LIVER TRANSPLANTATION 21:644–651, 2015
ORIGINAL ARTICLE
Macrosteatotic and Nonmacrosteatotic Grafts Respond Differently to Intermittent Hepatic Inflow Occlusion: Comparison of Recipient Survival Sangbin Han,1 Gyu-Seong Choi,2 Jong Man Kim,2 Ji Hye Kwon,1 Hyo-Won Park,1 Gaabsoo Kim,1 Choon Hyuck David Kwon,2 Mi Sook Gwak,1 Justin Sangwook Ko,1* and Jae-Won Joh2* 1 Department of Anesthesiology and Pain Medicine; and 2Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea
Intermittent hepatic inflow occlusion (IHIO) during liver graft procurement is known to confer protection against graft ischemia/reperfusion injury and thus may benefit the recipient’s outcome. We evaluated whether the protective effect of IHIO differs with the presence of macrosteatosis (MaS) and with an increase or decrease in the cumulative occlusion time. The subgroup of 188 recipients who received grafts with MaS was divided into 3 groups according to the number of total IHIO rounds during graft procurement: no IHIO, n 5 70; 1 to 2 rounds of IHIO, n 5 50; and 3 rounds of IHIO, n 5 68. Likewise, the subgroup of 200 recipients who received grafts without MaS was divided into 3 groups: no IHIO, n 5 108; 1 to 2 rounds of IHIO, n 5 40; and 3 rounds of IHIO, n 5 52. The Cox model was applied to evaluate the association between the number of total IHIO rounds and recipient survival separately in the subgroup of MaS recipients and the subgroup of non-MaS recipients. Analyzed covariables included the etiology, Milan criteria, transfusion, immunosuppression, and others. In the subgroup of MaS recipients, 1 to 2 rounds of IHIO were favorably associated with recipient survival [hazard ratio (HR), 0.29; 95% confidence interval (CI), 0.10-0.80; P 5 0.03 after Bonferroni correction], whereas 3 rounds of IHIO were not associated with recipient survival (HR, 0.56; 95% CI, 0.25-1.23). In the subgroup of non-MaS recipients, neither 1 to 2 rounds of IHIO (HR, 0.69; 95% CI, 0.30-1.61) nor 3 rounds of IHIO (HR, 0.91; 95% CI, 0.42-1.96) were associated with recipient survival. In conclusion, 1 to 2 rounds of IHIO may be used for the procurement of MaS grafts with potential benefit for recipient survival, whereas IHIO has a limited impact on recipient survival regardless of the cumulative occlusion time C 2015 AASLD. when it is used for non-MaS grafts. Liver Transpl 21:644-651, 2015. V Received August 16, 2014; accepted February 8, 2015. The severe shortage of available liver grafts has led to the increasing use of steatotic livers for living donor liver transplantation (LDLT). However, steatosis
decreases the tolerance of the liver to handle ischemia/reperfusion (IR) injury.1-4 As a multifactorial process leading to inflammation and cell death during
Abbreviations: CI, confidence interval; HBV, hepatitis B virus; HCV, hepatitis C virus; HR, hazard ratio; IHIO, intermittent hepatic inflow occlusion; IPC, ischemic preconditioning; IR, ischemia/reperfusion; LDLT, living donor liver transplantation; MaS, macrosteatosis; MELD, Model for End-Stage Liver Disease. Potential conflict of interest: Nothing to report. Financial support: Nothing to report. *These authors contributed equally to this work. Address reprint requests to Justin Sangwook Ko, M.D., Ph.D., Department of Anesthesiology and Pain Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-dong, Gangnam-gu, Seoul 135-710, Korea. Telephone: 82-2-3410-2470; FAX: 82-2-3410-0361; E-mail: [email protected]. Address reprint requests to Jae-Won Joh, M.D., Ph.D., Department of Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-dong, Gangnam-gu, Seoul 135-710, Korea. Telephone: 82-2-3410-3466; FAX: 82-2-3410-0040; E-mail: [email protected] DOI 10.1002/lt.24097 View this article online at wileyonlinelibrary.com. LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases
C 2015 American Association for the Study of Liver Diseases. V
LIVER TRANSPLANTATION, Vol. 21, No. 5, 2015
the course of graft procurement, preservation, and implantation, IR injury has been identified as a principal cause of impaired graft regeneration and clinical course.2,5 Thus, diverse therapeutic approaches for attenuating IR injury to steatotic livers have been intensely investigated.6 Among these approaches, the most promising results have been reported with ischemic preconditioning (IPC), which is the generation of a brief period of ischemia followed by reperfusion before a substantial IR insult.5,7,8 IPC was first described by Murry et al.,9 who reported that intermittent coronary inflow occlusion (5 minutes of ischemia alternating with 5 minutes of reperfusion) before an extended period (40 minutes) of ischemia was protective against myocardial infarction. Likewise, recent studies have suggested that intermittent hepatic inflow occlusion (IHIO) can provide protection for the liver as an IPC method.10-13 Those clinical studies demonstrated that IHIO was associated with less IR injury to the graft and thus might benefit posttransplant recipient outcomes. More recently, a randomized trial of living donors and their recipients corroborated the safe use of IHIO in LDLT.14 However, the aforementioned studies did not take into account the interaction between IHIO and macrosteatosis (MaS), which is a major risk form of steatosis15,16 and alters the response to hepatic ischemia.3,4 Also, they only examined whether or not IHIO was used and did not evaluate the effect of the cumulative occlusion time, which affects the degree of IR injury.3,17 Thus, we evaluated whether the protective effect of IHIO on the recipient differs with the presence of MaS and with the cumulative occlusion time. We used recipient survival as the study endpoint because overall biochemical and clinical effects of IHIO are supposed to be associated with posttransplant recipient death.
