Received: 7 April 2017 Accepted: 3 July 2017 DOI: 10.1111/liv.13514 ORIGINAL ARTICLE Combining albumin- bilirubin score with future liver remnant predicts post- hepatectomy liver failure in HBV- associated HCC patients Heng Zou Yu Wen Kun Yuan Xiong-Ying Miao Li Xiong Kui-Jie Liu Department of General Surgery, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China Correspondence Dr. Heng Zou and Dr. Yu Wen, Department of General Surgery, The Second Xiangya Hospital, Central South University, Changsha, Hunan, China. Emails: zhcsuxy@163.com, wenyu2861@163. com Handling Editor: Alejandro Forner Abstract Background and Aims: Accurate assessment of liver functional reserve pre- operatively is vital for safe hepatic resection. The ALBI score is a new model for assessing liver function. This study aimed to evaluate the value of combining ALBI score with sflr in predicting post- operative morbidity and PHLF in HCC patients who underwent hepatectomy. Methods: Patients undergoing three- dimensional CT reconstruction prior to hepatectomy for HCC between January 2015 and January 2017 were enrolled. The values of the CP score, ALBI score and sflr in predicting post- operative outcomes were evaluated. Results: A total of 229 HCC patients were enrolled; 24 (10.5%) experienced major complications and 21 (9.2%) developed PHLF. The incidence of major complications and PHLF increased with increasing ALBI grade. The ALBI grade classified patients with CP grade A into two subgroups with different incidences of PHLF (P=.029). sflr and ALBI scores were identified as independent predictors of PHLF. The AUC values for the CP score, ALBI score, sflr and sflr ALBI for predicting major complications were 0.600, 0.756, 0.660 and 0.790 respectively. The AUC values of the CP score, ALBI score, sflr and sflr ALBI for predicting PHLF were 0.646, 0.738, 0.758 and 0.884 respectively. Conclusions: The ALBI score showed superior predictive value of post- operative outcomes over CP score, and the combination of sflr and ALBI score was identified as a stronger predictor of post- operative outcomes than the sflr or ALBI score alone. KEYWORDS albumin-bilirubin score, child-pugh score, hepatocellular carcinoma, post-hepatectomy liver failure, standardized future liver remnant 1 INTRODUCTION Hepatocellular carcinoma (HCC) is the sixth most common malignancy and the second leading cause of death from cancer worldwide. 1 Hepatic resection is a potentially curative treatment and remains as the first- line treatment for HCC. Despite great improvements in surgical techniques and instruments, hepatic resection still carries a certain risk of post- hepatectomy liver failure (PHLF), which prolongs Abbreviations: ALBI, albumin-bilirubin; ALT, alanine aminotransferase; AUC, area under the ROC curve; BCLC, Barcelona Clinical Liver Cancer; BSA, body surface area; CP, Child-Pugh; CSPH, clinically significant portal pressure; etlv, estimated total liver volume; FLR, future liver remnant; HBsAg, hepatitis B surface antigen; HCC, hepatocellular carcinoma; INR, international normalized ratio; MELD, model for end-stage liver disease; PHLF, post-hepatectomy liver failure; ROC, receiver operating characteristic; sflr, standardized future liver remnant. Liver International. 2017;1 9. wileyonlinelibrary.com/journal/liv 2017 John Wiley & Sons A/S. 1 Published by John Wiley & Sons Ltd
2 ZOU et al. hospital stays and remains the major cause of post- operative mortality. Reducing post- operative morbidity and mortality is still a concern of hepatic surgeons. Accurate assessment of liver functional reserve prior to surgery is vital to avoid PHLF and post- operative mortality, especially for HCC patients with impaired liver function. The Child- Pugh (CP) grade is the most widely used method for assessment of hepatic function and selecting patients for hepatectomy in clinical practice. 2-4 However, some limitations of the CP grade have been described. 5,6 Recently, the albumin- bilirubin (ALBI) grade has been established as a more convenient and evidence- based model to assess hepatic function in patients with HCC, which only includes serum bilirubin and albumin and eliminates subjective parameters such as encephalopathy and ascites. 5 The ALBI grade has recently been proven to successfully replace the CP grade in predicting postoperative outcomes and survival in several studies. 7-10 Post- operative outcomes mainly depend on the quality and size of the future liver remnant (FLR). 11,12 FLR volume measured preoperatively by three- dimensional CT reconstruction can accurately reflect the size of the actual remnant liver. 13,14 Patients with a smaller FLR are at a higher risk of developing PHLF. 15,16 However, FLR volume does not exactly reflect the liver functional reserve of the FLR, which might be impaired by underlying liver parenchymal diseases, such as fibrosis, cirrhosis or steatosis. To better predict post- operative outcomes, the FLR volume should be complemented with a method that assesses hepatic function. In this study, we compared the role of CP grade with ALBI grade in predicting post- operative outcomes. By combining the FLR and ALBI score, we hypothesized whether it could gain a better prognosis of post- operative morbidity and PHLF in HBV- associated HCC patients undergoing hepatic resection. 