemans, prof. Henkjan Verkade, prof. Harry Kampinga voor de leuke, leerzame en sfeervolle cursussen van de JSM.

Verder wil ik Martijn van Faassen, Ido P. Kema en alle medewerkers van het ziekenhuislaboratorium bedanken voor hun medewerking. Maaike H. Bansema van de Groningen Graduate School of Medical Science, jou wil ik graag bedanken voor je informatie over de relevante cursussen. Je was altijd lief en vriendelijk.

A C K N O W L E D G E M E N T S [ D A N K W O O R D ]

______163

Ik wil mijn proefschrift sluiten met dankwoord aan mijn familie voor de steun en begrip tijdens mijn promotieonderzoek. Een speciaal dankwoord voor mijn ouders.

Grana Plar-jana (lieve vader), u bent de reden dat ik in Nederland arts ben geworden. In 2004 had u de mogelijkheid om een eind te maken aan de onzekere verblijfssituatie. Toch bleef u in Nederland vanwege mijn toekomst. U hebt uw eigen toekomst en vrijheid opgeofferd zodat ik arts kon worden. U bent altijd trots dat uw zoon in Nederland een arts is geworden. Maar het is ook andersom het geval. Ik ben er trots op dat ik uw zoon ben. Ik ben er trots op dat ik een vader heb die altijd klaar stond om mensen te helpen. Tot op de dag van vandaag hoor ik van alle kanten verhalen over uw liefdadigheid en behulpzaamheid. Ik ben u zeer dankbaar.

Granee moorjane (lieve moeder), woorden schieten te kort om de diepte van uw liefde en affectie te beschrijven. Dergelijke woorden bestaan eigenlijk niet. Daarvoor ben ik u erg dankbaar. Ik ben u ook dankbaar dat u samen met mijn vader mij uw normen en waarden hebt overgedragen. Normen en waarden waardoor mijn persoonlijkheid vorm kreeg en mij heeft gemaakt tot wie ik vandaag ben. Ik kan mij geen betere moeder wensen. U bent mijn wereld!

L I S T O F P U B L I C A T I O NS

______165

LIST OF PUBLICATIONS

Edris M. Alkozai, Ton Lisman, Robert J Porte; Bleeding in Liver Surgery: Prevention and Treatment Clinics in Liver Disease 2009; 13: 145-154

Edris M. Alkozai, Nijsten MW, KP de Jong, de Boer MT, Peeters PM, Slooff MJ, Porte RJ, Lisman T.;Immediate post-operative low platelet count is associated with delayed liver function recovery after partial liver resection Annals of Surgery, 2015; 41: 189-198

Edris M. Alkozai, Marco Senzolo; J. Adelmeijer, Alberto Zanetto, Paolo Simioni, Robert J Porte, Ton Lisman; Levels of Angiogenic Proteins in Plasma And Platelets are not Different in Patients with Infectious Hepatitis Related Cirrhosis and Patients with Cirrhosis and Hepatocellular Carcinoma Platelets, 2014;2:1-6

Edris M. Alkozai, Marco Senzolo; Paolo Simioni, Robert J Porte , Ton Lisman; No evidence for increased platelet activation in patients with hepatitis B- or C-related cirrhosis and hepatocellular carcinoma Thromb Res. 2015; 26: 577-82

Wilma Potze, Edris M. Alkozai (Shared first auteur), Jelle Adelmeijer, Robert J. Porte, Ton Lisman: Normal to increased thrombin generation in patients after a major partial liver resection despite prolonged conventional coagulation tests Alimentary Pharmacology and Therapeutics 2015 Jan;41:189-98.

Edris M. Alkozai, Martijn van Faassen, Ido P. Kema, Robert J. Porte, Ton Lisman; Evidence against a role of serotonin in liver regeneration in humans; Hepatology 2015; 62: 983

Edris M. Alkozai, Ton Lisman, Robert J Porte, Nijsten WNM. Early elevated serum gamma glutamyl transpeptidase after liver transplantation is associated with better survival. F1000Res. 2014 Apr 3;3:85

Edris M. Alkozai, Edris M. Alkozai, Bakhtawar K. Mahmoodi, Johan Decruyenaere, Robert J. Porte, Annemieke Oude Lansink – Hartgring, Ton Lisman, Maarten W. Nijsten. Systematic comparison of routine laboratory measurements with outcomes identifies a paradoxical relation of gamma-glutamyl transpeptidase with mortality: ICU-Labome, a large cohort study of critically ill patients; Manuscript submitted

L I S T O F P U B L I C A T I O NS

______166

Dafna Groeneveld, Edris M. Alkozai, Jelle Adelmeijer, Robert J. Porte, Ton Lisman: An imbalance between von Willebrand factor and ADAMTS13 following a major partial hepatectomy. Br J Surg. 2016; 10.1002

Kirschbaum M, Edris M. Alkozai, Jelle Adelmeijer, Robert J. Porte, Ton Lisman: Evidence against a role for platelet-derived molecules in liver regeneration after partial hepatectomy in humans. J Clin Transl Res, 2016; 2: 1 (Epub ahead of print)

C U R R I C U L U M V I T A E

______167

CURRICULUM VITAE

Edris Alkozai werd op 29 april 1984 geboren in Kabul, Afghanistan. Hij ronde zijn basis- en middelbare school in Afghanistan af. Hij vluchtte in 2001 naar Nederland. Na een eenjarige studie in het Noorderpoortcollege te Groningen behaalde hij het colloquium doctum. In 2004 behaalde hij zijn propedeuse farmacie aan de Rijksuniversiteit Groningen (RUG). In hetzelfde jaar begon hij aan de studie geneeskunde aan de RUG. Tijdens de studie verrichtte hij extracurriculaire activiteiten aan de Junior Scientific Masterclass van de RUG en sloot zich in 2008 aan bij de onderzoeksgroep van Prof. R.J. Porte en Prof. J.A. Lisman, waar hij zijn afstudeeronderzoek deed naar de relatie tussen de bloedplaatjes en leverfunctieherstel c.q. leverregeneratie bij de mensen. Hij volgde de reguliere coschappen in het Universitaire Medische Centrum Groningen (UMCG) en het Medische Centrum Leeuwaarden. Tijdens zijn coschappen behaalde Edris de prestigieuze Mozaiëk beurs van de Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO) waarmee hij zijn eigen onderzoeksproject startte. Hij deed zijn semi-artsstage bij de afdeling Hepatobiliaire Chirurgie van het Caroline Medical Center, te Charlotte, NC, Verenigde Staten en bij de afdeling Chirurgie van het Queen Elizabeth II Health Sciences Center, Halifax, Nova Scotia, Canada. In 2011 behaalde hij zijn artsenbul en begon als promovendus aan zijn promotietraject bij het Chirurgisch Onderzoekslaboratorium van het UMCG, met prof. dr. J.A. Lisman en prof. dr. R.J. Porte als promotoren en dr. M.W. Nijsten als co-promotor. Edris werkt op dit moment bij de afdeling Spoed Eisende Hulp van het Slingeland ziekenhuis en ambieert een academische carrière binnen geneeskunde.