PATIENTS AND METHODS Subjects A total of 388 recipients who underwent first adult-toadult LDLT using right hemiliver grafts procured via laparotomy between October 2008 and January 2013 were included in the study. Among these recipients, a subgroup of 188 recipients who received an MaS graft, which was defined as a graft with >5% MaS,18 was divided into 3 groups on the basis of the cumulative IHIO time: grafts procured without IHIO (n 5 70), grafts with 1 to 2 rounds of IHIO (n 5 50), and grafts with 3 rounds of IHIO (n 5 68).3 Likewise, a subgroup of 200 recipients who received a non-MaS graft was divided into 3 groups: grafts procured without IHIO (n 5 108), grafts with 1 to 2 rounds of IHIO (n 5 40), and grafts with 3 rounds of IHIO (n 5 52). To test the interaction between the cumulative IHIO time and MaS, the analysis was performed separately for the subgroups of MaS recipients and non-MaS recipients. The Institutional Review Board of Samsung Medical Center approved this retrospective cohort study (SMC 2013-01-081) and waived the require-
HAN ET AL. 645
ment for written informed consent. All data were obtained from computerized medical records and our liver transplantation database.
Histologic Evaluation Wedge liver biopsy was performed immediately after the laparotomy of the donor. All grafts were histologically evaluated with both hematoxylin-eosin and Oil Red O staining. Permanent biopsy sections were additionally prepared for better histological confirmation, and tissues were fixed in a 4% neutral buffered formalin solution, embedded in paraffin, and stained with hematoxylin-eosin. Steatosis was qualitatively classified as MaS (1 large fat droplet replacing most of the hepatocyte cytoplasm and pushing the nucleus to the edge of the cytoplasmic border) or microsteatosis (multiple small fatty droplets with a centrally placed nucleus). Quantitative classification was based on the average percentage of fatty droplets occupying the biopsied parenchyma. Data for the steatosis type and degree were obtained from the postoperative confirmatory report rather than the intraoperative frozen report in consideration of intrarater and interrater variability. The postoperative report was based on the review process of 5 to 8 serial frozen and permanent biopsy sections by multiple pathologists.
Operative Management The grafts consisted of hepatic segments 5 to 8 and excluded the middle hepatic vein trunk. Parenchymal dissection and biliary procedures were performed by 1 of 2 expert liver transplant surgeons (J.W.J. or C.H.K.). The transection plane was determined after temporary inflow occlusion via the hepatic artery and portal vein on the right side of the liver. A Cavitron ultrasonic aspirator and bipolar electrocautery were used for the resection; this avoided crushing by clamps. On the basis of a series of articles by Imamura et al.,10,12 who suggested protective effects of IHIO in LDLT, IHIO has been used liberally (without specific indications or contraindications) at our institution since 2008. Its use has been primarily dictated by the amount of bleeding during parenchymal dissection. The results of our clinical trial performed between July 2008 and August 2009 further led to IHIO becoming standard practice at our hospital (21 recipients in the present study were a part of this prior randomized trial).14 In this study, the donors and their recipients were randomized to the use or nonuse of IHIO during graft procurement, and the posttransplant outcomes did not differ, regardless of the use of IHIO. During each round of IHIO, the hepatic artery and the portal vein in the hepatoduodenal ligament were clamped for 15 minutes and unclamped for 5 minutes. The cut parenchymal surface was assessed during each unclamping period, and the clampingunclamping maneuver was repeated in a step-by-step fashion if the bleeding appeared to originate mainly
646 HAN ET AL.
LIVER TRANSPLANTATION, May 2015
TABLE 1. Univariable Cox Model for Posttransplant Death in Recipients With MaS Grafts Descriptive Variables Donor factors Age (year) Male sex (versus female) Macrosteatosis degree (%) Total IHIO rounds 0 1 to 2 (versus 0) 3 (versus 0) Procuring surgeon (versus J.W.Joh) J.H.D. Kwon Recipient factors Age (year) Male sex (versus female) Body mass index (kg/m2) ABO blood type incompatibility Primary etiology HBV-related HCV-related (versus HBV-related) Alcoholic (versus HBV-related) Other (versus HBV-related) Hepatocellular carcinoma None Within Milan (versus none) Beyond Milan (versus none) MELD score Preoperative creatinine (mg/dL) Graft-to-recipient weight ratio (%) Cold ischemia time (minutes) Operative time (minutes) Packed red blood cell transfusion (U) Platelets transfusion (U) Immunosuppression (versus tacrolimus) Cyclosporine-based
Statistics
HR
95% CI
P Value
32 (26-42) 131 (69.7) 10 (5-15)
0.99 0.96 1.02
0.96-1.02 0.49-1.90 0.98-1.06
0.33 0.91 0.30
70 (37.2) 50 (26.6) 68 (36.2)
0.37 1.11
0.47-1.00 0.57-2.19
0.05 0.76
117 (62.2)
1.11
0.59-2.10
0.75
54 (48-59) 142 (75.5) 24.5 (22.4-26.3) 18 (9.6)
1.00 1.28 1.00 0.86
0.99-1.04 0.59-2.80 0.99-1.01 0.27-2.82
0.93 0.53 0.74 0.81
128 22 19 19
(68.1) (11.7) (10.1) (10.1)
2.36 0.87 1.75
1.05-5.28 0.26-2.90 0.67-4.61
0.04 0.82 0.26
69 (36.7) 81 (43.1) 38 (20.2) 14 (10-18) 0.8 (0.7-1.0) 1.06 (0.90-1.25) 80 (69-96) 562 (494-619) 2 (0-4) 0 (0-6)
1.34 2.90 0.98 1.17 1.03 1.00 1.004 1.11 1.05
0.60-2.98 1.28-6.54 0.94-1.02 0.79-1.73 0.32-3.34 0.99-1.01 1.001-1.007 1.04-1.18 0.97-1.12
0.48 0.01 0.35 0.43 0.97 0.72 0.002 0.001 0.23
22 (11.7)
2.17
0.99-4.72
0.05
NOTE: Descriptive statistics are presented as medians and 25th and 75th percentiles or as numbers and percentages.