2 PATIENTS AND METHODS 2.1 Patients Between January 2015 and January 2017, all patients undergoing three- dimensional CT reconstruction prior to hepatectomy with curative intent for HCC at the Second Xiangya Hospital of Central South University were considered for this retrospective study. Patients who did not undergo three- dimensional CT reconstruction because of having no CT scan in our hospital were not enrolled. The inclusion criteria were as follows: open hepatectomy with curative intent performed by a single team of surgeons; no treatments for HCC before hepatectomy; and no cardiopulmonary, renal dysfunction or severe encephalopathy before hepatectomy. Informed consent for the clinical study was obtained from all patients, and this study was approved by the institutional review board of Central South University. 2.2 Diagnosis and definitions Diagnosis of HCC was based on pathological examination after surgery. The Barcelona Clinical Liver Cancer (BCLC) staging system was Key Points The incidence of post-operative major complications and PHLF increased with increasing ALBI grade; The ALBI grade classified patients with CP grade A into two subgroups with different incidences of PHLF; sflr and ALBI scores were identified as independent predictors of PHLF in multivariate logistic regression analysis; The combination of sflr and ALBI score was identified as a stronger predictor of post-operative morbidity and PHLF in HCC patients undergoing hepatic resection than the sflr or ALBI score alone. chosen as the HCC staging system. 17 PHLF was defined as a total serum bilirubin value >50 μmol L 1 and a prothrombin time index <50% (equal to international normalized ratio (INR) >1.7) on post- operative day 5 or thereafter, as proposed by the International Study Group of Liver Surgery (ISGLS). 18,19 Patients with grade A PHLF required no change in clinical management; those developing grade B PHLF required no- invasive treatment, such as administration of albumin, fresh- frozen plasma, daily diuretics and noninvasive ventilation; patients with grade C PHLF required invasive treatment including intubation and mechanical ventilation, circulatory support, haemodialysis and extracorporeal liver support. 19 Clinically significant portal pressure (CSPH) was defined as the presence of oesophageal varices or in those with a low platelet count (<100 10 9 /l) with splenomegaly. 20 A post- operative complication was defined according to the Dindo- Clavien classification. 21 A major complication was defined as grade 3 or above. Post- operative mortality was defined as death within 30 days after surgery. The CP score was calculated by five parameters: presence or absence of encephalopathy, ascites, serum total bilirubin, albumin concentration and prothrombin time. 22 According to the total score, the CP score can be divided into grade A, 5-6 points; grade B, 7-9 points and grade C, 10-15 points. The model for end- stage liver disease (MELD) score was calculated using the formula: 11.2 ln (international normalized ratio)+9.57 ln (creatinine, mg dl 1 )+3.78 ln (bilirubin, mg dl 1 )+6.43 (aetiology: 0 if cholestatic or alcoholic, 1 otherwise). 23 The ALBI score was calculated from the following formula: ALBI score=0.66 log 10 (total bilirubin [μmol l 1 ]) 0.085 (albumin [g L 1 ]), and was classified as grade 1 ( 2.60), grade 2 (> 2.60 to 1.39) or grade 3 (> 1.39). 5 2.3 Liver volumetry and calculation of the standardized future liver remnant Prior to surgery, all patients received a contrast- enhanced CT scan with a slice thickness of 0.9 mm. Automated volumetry of liver three- dimensional reconstruction was performed using medical image analysis software (Myrian XP Liver; Intrasense; France). The
ZOU et al. 3 results could be modified manually when needed. The gall bladder, hepatic inferior vena cava and main branches of the intrahepatic vascular structures were excluded from the reconstructional volume calculation, but the biliary structures were included. Preoperative virtual hepatic resection was performed according to the size and location of the tumours, which were evaluated by two experienced hepatic surgeons. The liver three- dimensional reconstruction and virtual hepatic resection are shown in Figure S1, supporting information. The future liver remnant volume was automatically calculated by the image analysis software in the computer. The standardized future liver remnant (sflr)=flr/estimated total liver volume (etlv) was used to represent the percentage of the liver that will remain after resection. 11 The etlv was calculated using the equation etlv (cm 3 )=706.2 BSA (m 2 )+2.4. 