Introduction

Hepatitis B virus (HBV) infection is related to 70–90% of the patients with hepatocellular carcinoma (HCC) in the Asia-Pacific regions, especially China.1 Partial hepatectomy is the preferred curative means in select HBV-related HCC patients.2,3 Although advances in hepatectomy and perioperative care techniques have greatly improved the safety of surgery, postoperative major complications, especially severe posthepatectomy liver failure (PHLF) induced by residual hepatic functional insufficiency, remain the major cause of postoperative death.4–8 Thus, it is of great significance to estimate liver function reserve prior to hepatectomy.

Currently, the Child–Pugh scoring system is the most commonly applied method to assess liver function reserve; however, its clinical applications is limited due to the use of two subjective and arbitrary indexes (hepatic encephalopathy and ascites) in its calculations.9,10 The model for end-stage liver disease (MELD), originally established to estimate the outcomes of cirrhotic patients, has been gradually recognized as a standard for assessing liver function reserve and sequencing transplant candidates. Nevertheless, the level of serum creatinine is strongly influenced by individual reasons, such as gender and age, leading to its limited application.11 The albumin–bilirubin (ALBI) score is the most recently recognized model for assessing hepatic functional reserve and is often used to predict the prognostic risk of different liver diseases, but it is still limited to accurately assess patients with obstructive jaundice.12 Therefore, there is still a need to explore better tools to estimate liver reserve function.

Indocyanine green (ICG) is a water-soluble fluorescent dye that binds to lipoprotein and albumin and excretes bile as it is after intravenous injection.13,14 As a quantitative excretory hepatic functional method to assess functional hepatocytes and liver blood flow, the ICG retention test at 15 minutes (ICG‐R15) became a standard preoperative parameter to evaluate liver function reserve in patients with different hepatic diseases, mostly in Asian series.15–18

In this study, we compared the abilities of ICG-R15, Child–Pugh, MELD and ALBI scores for assessing postoperative major complications and severe PHLF risk.

Methods Patient Population

In this study, 354 patients who were subjected to initial hepatectomy for HBV-related HCC between January 2017 and December 2018 in our hospital were included. HCC patients who received radiofrequency ablation, transarterial chemoembolization or other treatments for tumors prior to liver resection were excluded. This study was conducted with the written informed consent of each patient and approved by the Ethics Committee of Guangxi Medical University Cancer Hospital, as well as in accordance with the Helsinki Declaration.

Diagnosis and Definitions

Postoperative pathological examination was the basis for the diagnosis of HCC, and Barcelona Clinical Liver Cancer (BCLC) criterion was selected as the HCC stage. Splenomegaly or gastroesophageal varices with thrombocytopenia was defined as clinically significant portal hypertension (CSPH).19 Patients with hyperbilirubinemia and abnormal coagulation on postoperative day 5 was defined as PHLF. Grade A PHLF not needed any specific therapy, grade B PHLF not needed invasive treatments, and grade C PHLF needed invasive therapies. Among them, grade B or above PHLF was defined as severe PHLF.20 The severity of postoperative complications was classified based on the Dindo–Clavien grade, and grade III and above was defined as postoperative major complications.21

ICG Clearance

Generally, ICG clearance is performed using a continuous infusion technique during hepatic vein intubation. All enrolled patients in our study were received ICG clearance test prior to hepatectomy. After fasting overnight, an appropriate amount of ICG was quickly injected through a peripheral vein of forearm. Plasma ICG concentration was monitored by an optical probe connected to the patient, and the ICG-R15 value was measured by a pulsed dye density map analyzer (DDG3300K, Japan).