from the vascular inflows. The biliary procedures were performed as the final phase of the liver resection. Procured grafts were perfused through the portal vein by gravity flow with 2.5 to 3.0 L of histidine tryptophan ketoglutarate solution and subsequently perfused by a syringe through the hepatic artery with approximately 10 mL of the solution until the perfusate was clear. Graft implantation was performed with the standard piggyback technique. After the portal vein anastomosis was complete, the graft was reperfused by consecutive unclamping of the hepatic vein and portal vein. The hepatic artery was subsequently anastomosed, and this was followed by the biliary anastomosis.
Immunologic Regimens Immunosuppression was accomplished by a standardized quadruple regimen consisting of induction with methylprednisolone and basiliximab and maintenance with tacrolimus and mycophenolate mofetil. Basiliximab was injected immediately after graft reperfusion
and on postoperative day 4. Tacrolimus was initiated on postoperative day 3; blood trough concentrations were adjusted to 10 to 15 ng/mL during the first month and to 5 to 10 ng/mL thereafter. Starting on postoperative day 1, 750 mg of mycophenolate mofetil was administered twice a day. Cyclosporine was used if the recipients had renal failure, tacrolimus toxicity, or tacrolimus-refractory rejection.
Statistical Analysis The only study endpoint was overall recipient death. The 5-year recipient survival curve was depicted with the Kaplan-Meier method. The Cox model was used to evaluate the recipient death hazard rate according to the cumulative IHIO time (0, 1-2, and 3 rounds of IHIO) after adjustments for other relevant variables. First, a univariable Cox regression was performed to evaluate the association between the recipient death hazard rate and the following variables: donor/recipient age, donor/recipient sex, cumulative IHIO time, procuring surgeon, recipient body mass index, blood
LIVER TRANSPLANTATION, Vol. 21, No. 5, 2015
HAN ET AL. 647
reported P values were 2-sided, and P < 0.05 was considered statistically significant. SPSS 20.0 (SPSS, Inc, Chicago, IL) was used for the statistical analyses.
RESULTS Analysis for Recipients With MaS Grafts
Figure 1. Posttransplant overall survival for recipients who received (A) MaS grafts and (B) non-MaS grafts, which were procured with 0, 1 to 2, or 3 rounds of IHIO, respectively.
type compatibility, primary etiology of liver disease, Milan criteria, Model for End-Stage Liver Disease (MELD) score, preoperative recipient creatinine level, graft-to-recipient weight ratio, cold ischemia time, operative time, intraoperative blood product transfusion, and immunosuppression regimen. Second, variables with P < 0.2 during the univariable Cox regression were entered into a multivariable Cox model, and the results were described with the hazard ratio (HR) and 95% confidence interval (CI). The number of total IHIO rounds was entered into the multivariable Cox model, regardless of the results of the univariate analysis. P values were corrected with the Bonferroni method for multiple comparisons to decrease the risk of type I errors, as appropriate. The continuous variables are summarized as medians and 25th and 75th percentiles. The categorical variables are presented as numbers and percentages. All
The indications for transplantation for the 188 recipients of MaS grafts were as follows: hepatocellular carcinoma arising from viral cirrhosis (n 5 107), cirrhosis secondary to a viral etiology (n 5 42), alcoholic cirrhosis (n 5 13), hepatocellular carcinoma of an unknown origin (n 5 5), hepatocellular carcinoma arising from alcoholic cirrhosis (n 5 7), cryptogenic cirrhosis (n 5 5), autoimmune hepatitis (n 5 2), toxic hepatitis (n 5 1), primary biliary disease (n 5 4), metabolic disease (n 5 1), and Budd-Chiari syndrome (n 5 1). The degree of graft MaS was as follows: 5% (n 5 79), 10% (n 5 62), 15% (n 5 17), 20% (n 5 13), 25% (n 5 2), and 30% (n 5 15). The total number of IHIO rounds performed during graft procurement was as follows: 0 (n 5 70), 1 (n 5 13), 2 (n 5 37), 3 (n 5 53), 4 (n 5 9), 5 (n 5 5), and 6 (n 5 1). The results of univariable Cox regression are shown in Table 1. One to 2 rounds of IHIO during graft procurement versus no IHIO were associated with a decreased death hazard rate (HR, 0.37; 95% CI, 0.47-1.00) with marginal significance (P 5 0.05). In contrast, 3 rounds of IHIO versus no IHIO were not associated with the death hazard rate (HR, 1.11; 95% CI, 0.57-2.19). Survival curves of MaS recipients who received grafts procured with 0, 1 to 2, and 3 rounds of IHIO are shown in Fig. 1A. Hepatitis C virus (HCV)–related liver diseases and hepatocellular carcinoma beyond the Milan criteria were associated with an increased death hazard rate. The operative time and the amount of allogenic red blood cell transfusions during surgery were positively associated with the death hazard rate. Multivariable Cox regression confirmed that 1 to 2 rounds of IHIO were independently associated with a decreased death hazard rate after adjustments for potential confounders derived from the univariable analysis (HR, 0.29; 95% CI, 0.10-0.80; P 5 0.03 after Bonferroni correction). In contrast, 3 rounds of IHIO were not associated with the death hazard rate (HR, 0.56; 95% CI, 0.25-1.23; P 5 0.83 after Bonferroni Table 2).