24 The BSA was calculated using body weight and height: BSA (m 2 )=0.00607 height (cm)+0.0127 weight (kg)- 0.0698 for men and BSA (m 2 )=0.00586 height (cm) +0.0126 weight (kg)- 0.0461 for women. The sflr ALBI was calculated as a combinatorial measure of sflr and ALBI. 2.4 Surgical technique In each patient, laparotomy was performed with a right subcostal incision. Intra- operative ultrasonography was routinely used to locate the tumours and confirm the relationships between tumours and vascular. Once resectability had been confirmed, segment- oriented anatomical resection was undertaken according to the position of the tumours. Low central venous pressure was routinely taken by the anaesthetist during surgery. The Pringle manoeuvre or selective vascular occlusion of the hemiliver was chosen if needed. Bipolar coagulation, absorbable clips and sutures were used for haemostasis. The extent of hepatic resection was defined according to the number of Couinaud liver segments. Major hepatectomy was defined as a resection of three or more Couinaud liver segments, 25 whereas minor resection was defined as a resection of fewer than three segments. 2.5 Follow- up Patients were followed up at 1 month after discharging from hospital and then every 3 months. In the follow- up, routine examinations were conducted, including liver function test, α- foetoprotein level, chest radiography, abdominal ultrasonography, abdominal CT or magnetic resonance image. Patients with recurrence received partial hepatic resection, radiofrequency ablation, transcatheter arterial chemoembolization or sorafenib treatment, according to the liver functional reserve, extent of the tumour and patients general performance status. The last date of follow- up was March 15, 2017. 2.6 Statistical analysis Continuous variables are expressed as medians (range) and were compared using the Mann- Whitney U test. Discrete variables were compared using the χ 2 test. Univariate analysis and multivariate logistic regression analysis were performed to identify risk factors associated with PHLF. The cut- off values for the occurrence of post- operative major complications and PHLF were determined by receiver operating characteristic (ROC) curve analysis. Statistical analyses were performed using SPSS 17.0 (SPSS, Inc., Chicago, IL, USA). P<.05 was considered to indicate a statistically significant difference. 3 RESULTS 3.1 Clinicopathological characteristics A total of 229 HCC patients were enrolled in the present study, comprising 197 men and 32 women, with a median age of 52 years (range, 18-74 years). The majority (93.0%) had hepatitis B virus infection. Cirrhosis occurred in 196 (85.6%) patients. Based on the BCLC model, 1.7% (4/229) of the patients was classified into 0, 60.7% (139/229) as A, 18.3% (42/229) as B and 19.2% (44/229) as C. Among BCLC C patients, three had lymph node metastasis and 41 had portal vein invasion. According to the CP grade, the majority of patients had grade A (210/229, 91.7%), and the remaining patients had grade B. According to the ALBI grade, 45.9% (105/229) of the patients were stratified into grade 1, 53.3% (122/229) as grade 2 and 0.9% (2/229) as grade 3. The relationship between CP grade and ALBI grade was presented in Table S1, supporting information. Major hepatic resections were performed in 62 (27.1%) patients. The median sflr measured prior to surgery was 0.637 (range, 0.361-0.995). Patient characteristics and operative variables are presented in Table 1. 3.2 Post- operative morbidity, PHLF and mortality Among the 229 patients, 102 (44.5%) experienced post- operative complications (Table 2). The most frequent complication was pleural effusion and ascites, which occurred in 29 (12.7%) patients, followed by pneumonia in 21 (9.2%). Seventy- eight (34.1%) patients developed minor complications, including grade I complications in 22 (9.6%) patients and grade II complications in 56 (24.5%) patients. Major complications occurred in 24 (10.5%) patients, including grade IIIa complications (n=16, 7.0%), grade IIIb complications (n=3, 1.3%), grade IVa complications (n=1, 0.4%), grade IVb complications (n=1, 0.4%) and grade V complications (n=3, 1.3%). Twenty- one (9.2%) patients developed PHLF. Among them, four (1.7%) patients experienced grade A PHLF, eight (3.5%) patients experienced grade B PHLF and nine (3.9%) patients experienced grade C PHLF. Three patients died as a result of PHLF within 30 days after surgery, with a post- operative mortality rate of 1.3%. A total of 25 patients died until the last follow- up date, with an overall mortality of 10.9%. 3.3 Correlations between Child- Pugh grade, ALBI grade and post- operative complications A total of 18 of the 210 (8.6%) CP grade A patients developed major complications and six of the 19 (31.6%) CP grade B patients developed