Hepatectomy and Follow-Up

Before hepatectomy, abdominal CT or MRI was carried out to estimate cancer situation and surgical safety. The Child–Pugh scoring system and residual hepatic volume were measured to assess hepatic function reserve. The surgical treatment of liver tumors were based on segmental anatomical resection. The extent of hepatectomy can be divided into major resection (removal of three or more Couinaud segments) or minor resection (removal of one or two segments or wedge resection) based on the number of liver segments resected.22 More details and indications of liver resection procedures were described in previous research.23

All patients were routinely reviewed 1 month after discharge, every 2–3 months in the first postoperative year, and every 3–6 months in the second year. Routine re-examinations include serum biochemistry, α‐fetoprotein, abdominal ultrasonography, CT or MRI, and so on.

Statistical Analyses

Categorical variables were shown as frequencies and proportions and were compared using χ2 test. Continuous variables were shown as median (Q25-Q75) and were compared using Mann–Whitney U-tests.

Using univariate and multivariate logistic regression analyses, we confirmed independent risk parameters that predicted postoperative major complications and severe PHLF. Predictive abilities of Child–Pugh, MELD, ALBI and ICG‐R15 to predict postoperative major complications and severe PHLF were tested via the areas under the receiver-operating characteristic (ROC) curves (AUCs) and decision curve analysis (DCA).24 Additionally, three risk groups were generated by splitting its linear predictor at the 50th and 85th percentiles of ICG-R15. The low-risk group was less than 50%, the intermediate risk group was between the 50th and 85th percentiles, and the last 15% was high-risk.

SPSS software (version 25.0, IBM, USA) was used for statistical analyses. P < 0.05 was considered to be statistically significant.

Results Patients’ Characteristics

The clinical characteristics of 354 HBV-related HCC patients enrolled are shown in Table 1. The patients included 36 females and 318 males with a median age of 51 years. And, 9.4% of the patients suffered from CSPH, while most patients (60.2%) had cirrhosis. Moreover, most patients (86.2%) were categorized as Child–Pugh grade A, and the rest patients was grade B. The median MELD was 5 (4 to 7), the median ALBI was −2.38 (−2.59 to −2.16), and the median ICG‐R15 was 4.6 (3.2 to 7.35).

Table 1 Baseline Characteristics of the Included 354 Patients

Based on the BCLC grade system, 3.4% of the patients were grade 0, 57.9% were grade A, 20.3% were grade B, and 18.4% were grade C. The surgical resection included 235 major hepatectomy and 117 minor hepatectomy.

Postoperative Complications

Of the 354 patients, 199 patients (56.2%) had postoperative complications (Supplementary Table 1). The most postoperative complication was ascites or pleural effusion in 115 cases (32.5%), followed by PHLF in 109 cases (30.8%). Among them, 32 patients (9.1%) developed postoperative major complications, and 109 patients (30.8%) developed PHLF: (grade A: 14.7% [n = 52]; grade B: 15.0% [n = 53]; and grade C: 1.1% [n = 4]), of whom 57 patients (16.1%) developed severe PHLF.

Independent Predictors of Postoperative Major Complications

Factors related to postoperative major complications in univariate logistic regression analyses, included male, prealbumin, albumin, aspartate aminotransferase, creatinine, Child–Pugh, MELD, ALBI, ICG-R15, tumor size, blood loss and major resection (Table 2, P < 0.05 for all). For multivariate analysis, aspartate aminotransferase, ICG-R15 and major hepatectomy were confirmed as independent predictors of postoperative major complications in HBV-related HCC patients (Table 2, P < 0.05 for all).

Table 2 Univariate and Multivariate Analyses to Identify Factors Predicting Postoperative Major Complications

Independent Predictors of Severe PHLF

Univariate logistic regression analyses indicated prothrombin time, prealbumin, albumin, CSPH, cirrhosis, Child–Pugh, MELD, ALBI, ICG-R15, tumor size, portal invasion or extrahepatic spread and major hepatectomy were related to severe PHLF (Table 3, P < 0.05 for all). Then, in a multivariate analysis, prothrombin time, cirrhosis, ICG-R15 and major hepatectomy were identified as independent predict variables of severe PHLF in HBV-related HCC patients (Table 3, P < 0.05 for all).