Analysis for Recipients With Non-MaS Grafts The indications for transplantation for the 200 recipients of non-MaS grafts were as follows: hepatocellular carcinoma arising from viral cirrhosis (n 5 103), cirrhosis secondary to a viral etiology (n 5 54), alcoholic cirrhosis (n 5 12), hepatocellular carcinoma of an unknown origin (n 5 3), hepatocellular carcinoma arising from alcoholic cirrhosis (n 5 2), cryptogenic cirrhosis (n 5 4), autoimmune hepatitis (n 5 6), toxic hepatitis (n 5 5), primary biliary disease (n 5 7), metabolic disease (n 5 2), and Budd-Chiari syndrome
648 HAN ET AL.
LIVER TRANSPLANTATION, May 2015
TABLE 2. Multivariable Cox Model for Posttransplant Death in Recipients With MaS Grafts Variables Total IHIO rounds (versus 0) 1-2 3 Primary etiology (versus HBV-related) HCV-related Alcoholic Other Hepatocellular carcinoma (versus none) Within Milan Beyond Milan Operative time (minutes) Packed red blood cell transfusion (U) Cyclosporine-based immunosuppression (versus tacrolimus)
HR
95% CI
P Value
0.29 0.56
0.10-0.80 0.25-1.23
0.03* 0.83*
1.98 0.75 2.67
0.92-4.75 0.22-2.53 0.87-8.18
0.39* >0.99* 0.26*
1.60 4.77 1.00 1.10 1.80
0.61-4.19 1.79-12.72 0.99-1.01 1.01-1.20 0.73-4.40
0.67* 0.004* 0.10 0.04 0.20
*P values were corrected by the Bonferroni method for multiple comparisons.
(n 5 2). The total number of IHIO rounds performed during graft procurement was as follows: 0 (n 5 107), 1 (n 5 10), 2 (n 5 31), 3 (n 5 39), 4 (n 5 10), 5 (n 5 1), and 6 (n 5 2).
The results of a univariable Cox regression are shown in Table 3. Neither 1 to 2 rounds of IHIO (HR, 0.81; 95% CI, 0.37-1.78) nor 3 rounds of IHIO (HR, 0.86; 95% CI, 0.43-1.73) were significantly associated
TABLE 3. Univariable Cox Model for Posttransplant Death in Recipients With Non-MaS Grafts Variables Donor factors Age (year) Male sex (versus female) Total IHIO rounds 0 1-2 (versus 0) 3 (versus 0) Procuring surgeon (versus J.W.Joh) J.H.D. Kwon Recipient factors Age (year) Male sex (versus female) Body mass index (kg/m2) ABO blood type incompatibility Primary etiology Viral* Alcoholic (versus viral) Other (versus viral) Hepatocellular carcinoma None Within Milan (versus none) Beyond Milan (versus none) MELD score Preoperative creatinine (mg/dL) Graft-to-recipient weight ratio (%) Cold ischemia time (minutes) Operative time (minutes) Packed red blood cell transfusion (U) Platelets transfusion (U) Immunosuppression (versus tacrolimus) Cyclosporine-based
Descriptive Statistics
HR
95% CI
P Value
29 (23-41) 118 (59.0)
1.01 0.50
0.98-1.03 0.28-0.90
0.67 0.02
107 (53.5) 41 (20.5) 52 (26.0)
0.81 0.86
0.37-1.78 0.43-1.73
0.60 0.67
71 (35.5)
1.24
0.69-2.25
0.47
52 (47-57) 158 (79.0) 23.9 (22.0-26.3) 18 (9.0)
0.99 1.58 0.98 0.73
0.96-1.02 0.71-3.53 0.90-1.06 0.23-2.37
0.39 0.27 0.58 0.60
152 (76.0) 15 (7.5) 33 (16.5)
1.50 1.76
0.53-4.25 0.81-3.83
0.52 0.20
93 (46.5) 70 (35.0) 37 (18.5) 15 (10-24) 0.8 (0.7-1.0) 0.98 (0.86-1.16) 82 (68-100) 547 (490-604) 1 (0-4) 0 (0-6)
0.55 1.14 1.03 1.22 0.34 1.00 1.00 1.10 1.05
0.27-1.12 0.56-2.33 1.01-1.05 0.90-1.65 0.09-1.26 0.99-1.01 1.00-1.01 1.05-1.15 1.00-1.11
0.10 0.72 0.03 0.21 0.11 0.26 0.06
ORIGINAL ARTICLE
Macrosteatotic and Nonmacrosteatotic Grafts Respond Differently to Intermittent Hepatic Inflow Occlusion: Comparison of Recipient Survival Sangbin Han,1 Gyu-Seong Choi,2 Jong Man Kim,2 Ji Hye Kwon,1 Hyo-Won Park,1 Gaabsoo Kim,1 Choon Hyuck David Kwon,2 Mi Sook Gwak,1 Justin Sangwook Ko,1* and Jae-Won Joh2* 1 Department of Anesthesiology and Pain Medicine; and 2Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea
Intermittent hepatic inflow occlusion (IHIO) during liver graft procurement is known to confer protection against graft ischemia/reperfusion injury and thus may benefit the recipient’s outcome. We evaluated whether the protective effect of IHIO differs with the presence of macrosteatosis (MaS) and with an increase or decrease in the cumulative occlusion time. The subgroup of 188 recipients who received grafts with MaS was divided into 3 groups according to the number of total IHIO rounds during graft procurement: no IHIO, n 5 70; 1 to 2 rounds of IHIO, n 5 50; and 3 rounds of IHIO, n 5 68. Likewise, the subgroup of 200 recipients who received grafts without MaS was divided into 3 groups: no IHIO, n 5 108; 1 to 2 rounds of IHIO, n 5 40; and 3 rounds of IHIO, n 5 52. The Cox model was applied to evaluate the association between the number of total IHIO rounds and recipient survival separately in the