4 ZOU et al. TABLE 1 Patient characteristics Variables No. of patients (n=229) Age, y a 52 (18-74) Sex ratio (M/F) 197/32 Platelet count, x10 9 /l a 150 (28-535) Alanine aminotransferase, U/l a 34.5 (8.5-302.6) Total bilirubin, μmol/l a 13.4 (3.9-49.5) Albumin, g/l a 38.5 (23.6-48.9) Prothrombin time, s a 13.1 (9.9-17.7) Positive HBsAg b 213 (93.0) Tumour diameter, cm a 5.5 (1.5-18.0) Inflow occlusion b 132 (57.6) Blood loss, ml a 400 (40-6500) Operating time, min a 179 (85-353) Major hepatic resection b 62 (27.1) Cirrhosis b 196 (85.6) etlv, ml a 1220.4 (981.4-1495.6) FLR, ml a 784.4 (425.6-1342.4) sflr a 0.637 (0.361-0.995) MELD score a 7 (6 15) MELD score b 9 45 (19.7) <9 184 (80.3) Child- Pugh score a 5 (5-9) Child- Pugh grade b A 210 (91.7) B 19 (8.3) C 0 (0) ALBI score a ALBI grade b 2.519 ( 3.474-1.140) 1 105 (45.9) 2 122 (53.3) 3 2 (0.9) BCLC stage b 0 4 (1.7) A 139 (60.7) B 42 (18.3) C 44 (19.2) a Values are median (range). b Values are number (%). ALBI, albumin- bilirubin; BCLC, barcelona clinic liver Cancer; etlv, estimated total liver volume; FLR, future liver remnant; HBsAg, hepatitis B surface antigen; MELD, model for end- stage liver disease; sflr, standardized future liver remnant. major complications (P=.008). Nineteen of the 122 (15.6%) ALBI grade 2 patients developed major complications, but only three of the 105 (2.9%) ALBI grade 1 patients developed major complications (P=.001). sflr was significantly smaller in patients who experienced major complications with those without major complications (P=.01). ALBI was significantly greater in patients who experienced major complications compared with those without major complications (P<.001). However, no significant difference was found between patients who experienced major complications and those without when classified according to CP scores (P=.058). ROC curve analyses of CP score, ALBI score, sflr and sflr ALBI were performed to predict post- operative major complications (Figure 1A). The area under the ROC curve (AUC) for CP score, ALBI score, sflr and sflr ALBI were 0.600 (P=.11, 0.470-0.731), 0.756 (P<.001, 0.655-0.857), 0.660 (P=.01, 0.542-0.778) and 0.790 (P<.001, 0.684-0.897) respectively. ALBI score had a greater AUC for predicting post- operative major complications than that of the CP score, and sfl- R ALBI had the largest AUC in predicting major complications. 3.4 Univariate and multivariate analyses for PHLF Univariate logistic regression analyses revealed that CSPH, MELD score, sflr and ALBI score were associated with PHLF (P<.05) (Table 3). A multivariate logistic regression analysis was performed to identify risk factors for PHLF. Two different logistic regression analysis models for the CP score and ALBI score were undertaken to avoid collinearity. CSPH, MELD score and a small sflr were identified as significantly independent predictors of PHLF for the CP model, and CSPH, a small sflr and large ALBI score were identified as independent predictors of PHLF for the ALBI model (Table 3). 3.5 Correlations between Child- Pugh grade, ALBI grade and PHLF A total of 15 of the 210 (7.1%) CP grade A patients developed PHLF, and six of the 19 (31.6%) CP grade B patients developed PHLF (P=.004). According to the ALBI grade criteria, 105 patients were of ALBI grade 1, 122 patients were of ALBI grade 2 and two patients were of ALBI grade 3. Sixteen of the 122 (13.1%) ALBI grade 2 patients developed PHLF, but only three of the 105 (2.9%) ALBI grade 1 patients developed PHLF (P=.003). Both two ALBI grade 3 patients developed PHLF. In the CP grade A group, 105 (50.0%) patients were ALBI grade 1 patients and 105 (50.0%) patients were ALBI grade 2 patients. Patients with ALBI grade 2 had a higher incidence of PHLF than those with ALBI grade 1 (12 of 105 and three of 105; P=.029). 3.6 Defining the cut- off value for CP score, ALBI score, sflr and sflr ALBI for PHLF The area under the ROC curve values for the CP score, ALBI score, sflr and sflr ALBI for predicting PHLF were 0.646, 0.738, 0.758 and 0.884 respectively (Figure 1B). The sflr ALBI had the greatest AUC for predicting PHLF. The cut- off value of sflr for predicting PHLF was 53.8%, with 77.9% sensitivity and 61.9% specificity. Patients with sflr below 53.8% had a higher incidence of PHLF than patients with sflr of 53.8% or above: 13 (22.0%) of 59 and eight (4.7%) of 170 respectively (P<.001). The cut- off value of ALBI score
ZOU et al. 5 TABLE 2 Post- operative morbidity of HCC patients after hepatic resection Clavien- Dindo classification Complications I II IIIa IIIb IVa IVb V Total (%) Pleural effusion or ascites 21 6 2 29 (12.7) Urinary tract infection 3 3 (1.3) Pneumonia 18 2 1 21 (9.2) Upper gastrointestinal haemorrhage 2 2 (0.9) Arrhythmia 1 2 3 (1.3) Biliary leakage 6 1 7 (3.1) Wound infection 3 3 3 9 (3.9) Deep vein thrombosis 2 1 3 (1.3) Intra- abdominal abscess 2 2 (0.9) Intra- abdominal haemorrhage 4 4 (1.7) Small bowel obstruction 2 2 (0.9) Post- operative liver failure 8 5 1 3 17 (7.4) Total (%) 22 (9.6) 56 (24.5) 16 (7.0) 3 (1.3) 1 (0.4) 1 (0.4) 3 (1.3) 102 (44.5) FIGURE 1 Receiver operating characteristic (ROC) curve analyses of Child- Pugh score, ALBI score, sflr and sflr ALBI to predict (A) postoperative major complications and (B) PHLF. The area under the ROC curve values for the Child- Pugh score, ALBI score, sflr and sflr ALBI for predicting major complications were 0.600, 0.756, 0.660 and 0.790 respectively. The