Table 3 Univariate and Multivariate Analyses to Identify Factors Predicting Severe PHLF

Discriminative Abilities of the Models for Major Complications

The AUC of the ICG-R15 (AUC 0.789, 95% confidence interval (c.i.) 0.707 to 0.872) for predicting postoperative major complications was higher than the Child–Pugh (AUC 0.619, 95% c.i. 0.515 to 0.723), MELD (AUC 0.617, 95% c.i. 0.516 to 0.721) and ALBI (AUC 0.666, 95% c.i. 0.561 to 0.771) scores (Figure 1A, P < 0.05 for all). Furthermore, the DCA plot showed that ICG-R15 has a better net benefit and a wider threshold possibilities in assessing postoperative major complications (Figure 1B). Accordingly, the ICG-R15 was superior in estimating postoperative major complications risk.

Figure 1 (A) ROC curves and (B) DCA plot analyses of ICG‐R15, Child–Pugh, MELD and ALBI scores for assessing postoperative major complications.

Abbreviations: ICG‐R15, indocyanine green retention test at 15 minutes; MELD, model for end‐stage liver disease; ALBI, albumin–bilirubin; ROC, receiver operating characteristic; DCA, decision curve analysis.

Discriminative Abilities of the Models for Severe PHLF

The AUC of the ICG-R15 (AUC 0.823, 95% c.i. 0.775 to 0.871) to predict severe PHLF was remarkably higher than Child–Pugh (AUC 0.641, 95% c.i. 0.564 to 0.718), MELD (AUC 0.604, 95% c.i. 0.518 to 0.690) and ALBI (AUC 0.691, 95% c.i. 0.612 to 0.769) scores (Figure 2A, P < 0.05 for all). In addition, the DCA plot also indicated that ICG-R15 has a better net benefit and a wider threshold possibilities in predicting severe PHLF (Figure 2B). Thus, ICG-R15 also showed a significant advantage in predicting severe PHLF.

Figure 2 (A) ROC curves and (B) DCA plot analyses of ICG‐R15, Child–Pugh, MELD and ALBI scores for assessing severe PHLF.

Abbreviations: ICG‐R15, indocyanine green retention test at 15 minutes; MELD, model for end‐stage liver disease; ALBI, albumin–bilirubin; ROC, receiver operating characteristic; DCA, decision curve analysis; PHLF, posthepatectomy liver failure.

Subgroup Analyses

Subgroup analyses were performed according to the cirrhosis conditions, intraoperative status (hepatectomy, blood loss and blood transfusion), and tumor stage. In all subgroups, the AUCs values of ICG-R15 in predicting major postoperative complications (Figure 3 and Supplementary Table 2; P < 0.05 for all) and severe PHLF (Figure 4 and Supplementary Table 3; P < 0.05 for all) were greatly higher than the other scoring systems.

Figure 3 ROC curves of ICG‐R15, Child–Pugh, MELD and ALBI scores for assessing postoperative major complications in the HBV-related HCC patients subgroups. Subgroups include (A) cirrhosis, (B) no cirrhosis, (C) major hepatectomy, (D) minor hepatectomy, (E) blood loss ≥400 mL, (F) blood loss <400 mL, (G) blood transfusion, (H) no blood transfusion, (I) BCLC-0 or -A stage, and (J) BCLC-B or -C stage.

Abbreviations: ICG‐R15, indocyanine green retention test at 15 minutes; MELD, model for end‐stage liver disease; ALBI, albumin–bilirubin; ROC, receiver operating characteristic; PHLF, posthepatectomy liver failure; BCLC, Barcelona Clinical Liver Cancer; HBV, hepatitis B virus; HCC, hepatocellular carcinoma.