subgroup of MaS recipients and the subgroup of non-MaS recipients. Analyzed covariables included the etiology, Milan criteria, transfusion, immunosuppression, and others. In the subgroup of MaS recipients, 1 to 2 rounds of IHIO were favorably associated with recipient survival [hazard ratio (HR), 0.29; 95% confidence interval (CI), 0.10-0.80; P 5 0.03 after Bonferroni correction], whereas 3 rounds of IHIO were not associated with recipient survival (HR, 0.56; 95% CI, 0.25-1.23). In the subgroup of non-MaS recipients, neither 1 to 2 rounds of IHIO (HR, 0.69; 95% CI, 0.30-1.61) nor 3 rounds of IHIO (HR, 0.91; 95% CI, 0.42-1.96) were associated with recipient survival. In conclusion, 1 to 2 rounds of IHIO may be used for the procurement of MaS grafts with potential benefit for recipient survival, whereas IHIO has a limited impact on recipient survival regardless of the cumulative occlusion time C 2015 AASLD. when it is used for non-MaS grafts. Liver Transpl 21:644-651, 2015. V Received August 16, 2014; accepted February 8, 2015. The severe shortage of available liver grafts has led to the increasing use of steatotic livers for living donor liver transplantation (LDLT). However, steatosis
decreases the tolerance of the liver to handle ischemia/reperfusion (IR) injury.1-4 As a multifactorial process leading to inflammation and cell death during
Abbreviations: CI, confidence interval; HBV, hepatitis B virus; HCV, hepatitis C virus; HR, hazard ratio; IHIO, intermittent hepatic inflow occlusion; IPC, ischemic preconditioning; IR, ischemia/reperfusion; LDLT, living donor liver transplantation; MaS, macrosteatosis; MELD, Model for End-Stage Liver Disease. Potential conflict of interest: Nothing to report. Financial support: Nothing to report. *These authors contributed equally to this work. Address reprint requests to Justin Sangwook Ko, M.D., Ph.D., Department of Anesthesiology and Pain Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-dong, Gangnam-gu, Seoul 135-710, Korea. Telephone: 82-2-3410-2470; FAX: 82-2-3410-0361; E-mail: [email protected]. Address reprint requests to Jae-Won Joh, M.D., Ph.D., Department of Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-dong, Gangnam-gu, Seoul 135-710, Korea. Telephone: 82-2-3410-3466; FAX: 82-2-3410-0040; E-mail: [email protected] DOI 10.1002/lt.24097 View this article online at wileyonlinelibrary.com. LIVER TRANSPLANTATION.DOI 10.1002/lt. Published on behalf of the American Association for the Study of Liver Diseases
C 2015 American Association for the Study of Liver Diseases. V
LIVER TRANSPLANTATION, Vol. 21, No. 5, 2015
the course of graft procurement, preservation, and implantation, IR injury has been identified as a principal cause of impaired graft regeneration and clinical course.2,5 Thus, diverse therapeutic approaches for attenuating IR injury to steatotic livers have been intensely investigated.6 Among these approaches, the most promising results have been reported with ischemic preconditioning (IPC), which is the generation of a brief period of ischemia followed by reperfusion before a substantial IR insult.5,7,8 IPC was first described by Murry et al.,9 who reported that intermittent coronary inflow occlusion (5 minutes of ischemia alternating with 5 minutes of reperfusion) before an extended period (40 minutes) of ischemia was protective against myocardial infarction. Likewise, recent studies have suggested that intermittent hepatic inflow occlusion (IHIO) can provide protection for the liver as an IPC method.10-13 Those clinical studies demonstrated that IHIO was associated with less IR injury to the graft and thus might benefit posttransplant recipient outcomes. More recently, a randomized trial of living donors and their recipients corroborated the safe use of IHIO in LDLT.14 However, the aforementioned studies did not take into account the interaction between IHIO and macrosteatosis (MaS), which is a major risk form of steatosis15,16 and alters the response to hepatic ischemia.3,4 Also, they only examined whether or not IHIO was used and did not evaluate the effect of the cumulative occlusion time, which affects the degree of IR injury.3,17 Thus, we evaluated whether the protective effect of IHIO on the recipient differs with the presence of MaS and with the cumulative occlusion time. We used recipient survival as the study endpoint because overall biochemical and clinical effects of IHIO are supposed to be associated with posttransplant recipient death.