area under the ROC curve values for the Child- Pugh score, ALBI score, sflr and sflr ALBI for predicting PHLF were 0.646, 0.738, 0.758 and 0.884 respectively. ALBI, albumin- bilirubin; sflr, standardized future liver remnant. PHLF, post- hepatectomy liver failure for predicting PHLF was 2.28, with 81.7% sensitivity and 57.1% specificity. Patients with an ALBI score above 2.28 had a higher incidence of PHLF than patients with ALBI of 2.28 or below: 12 (24.0%) of 50 and nine (5.0%) of 179 respectively (P<.001). The cut- off value of sflr ALBI for predicting PHLF was 1.26, with 81.3% sensitivity and 90.5% specificity. Patients with sflr ALBI above 1.26 had a higher incidence of PHLF than patients with sflr ALBI of 1.26 or below: 19 (32.8%) of 58 and two (1.2%) of 171 respectively (P<.001). 3.7 Analyses of factors predicting PHLF in subgroups stratified by CSPH According to the definition of CSPH, 49 (21.4%) patients had CSPH and 180 (78.6%) patients had no CSPH. Univariate analysis showed that a small sflr and a greater ALBI score were associated with PHLF in non- CSPH subgroup (P<.05) (Table 4). On multivariable analyses, only sflr (P=.002) was an independent risk factor associated with PHLF for the CP model, and age (P=.047), sflr (P=.001) and ALBI score (P=.014) were identified as independent predictors of PHLF for the ALBI model in non- CSPH subgroup (Table 4). ROC curve analyses of CP score, ALBI score, sflr and sflr ALBI were performed to predict PHLF in subgroups stratified by CSPH (Figure 2). The AUC values for the CP score, ALBI score, sflr and sflr ALBI for predicting PHLF were 0.617 (P=.278), 0.706 (P=.056), 0.808 (P=.004) and 0.928 (P<.001) in CSPH subgroup respectively (Figure 2A). The AUC values for the CP score, ALBI score, sflr and sflr ALBI for predicting PHLF were 0.651 (P=.081), 0.751 (P=.004), 0.802 (P<.001) and 0.893 (P<.001) in non- CSPH subgroup respectively (Figure 2B). 3.8 Discriminative ability of sflr, ALBI score and CSPH for PHLF The function of sflr, ALBI score and CSPH in predicting PHLF was assessed in Figure 3. According to the distribution of PHLF, patients
6 ZOU et al. TABLE 3 Risk factors associated with post- hepatectomy liver failure Univariable logistic regression Multivariable logistic regression Child- Pugh model ALBI model Variables Odds ratio P Odds ratio P Odds ratio P Age (>60 y) 1.49 (0.55-4.07).435 2.34 (0.69-7.87).171 3.17 (0.86-11.70).083 Male sex 0.66 (0.21-2.11).484 0.58 (0.13-2.48).458 0.66 (0.13-3.31).617 ALT, U/l 2.36 (0.95-5.87).064 2.90 (0.95-8.93).063 2.65 (0.83-8.43).100 Prothrombin time, s 1.50 (0.61-3.70).383 1.00 (0.33-3.09).995 1.11 (0.33-3.73).872 Tumour diameter, cm 1.34 (0.53-3.37).534 0.75 (0.21-2.73).660 0.64 (0.16-2.52).522 Tumour number ( 3) 0.47 (0.11-2.11).325 0.23 (0.04-1.39).109 0.27 (0.04-1.61).149 CSPH 3.15 (1.24-7.99).016 5.59 (1.42-22.00).014 6.58 (1.51-28.78).012 MELD score 2.84 (1.10-7.35).031 4.95 (1.38-17.70).014 2.58 (0.65-10.25).180 Operating time, min 1.04 (0.42-2.56).931 1.18 (0.38-3.69).775 1.07 (0.31-3.75).910 Inflow occlusion 1.94 (0.73-5.21).186 1.89 (0.60-6.00).280 1.93 (0.54-6.87).309 Cirrhosis 1.01 (0.28-3.64).986 0.87 (0.19-4.02).861 0.54 (0.10-2.80).462 Blood loss, ml 1.72 (0.76-3.90).194 1.05 (0.33-3.39).931 0.78 (0.22-2.83).707 sflr 4.57 (1.81-11.50).001 11.02 (2.80-43.34).001 16.84 (3.76-75.37) <.001 Child- Pugh score 2.12 (0.86-5.24).102 1.53 (0.52-4.54).442 ALBI score 6.16 (2.42-15.69) <.001 8.82 (2.42-32.14) 0.001 ALBI: albumin- bilirubin; ALT, alanine aminotransferase; CSPH, clinically significant portal pressure; MELD, model for end- stage liver disease; sflr, standardized future liver remnant. TABLE 4 Risk factors associated with post- hepatectomy liver failure in non- CSPH patients (n=180) Univariable logistic regression Multivariable logistic regression Child- Pugh model ALBI model Variables Odds ratio P Odds ratio P Odds ratio P Age (>60 y) 1.83 (0.52-6.43).344 3.93 (0.83-18.59).085 5.57 (1.02-30.36).047 Male sex 1.92 (0.24-15.56).540 1.38 (0.12-15.93).796 1.74 (0.12-26.24).690 ALT, U/l 2.28 (0.69-7.47).175 2.40 (0.54-10.69).251 2.41 (0.53-10.91).254 Prothrombin time, s 0.65 (0.19-2.25).496 0.64 (0.15-2.76).551 0.82 (0.18-3.64).790 Tumour diameter, cm 1.99 (0.52-7.62).315 0.64 (0.11-3.78).622 0.60 (0.09-3.92).591 Tumour number ( 3) 0.39 (0.05-3.10).371 0.39 (0.04-3.70).410 0.52 (0.05-5.15).574 MELD score 2.10 (0.53-8.34).291 5.08 (0.88-29.33).069 2.33 (0.36-14.89).373 Operating time, min 2.15 (0.62-7.41).226 3.04 (0.57-16.33).195 3.04 (0.48-19.19).236 Inflow occlusion 2.09 (0.55-8.00).281 1.66 (0.33-8.30).540 1.40 (0.25-7.92).705 Cirrhosis 1.04 (0.21-5.01).958 1.24 (0.21-7.21).812 0.85 (0.13-5.57).868 Blood loss, ml 1.16 (0.39-3.40).793 0.34 (0.06-1.85).213 0.22 (0.03-1.45).114 sflr 8.46 (2.19-32.65).002 15.17 (2.64-87.09).002 25.27 (3.56-179.24).001 Child- Pugh score 2.17 (0.67-7.04).197 2.06 (0.48-8.82).330 ALBI score 5.00 (1.50-16.64).009 8.46 (1.54-46.62).014 ALBI: albumin- bilirubin; ALT, alanine aminotransferase; CSPH, clinically significant portal pressure; MELD, model for end- stage liver disease; sflr, standardized future liver remnant. with CSPH tended to need a higher sflr (Figure 3A) and patients with non- CSPH could tolerate a relatively higher ALBI score to avoid PHLF (Figure 3B). However, these results needed further studies to demonstrate. 3.9 Relationships of sflr and ALBI score with PHLF The relationships of the sflr and ALBI score with PHLF are shown in Figure 4. An oblique line was identified according to the distribution