Figure 4 ROC curves of ICG‐R15, Child–Pugh, MELD and ALBI scores for assessing severe PHLF in the HBV-related HCC patients. Subgroups include (A) cirrhosis, (B) no cirrhosis, (C) major hepatectomy, (D) minor hepatectomy, (E) blood loss ≥400 mL, (F) blood loss <400 mL, (G) blood transfusion, (H) no blood transfusion, (I) BCLC-0 or -A stage, and (J) BCLC-B or -C stage.

Abbreviations: ICG‐R15, indocyanine green retention test at 15 minutes; MELD, model for end‐stage liver disease; ALBI, albumin–bilirubin; ROC, receiver operating characteristic; PHLF, posthepatectomy liver failure; BCLC, Barcelona Clinical Liver Cancer; HBV, hepatitis B virus; HCC, hepatocellular carcinoma.

Application of the ICG-R15 in Patients Risk Stratification

The 50th percentile of ICG-R15 was 4.6%, and 85th percentile was 9.9%. Then, three risk groups were generated (low-risk ≤4.6%, intermediate-risk 4.6–9.9%, and high-risk >9.9%). The incidence of postoperative major complications and severe PHLF was significantly different among all enrolled patients in the ICG-R15 risk subgroups (Figure 5 and Supplementary Table 4; P <0.05 for all). Moreover, similar findings were yielded for all the HCC patients’ subgroup analyses that assessed postoperative major complications (Supplementary Figure 1 and Supplementary Table 4; P <0.05 for all) and severe PHLF (Supplementary Figure 2 and Supplementary Table 4; P <0.05 for all).

Figure 5 Relationship between the incidence of (A) postoperative major complications and (B) severe PHLF based upon risk group stratification assessed using the ICG-R15 in all included HBV-related HCC patients.

Abbreviations: PHLF, posthepatectomy liver failure; ICG‐R15, indocyanine green retention test at 15 minutes; HBV, hepatitis B virus; HCC, hepatocellular carcinoma.

Discussion

In this research, we compared the differences of four methods (Child–Pugh, MELD, ALBI and ICG-R15) in assessing postoperative major complications and severe PHLF in HBV-related HCC patients after hepatectomy. We found that ICG-R15 was an independent predictor of postoperative major complications and severe PHLF, and the predictive abilities of ICG‐R15 wwere greatly higher than other scoring systems. Furthermore, the ICG‐R15 also has great advantages in predicting postoperative major complications and severe PHLF in subgroup analyses based on cirrhosis condition, intraoperative status (hepatectomy, blood loss and blood transfusion), and tumor stage. In addition, the incidence of postoperative major complications and severe PHLF risk also increased with ICG-R15-based risk stratification.

PHLF is the most serious complication after hepatectomy and may lead to death of patients.4–8 To reduce the risk of postoperative major complications and severe PHLF, it is of great significance to estimate hepatic functional reserve prior to surgery. Commonly, the Child–Pugh, MELD and ALBI scores are three applied tools for hepatic functional reserve assessment, but they all have obvious defects that limit their wide clinical application.9–12 Recently, with the development of noninvasive pulse spectrophotometers, ICG-R15 test have became a standard preoperative parameter to assess liver function reserve is possible prior to hepatectomy in patients with sepsis in intensive care units, hepatosteatosis, acute hepatitis, or receiving chemotherapy.13–18 However, it is not clear which of the four mentioned models is the optimal method to assess liver function reserve in HBV-related HCC patients prior to hepatectomy. To solve this issue, we first carried out univariate logistic regression analyses to find indicators for predicting postoperative major complications and severe PHLF. As expected, all four mentioned methods showed significant differences in predicting major postoperative complications and severe PHLF alone. However, only ICG-R15 of the four methods can be used as an independent predictor of postoperative major complications and severe PHLF when the multivariate logistic analysis of other factors is taken into account. These findings preliminarily revealed that ICG‐R15 is a better predictor of postoperative major complications and severe PHLF than other models. Furthermore, the ROC curve analyses showed that ICG-R15 had higher AUCs for predicting postoperative major complications and severe PHLF compared to the other three models, and DCA plots suggest that ICG-R15 had a better net benefit and a wider range of threshold possibilities in predicting postoperative major complications and severe PHLF. These results further verified that ICG-R15 has significantly higher predictive power than the other three models in assessing postoperative major complications and severe PHLF.