PATIENTS AND METHODS Subjects A total of 388 recipients who underwent first adult-toadult LDLT using right hemiliver grafts procured via laparotomy between October 2008 and January 2013 were included in the study. Among these recipients, a subgroup of 188 recipients who received an MaS graft, which was defined as a graft with >5% MaS,18 was divided into 3 groups on the basis of the cumulative IHIO time: grafts procured without IHIO (n 5 70), grafts with 1 to 2 rounds of IHIO (n 5 50), and grafts with 3 rounds of IHIO (n 5 68).3 Likewise, a subgroup of 200 recipients who received a non-MaS graft was divided into 3 groups: grafts procured without IHIO (n 5 108), grafts with 1 to 2 rounds of IHIO (n 5 40), and grafts with 3 rounds of IHIO (n 5 52). To test the interaction between the cumulative IHIO time and MaS, the analysis was performed separately for the subgroups of MaS recipients and non-MaS recipients. The Institutional Review Board of Samsung Medical Center approved this retrospective cohort study (SMC 2013-01-081) and waived the require-
HAN ET AL. 645
ment for written informed consent. All data were obtained from computerized medical records and our liver transplantation database.
Histologic Evaluation Wedge liver biopsy was performed immediately after the laparotomy of the donor. All grafts were histologically evaluated with both hematoxylin-eosin and Oil Red O staining. Permanent biopsy sections were additionally prepared for better histological confirmation, and tissues were fixed in a 4% neutral buffered formalin solution, embedded in paraffin, and stained with hematoxylin-eosin. Steatosis was qualitatively classified as MaS (1 large fat droplet replacing most of the hepatocyte cytoplasm and pushing the nucleus to the edge of the cytoplasmic border) or microsteatosis (multiple small fatty droplets with a centrally placed nucleus). Quantitative classification was based on the average percentage of fatty droplets occupying the biopsied parenchyma. Data for the steatosis type and degree were obtained from the postoperative confirmatory report rather than the intraoperative frozen report in consideration of intrarater and interrater variability. The postoperative report was based on the review process of 5 to 8 serial frozen and permanent biopsy sections by multiple pathologists.
Operative Management The grafts consisted of hepatic segments 5 to 8 and excluded the middle hepatic vein trunk. Parenchymal dissection and biliary procedures were performed by 1 of 2 expert liver transplant surgeons (J.W.J. or C.H.K.). The transection plane was determined after temporary inflow occlusion via the hepatic artery and portal vein on the right side of the liver. A Cavitron ultrasonic aspirator and bipolar electrocautery were used for the resection; this avoided crushing by clamps. On the basis of a series of articles by Imamura et al.,10,12 who suggested protective effects of IHIO in LDLT, IHIO has been used liberally (without specific indications or contraindications) at our institution since 2008. Its use has been primarily dictated by the amount of bleeding during parenchymal dissection. The results of our clinical trial performed between July 2008 and August 2009 further led to IHIO becoming standard practice at our hospital (21 recipients in the present study were a part of this prior randomized trial).14 In this study, the donors and their recipients were randomized to the use or nonuse of IHIO during graft procurement, and the posttransplant outcomes did not differ, regardless of the use of IHIO. During each round of IHIO, the hepatic artery and the portal vein in the hepatoduodenal ligament were clamped for 15 minutes and unclamped for 5 minutes. The cut parenchymal surface was assessed during each unclamping period, and the clampingunclamping maneuver was repeated in a step-by-step fashion if the bleeding appeared to originate mainly
646 HAN ET AL.
LIVER TRANSPLANTATION, May 2015
TABLE 1. Univariable Cox Model for Posttransplant Death in Recipients With MaS Grafts Descriptive Variables Donor factors Age (year) Male sex (versus female) Macrosteatosis degree (%) Total IHIO rounds 0 1 to 2 (versus 0) 3 (versus 0) Procuring surgeon (versus J.W.Joh) J.H.D. Kwon Recipient factors Age (year) Male sex (versus female) Body mass index (kg/m2) ABO blood type incompatibility Primary etiology HBV-related HCV-related (versus HBV-related) Alcoholic (versus HBV-related) Other (versus HBV-related) Hepatocellular carcinoma None Within Milan (versus none) Beyond Milan (versus none) MELD score Preoperative creatinine (mg/dL) Graft-to-recipient weight ratio (%) Cold ischemia time (minutes) Operative time (minutes) Packed red blood cell transfusion (U) Platelets transfusion (U) Immunosuppression (versus tacrolimus) Cyclosporine-based
Statistics
HR
95% CI
P Value
32 (26-42) 131 (69.7) 10 (5-15)
0.99 0.96 1.02
0.96-1.02 0.49-1.90 0.98-1.06
0.33 0.91 0.30
70 (37.2) 50 (26.6) 68 (36.2)
0.37 1.11
0.47-1.00 0.57-2.19
0.05 0.76
117 (62.2)
1.11
0.59-2.10
0.75
54 (48-59) 142 (75.5) 24.5 (22.4-26.3) 18 (9.6)
1.00 1.28 1.00 0.86
0.99-1.04 0.59-2.80 0.99-1.01 0.27-2.82
0.93 0.53 0.74 0.81
128 22 19 19
(68.1) (11.7) (10.1) (10.1)
2.36 0.87 1.75
1.05-5.28 0.26-2.90 0.67-4.61
0.04 0.82 0.26
69 (36.7) 81 (43.1) 38 (20.2) 14 (10-18) 0.8 (0.7-1.0) 1.06 (0.90-1.25) 80 (69-96) 562 (494-619) 2 (0-4) 0 (0-6)
1.34 2.90 0.98 1.17 1.03 1.00 1.004 1.11 1.05
0.60-2.98 1.28-6.54 0.94-1.02 0.79-1.73 0.32-3.34 0.99-1.01 1.001-1.007 1.04-1.18 0.97-1.12
0.48 0.01 0.35 0.43 0.97 0.72 0.002 0.001 0.23
22 (11.7)
2.17
0.99-4.72
0.05
NOTE: Descriptive statistics are presented as medians and 25th and 75th percentiles or as numbers and percentages.