ZOU et al. 7 FIGURE 2 Receiver operating characteristic (ROC) curve analyses of CP score, ALBI score, sflr and sflr ALBI to predict PHLF stratified by CSPH. (A) The area under the ROC curve values for the CP score, ALBI score, sflr and sflr ALBI for predicting PHLF were 0.617, 0.706, 0.808 and 0.928 in CSPH subgroup respectively. (B) The area under the ROC curve values for the CP score, ALBI score, sflr and sflr ALBI for predicting PHLF were 0.651, 0.751, 0.802 and 0.893 in non- CSPH subgroup respectively. CP, Child- Pugh; ALBI, albumin- bilirubin; sflr, standardized future liver remnant; PHLF, post- hepatectomy liver failure; CSPH: clinically significant portal pressure FIGURE 3 Usefulness of clinically significant portal pressure (CSPH), standardized future liver remnant (sflr) and albumin- bilirubin (ALBI) score in predicting post- hepatectomy liver failure FIGURE 4 Distribution of post- hepatectomy liver failure in patients with different ALBI scores and different sflrs. ALBI, albumin- bilirubin; sflr, standardized future liver remnant of PHLF patients. The incidence of PHLF above the line was 1.2% (2/172), and that below the line was 33.3% (19/57), showing a significant difference (P<.001). 4 DISCUSSION In the current study, we validated the value of CP score, ALBI score and sflr in predicting post- operative morbidity and PHLF in HBV- associated HCC patients who underwent hepatectomy. For years, the CP grade has been used to estimate the risk of hepatic resection. 2,3 Traditionally, CP grade A HCC patients were often regarded as suitable candidates for hepatic resection. However, in clinical practice, some CP grade A patients may already have underlying liver functional impairment with decreased serum albumin, obviously rising bilirubin or ascites. PHLF accidentally occurs in some CP grade A patients as in our study, revealing that the heterogeneity of liver functional reserve may exist in CP grade A patients. Recent studies have shown that dissection of CP grade A by ALBI grade was of great prognostic importance in intermediate or advanced stage of HCC. 26-28 Our results showed that the CP grade A patients could be successfully reclassified into two subgroups according to ALBI grade and that the incidence of PHLF in the ALBI grade 2 subgroup was higher compared with those with grade 1 (P=.029). These data revealed that the ALBI grade served as a useful predictor of PHLF to reclassify CP grade and discriminated those patients with worse liver functional reserve in a CP grade. In addition, our results showed that the incidence of major complications and PHLF for ALBI grade 2 was significantly larger than ALBI grade 1. This was in accordance with the recent study reported by Andreatos N. 29 With a very low incidence of major complications and PHLF, patients with ALBI grade 1 appeared to show a better tolerance to hepatic resection than those with CP grade A. ROC curve analysis revealed that ALBI score had a larger AUC for predicting postoperative major complications and PHLF than that of the CP score, indicating that the ALBI score was a stronger predictor of post- operative major complications and PHLF compared with CP score. Moreover,
8 ZOU et al. multivariate logistic regression analysis showed that the ALBI score was an independent predictor of PHLF, but the CP score was not, further demonstrating that the ALBI score was a superior predictor of PHLF than the CP score. CSPH has been reported to be closely associated with postoperative outcomes. 20,30 This was consistent with our results. We found CSPH was an independent predictor of PHLF both in CP model and ALBI model. By integrating CSPH with sflr and ALBI score, it was shown that patients with CSPH tended to need a higher sflr and patients with non- CSPH could tolerate a relatively higher ALBI score to avoid PHLF, this might help to decrease the risk of developing PHLF. When analysing stratified by CSPH, only ALBI score was found to be an independent predictor of PHLF, but the CP score was not in non- CSPH subgroup. Additionally, ROC curve analyses revealed that ALBI score had a larger AUC for predicting PHLF than that of the CP score both in CSPH and non- CSPH subgroups. All these results further verified that the ALBI grade was a more accurate predictor of PHLF than CP grade. The post- operative remnant liver quality and volume were the determinants of post- operative outcomes. More recently, the importance of pre- operative volumetric analysis for major hepatic resection has been increasingly valued. FLR has been shown to be a very important predictor of PHLF. 15,31,32 Our results showed that sflr was an independent risk factor for PHLF regardless of using the CP model or ALBI model. ROC curve analysis revealed that the cut- off value of sflr for predicting PHLF was 53.8%, which was larger than most of the previous studies, 31,33,34 probably because 92.6% of the patients had hepatitis B virus infection, and 85.6% of the patients had Cirrhosis in the present study. A shortcoming of volumetry is the fact that volumetric assessment of the remnant liver does not consider the quality of the remnant liver parenchyma and is therefore not reliable in predicting PHLF in patients with underlying liver diseases. As in the present study, two of the 21 patients who developed PHLF had a sflr of more than 70%. It was strongly recommended to give special consideration to liver function for this type of patient. Therefore, we used sflr ALBI as the combination of remnant liver quality and volume. ROC curve analyses revealed that sflr ALBI had a larger AUC for predicting post- operative major complications and PHLF than sflr or ALBI alone and the cut- off value of sflr ALBI could predict postoperative major complications and PHLF more accurately than sflr or ALBI alone. Besides, we got the same results both in CSPH and non- CSPH subgroups. This indicates that sflr ALBI is a more accurate predictor of post- operative morbidity and hepatic function than sflr or ALBI. Theoretically, with a larger ALBI score, a larger sflr is required to avoid PHLF. Here, we designed a method to analyse the PHLF distribution together with sflr and ALBI score and found an oblique line (Figure 4). The incidence of PHLF below the line was significantly higher than above. Patients who were located below the line should be given special considerations by hepatic surgeons before an operation and may be not ideal candidates for hepatic resection. The current study has a number of limitations. Firstly, all the subjects in the study were enrolled from a single study centre and a multicentre study recruiting more patients is required. In addition, most of the patients in the study were infected with HBV. It is possible that the results do not apply to HCC patients in other countries where HCV infection or alcoholic cirrhosis may be the most common cause of HCC. Therefore, these study results require validation in western patient populations. Finally, as the follow- up time was short, we could not analyse the relationship between ALBI grade and patient survival. This is the work we need to do in the future. In conclusion, the ALBI score showed better prediction of post- operative outcomes than the CP score, and the combination of sflr and ALBI has the advantage of predicting post- operative morbidity and PHLF in HBV- associated HCC patients undergoing hepatic resection than sflr or ALBI alone. CONFLICT OF INTEREST None declared. FINANCIAL SUPPORT None. REFERENCES 1. Torre LA, Bray F, Siegel RL, et al. Global cancer statistics, 2012. CA Cancer J Clin. 2015;65(2):87 108. 2. Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet. 2012;379:1245 1255. 3. Bruix J, Reig M, Sherman M. Evidence- based diagnosis, staging, and treatment of patients with hepatocellular carcinoma. Gastroenterology. 2016;150(4):835 853. 4. Clavien PA, Petrowsky H, DeOliveira ML, Graf R. Strategies for safer liver surgery and partial liver transplantation. N Engl J Med. 2007;356:1545 1559. 5. Johnson PJ, Berhane S, Kagebayashi C, et al. Assessment of liver function in patients with hepatocellular carcinoma: a new evidence- based approach- the ALBI grade. J Clin Oncol. 2015;33(6):550 558. 6. Kaplan DE, Dai F, Aytaman A, et al. Development and performance of an algorithm to estimate the Child- Turcotte- Pugh score from a national electronic healthcare database. Clin Gastroenterol Hepatol. 2015;13(13):2333 2341. 7. Wang YY, Zhong JH, Su ZY, JF H. Albumin- bilirubin versus Child- Pugh score as a predictor of outcome after liver resection for hepatocellular carcinoma. Br J Surg. 2016;103:725 734. 8. Chan AW, Chan RC, Wong GL, et al. New simple prognostic score for primary biliary cirrhosis: albumin- bilirubin score. J Gastroenterol Hepatol. 2015;30(9):1391 1396. 9. Hiraoka A, Kumada T, Michitaka K, et al. Usefulness of albuminbilirubin grade for evaluation of prognosis of 2584 Japanese patients with hepatocellular carcinoma. J Gastroenterol Hepatol. 2016;31(5):1031 1036. 10. Li MX, Zhao H, Bi XY, et al. Prognostic value of the albumin- bilirubin grade in patients with hepatocellular carcinoma: validation in a Chinese cohort. Hepatol Res. 2016;47:731 741. 11. Cieslak KP, Runge JH, Heger M, et al. New perspectives in the assessment of future remnant liver. Dig Surg. 2014;31:255 268. 12. Hoekstra LT, de Graaf W, Nibourg GAA, et al. Physiological and biochemical basis of clinical liver function tests: a review. Ann Surg. 2013;257:27 36. 13. Wigmore SJ, Redhead DN, Yan XJ, et al. Virtual hepatic resection using three- dimensional reconstruction of helical computed tomography angioportograms. Ann Surg. 2001;233:221 226.