from the vascular inflows. The biliary procedures were performed as the final phase of the liver resection. Procured grafts were perfused through the portal vein by gravity flow with 2.5 to 3.0 L of histidine tryptophan ketoglutarate solution and subsequently perfused by a syringe through the hepatic artery with approximately 10 mL of the solution until the perfusate was clear. Graft implantation was performed with the standard piggyback technique. After the portal vein anastomosis was complete, the graft was reperfused by consecutive unclamping of the hepatic vein and portal vein. The hepatic artery was subsequently anastomosed, and this was followed by the biliary anastomosis.
Immunologic Regimens Immunosuppression was accomplished by a standardized quadruple regimen consisting of induction with methylprednisolone and basiliximab and maintenance with tacrolimus and mycophenolate mofetil. Basiliximab was injected immediately after graft reperfusion
and on postoperative day 4. Tacrolimus was initiated on postoperative day 3; blood trough concentrations were adjusted to 10 to 15 ng/mL during the first month and to 5 to 10 ng/mL thereafter. Starting on postoperative day 1, 750 mg of mycophenolate mofetil was administered twice a day. Cyclosporine was used if the recipients had renal failure, tacrolimus toxicity, or tacrolimus-refractory rejection.
Statistical Analysis The only study endpoint was overall recipient death. The 5-year recipient survival curve was depicted with the Kaplan-Meier method. The Cox model was used to evaluate the recipient death hazard rate according to the cumulative IHIO time (0, 1-2, and 3 rounds of IHIO) after adjustments for other relevant variables. First, a univariable Cox regression was performed to evaluate the association between the recipient death hazard rate and the following variables: donor/recipient age, donor/recipient sex, cumulative IHIO time, procuring surgeon, recipient body mass index, blood
LIVER TRANSPLANTATION, Vol. 21, No. 5, 2015
HAN ET AL. 647
reported P values were 2-sided, and P < 0.05 was considered statistically significant. SPSS 20.0 (SPSS, Inc, Chicago, IL) was used for the statistical analyses.
RESULTS Analysis for Recipients With MaS Grafts
Figure 1. Posttransplant overall survival for recipients who received (A) MaS grafts and (B) non-MaS grafts, which were procured with 0, 1 to 2, or 3 rounds of IHIO, respectively.
type compatibility, primary etiology of liver disease, Milan criteria, Model for End-Stage Liver Disease (MELD) score, preoperative recipient creatinine level, graft-to-recipient weight ratio, cold ischemia time, operative time, intraoperative blood product transfusion, and immunosuppression regimen. Second, variables with P < 0.2 during the univariable Cox regression were entered into a multivariable Cox model, and the results were described with the hazard ratio (HR) and 95% confidence interval (CI). The number of total IHIO rounds was entered into the multivariable Cox model, regardless of the results of the univariate analysis. P values were corrected with the Bonferroni method for multiple comparisons to decrease the risk of type I errors, as appropriate. The continuous variables are summarized as medians and 25th and 75th percentiles. The categorical variables are presented as numbers and percentages. All
The indications for transplantation for the 188 recipients of MaS grafts were as follows: hepatocellular carcinoma arising from viral cirrhosis (n 5 107), cirrhosis secondary to a viral etiology (n 5 42), alcoholic cirrhosis (n 5 13), hepatocellular carcinoma of an unknown origin (n 5 5), hepatocellular carcinoma arising from alcoholic cirrhosis (n 5 7), cryptogenic cirrhosis (n 5 5), autoimmune hepatitis (n 5 2), toxic hepatitis (n 5 1), primary biliary disease (n 5 4), metabolic disease (n 5 1), and Budd-Chiari syndrome (n 5 1). The degree of graft MaS was as follows: 5% (n 5 79), 10% (n 5 62), 15% (n 5 17), 20% (n 5 13), 25% (n 5 2), and 30% (n 5 15). The total number of IHIO rounds performed during graft procurement was as follows: 0 (n 5 70), 1 (n 5 13), 2 (n 5 37), 3 (n 5 53), 4 (n 5 9), 5 (n 5 5), and 6 (n 5 1). The results of univariable Cox regression are shown in Table 1. One to 2 rounds of IHIO during graft procurement versus no IHIO were associated with a decreased death hazard rate (HR, 0.37; 95% CI, 0.47-1.00) with marginal significance (P 5 0.05). In contrast, 3 rounds of IHIO versus no IHIO were not associated with the death hazard rate (HR, 1.11; 95% CI, 0.57-2.19). Survival curves of MaS recipients who received grafts procured with 0, 1 to 2, and 3 rounds of IHIO are shown in Fig. 1A. Hepatitis C virus (HCV)–related liver diseases and hepatocellular carcinoma beyond the Milan criteria were associated with an increased death hazard rate. The operative time and the amount of allogenic red blood cell transfusions during surgery were positively associated with the death hazard rate. Multivariable Cox regression confirmed that 1 to 2 rounds of IHIO were independently associated with a decreased death hazard rate after adjustments for potential confounders derived from the univariable analysis (HR, 0.29; 95% CI, 0.10-0.80; P 5 0.03 after Bonferroni correction). In contrast, 3 rounds of IHIO were not associated with the death hazard rate (HR, 0.56; 95% CI, 0.25-1.23; P 5 0.83 after Bonferroni Table 2).