ZOU et al. 9 14. Simpson AL, Geller DA, Hemming AW, et al. Liver planning software accurately predicts postoperative liver volume and measures early regeneration. J Am Coll Surg. 2014;219(2):199 207. 15. Kim HJ, Kim CY, Park EK, et al. Volumetric analysis and indocyanine green retention rate at 15 min as predictors of post- hepatectomy liver failure. HPB. 2015;17(2):159 167. 16. Kishi Y, Abdalla EK, Chun YS, et al. Three hundred and one consecutive extended right hepatectomies. Ann Surg. 2009;250:540 548. 17. Bruix J, Sherman M. Management of hepatocellular carcinoma: an update. Hepatology. 2011;53(3):1020 1022. 18. 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(6):824 828. 19. Rahbari NN, Garden OJ, Padbury R, et al. Posthepatectomy liver failure: a definiction and grading by the International Study Group of Liver Surgery(ISGLS). Surgery. 2011;149(5):713 724. 20. Chen X, Zhai J, Cai X, et al. Severity of portal hypertension and prediction of postoperative liver failure after liver resection in patients with Child- Pugh grade A cirrhosis. Br J Surg. 2012;99(12):1701 1710. 21. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240(2):205 213. 22. Pugh RN, Murray-Lyon IM, Dawson JL, et al. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg. 1973;60(8):646 649. 23. Patrick S, Kamath RH, Michael MM, et al. A model to predict survival in patients with end- stage liver disease. Hepatology. 2001;33:464 470. 24. Urata K, Kawasaki S, Matsunami H, et al. Calculation of child and adult standard liver volume for liver transplantation. Hepatology. 1995;21(5):1317 1321. 25. Mullen JT, Ribero D, Reddy SK, et al. Hepatic insufficiency and mortality in 1059 noncirrhotic patients undergoing magor hepatectomy. J Am Coll Surg. 2007;204(5):854 862. 26. Pinato DJ, Yen C, Bettinger D, et al. The albumin bilirubin grade improves hepatic reserve estimation post- sorafenib failure: implications for drug development. Aliment Pharmacol Ther. 2017;45(5):714 722. 27. Pinato DJ, Sharma R, Allara E, et al. The ALBI grade provides objective hepatic reserve estimation across each BCLC stage of hepatocellular carcinoma. J Hepatol. 2017;66(2):338 346. 28. Edeline J, Blanc J-F, Johnson P, et al. A multicentre comparison between Child Pugh and albumin- Bilirubin scores in patients treated with sorafenib for hepatocellular carcinoma. Liver Int. 2016;36(12):1821 1828. 29. Andreatos N, Amini N, Gani F, et al. Albumin- Bilirubin score: predicting short- term outcomes including bile leak and post- hepatectomy liver failure following hepatic resection. J Gastrointest Surg. 2017;21(2):238 248. 30. Allard MA, Adam MR, Bucur PO, et al. Posthepatectomy portal vein pressure predicts liver failure and mortality after major liver resection on noncirrhotic liver. Ann Surg. 2013;258:822 830. 31. Narita M, Oussoultzoglou E, Fuchshuber P, et al. What is a safe future liver remnant size in patients undergoing major hepatectomy for colorectal liver metastases and treated by intensive preoperative chemotherapy? Ann Surg Oncol. 2012;19:2526 2538. 32. Yokoyama Y, Nishio H, Ebata T, et al. Value of indocyanine green clearance of the future liver remnant in predicting outcome after resection for biliary cancer. Br J Surg. 2010;97:1260 1268. 33. Truant S, Boleslawski E, Sergent G, et al. Liver function following extended hepatectomy can be accurately predicted using remnant liver volume to body weight ratio. World J Surg. 2015;39(5):1193 1201. 34. Kishi Y, Abdalla EK, Chun YS, et al. Three hundred and one consecutive extended right hepatectomies: evaluation of outcome based on systematic liver volumetry. Ann Surg. 2009;250(4):540 548. SUPPORTING INFORMATION Additional Supporting Information may be found online in the supporting information tab for this article. How to cite this article: Zou H, Wen Y, Yuan K, Miao X-Y, Xiong L, Liu K-J. Combining albumin- bilirubin score with future liver remnant predicts post- hepatectomy liver failure in HBV- associated HCC patients. Liver Int. 2017;00:1 9. https:// doi.org/10.1111/liv.13514
本文献由 学霸图书馆 - 文献云下载 收集自网络, 仅供学习交流使用 学霸图书馆 (www.xuebalib.com) 是一个 整合众多图书馆数据库资源, 提供一站式文献检索和下载服务 的 24 小时在线不限 IP 图书馆 图书馆致力于便利 促进学习与科研, 提供最强文献下载服务 图书馆导航 : 图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具