Analysis for Recipients With Non-MaS Grafts The indications for transplantation for the 200 recipients of non-MaS grafts were as follows: hepatocellular carcinoma arising from viral cirrhosis (n 5 103), cirrhosis secondary to a viral etiology (n 5 54), alcoholic cirrhosis (n 5 12), hepatocellular carcinoma of an unknown origin (n 5 3), hepatocellular carcinoma arising from alcoholic cirrhosis (n 5 2), cryptogenic cirrhosis (n 5 4), autoimmune hepatitis (n 5 6), toxic hepatitis (n 5 5), primary biliary disease (n 5 7), metabolic disease (n 5 2), and Budd-Chiari syndrome
648 HAN ET AL.
LIVER TRANSPLANTATION, May 2015
TABLE 2. Multivariable Cox Model for Posttransplant Death in Recipients With MaS Grafts Variables Total IHIO rounds (versus 0) 1-2 3 Primary etiology (versus HBV-related) HCV-related Alcoholic Other Hepatocellular carcinoma (versus none) Within Milan Beyond Milan Operative time (minutes) Packed red blood cell transfusion (U) Cyclosporine-based immunosuppression (versus tacrolimus)
HR
95% CI
P Value
0.29 0.56
0.10-0.80 0.25-1.23
0.03* 0.83*
1.98 0.75 2.67
0.92-4.75 0.22-2.53 0.87-8.18
0.39* >0.99* 0.26*
1.60 4.77 1.00 1.10 1.80
0.61-4.19 1.79-12.72 0.99-1.01 1.01-1.20 0.73-4.40
0.67* 0.004* 0.10 0.04 0.20
*P values were corrected by the Bonferroni method for multiple comparisons.
(n 5 2). The total number of IHIO rounds performed during graft procurement was as follows: 0 (n 5 107), 1 (n 5 10), 2 (n 5 31), 3 (n 5 39), 4 (n 5 10), 5 (n 5 1), and 6 (n 5 2).
The results of a univariable Cox regression are shown in Table 3. Neither 1 to 2 rounds of IHIO (HR, 0.81; 95% CI, 0.37-1.78) nor 3 rounds of IHIO (HR, 0.86; 95% CI, 0.43-1.73) were significantly associated
TABLE 3. Univariable Cox Model for Posttransplant Death in Recipients With Non-MaS Grafts Variables Donor factors Age (year) Male sex (versus female) Total IHIO rounds 0 1-2 (versus 0) 3 (versus 0) Procuring surgeon (versus J.W.Joh) J.H.D. Kwon Recipient factors Age (year) Male sex (versus female) Body mass index (kg/m2) ABO blood type incompatibility Primary etiology Viral* Alcoholic (versus viral) Other (versus viral) Hepatocellular carcinoma None Within Milan (versus none) Beyond Milan (versus none) MELD score Preoperative creatinine (mg/dL) Graft-to-recipient weight ratio (%) Cold ischemia time (minutes) Operative time (minutes) Packed red blood cell transfusion (U) Platelets transfusion (U) Immunosuppression (versus tacrolimus) Cyclosporine-based
Descriptive Statistics
HR
95% CI
P Value
29 (23-41) 118 (59.0)
1.01 0.50
0.98-1.03 0.28-0.90
0.67 0.02
107 (53.5) 41 (20.5) 52 (26.0)
0.81 0.86
0.37-1.78 0.43-1.73
0.60 0.67
71 (35.5)
1.24
0.69-2.25
0.47
52 (47-57) 158 (79.0) 23.9 (22.0-26.3) 18 (9.0)
0.99 1.58 0.98 0.73
0.96-1.02 0.71-3.53 0.90-1.06 0.23-2.37
0.39 0.27 0.58 0.60
152 (76.0) 15 (7.5) 33 (16.5)
1.50 1.76
0.53-4.25 0.81-3.83
0.52 0.20
93 (46.5) 70 (35.0) 37 (18.5) 15 (10-24) 0.8 (0.7-1.0) 0.98 (0.86-1.16) 82 (68-100) 547 (490-604) 1 (0-4) 0 (0-6)
0.55 1.14 1.03 1.22 0.34 1.00 1.00 1.10 1.05
0.27-1.12 0.56-2.33 1.01-1.05 0.90-1.65 0.09-1.26 0.99-1.01 1.00-1.01 1.05-1.15 1.00-1.11
0.10 0.72 0.03 0.21 0.11 0.26 0.06