In addition, many studies have shown that liver cirrhosis background, intraoperative status (hepatectomy, blood loss and blood transfusion) and tumor stage were also independent predictors for assessing postoperative complications.6,19,25 In our research, only major hepatectomy has always been an independent risk parameter for predicting major complications and severe PHLF, while cirrhosis was only an independent predictor for severe PHLF. Then, according to these different subgroups, we continued to compare the predictive ability of those mentioned four methods to assess postoperative major complications and severe PHLF. Surprisingly, in all the subgroup analyses, the ICG-R15 showed stable and satisfactory predictive performance in assessing postoperative major complications and severe PHLF and was superior to the other three models.

On the basis of risk stratification, this study further analyzed the relationship between ICG-R15 and postoperative major complications and severe PHLF. Our study showed that the incidence of postoperative major complications and severe PHLF differed significantly among the three risk groups. Unsurprisingly, the incidence of major postoperative complications and severe PHLF was greatly higher in the high-risk cohort than in the other two groups. Therefore, from our results, it can be concluded that hepatectomy should be carefully selected for high-risk population.

However, there are also some limitations in our research. Firstly, all included patients were HBV-related HCC patients, and other etiologies, such as hepatitis C virus or alcoholic liver disease, still need to be studied. Moreover, this is a retrospective and single-center project, and further larger and multicentric researches are required to verify our findings.

Conclusion

Compared with Child–Pugh, MELD and ALBI scores, preoperative ICG-R15 can more accurately predict the postoperative major complications and severe PHLF risk after hepatectomy in HBV-related HCC patients.

Abbreviations

HBV, hepatitis B virus; HCC, hepatocellular carcinoma; PHLF, posthepatectomy liver failure; MELD, model for end-stage liver disease; ALBI, albumin–bilirubin; ICG-R15, indocyanine green retention test at 15 min; AUC, area under the operating characteristic curve; DCA, decision curve analysis; BCLC, Barcelona Clinical Liver Cancer.

Data Sharing Statement

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Ethics Approval and Consent to Participate

The study was conducted in compliance with the Helsinki Declaration and approved by the institutional Ethics Committee of Guangxi Medical University Cancer Hospital, and all patients provided written informed consent.

Author Contributions

RYM and TB contributed equally to this work. All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Funding

The study was supported by the Natural Science Foundation of Guangxi (NO. 8186110284) and Guangxi Traditional Chinese Medicine AppropriateTechnology Development and Promotion Project (GZSY20‑18).

Disclosure

The authors declare that they have no competing interests.

References

1. Zhou J, Sun H, Wang Z, et al. Guidelines for diagnosis and treatment of primary liver cancer in China (2017 Edition). Liver Cancer. 2018;7:235–260. doi:10.1159/000488035

2. Heimbach JK, Kulik LM, Finn RS, et al. AASLD guidelines for the treatment of hepatocellular carcinoma. Hepatology. 2018;67(1):358–380. doi:10.1002/hep.29086

3. Galle PR, Forner A, Llovet JM, Mazzaferro V, Piscaglia F, Raoul JL. EASL clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol. 2018;69(1):182–236. doi:10.1016/j.jhep.2018.03.019

4. 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(5):713–724. doi:10.1016/j.surg.2010.10.001

5. 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):824–8, discussion 828–9. doi:10.1097/01.sla.0000189131.90876.9e

6. 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(4):698–708; discussion 708–10. doi:10.1097/01.sla.0000141195.66155.0c

7. Wei AC, Tung-Ping PR, Fan ST, Wong J. Risk factors for perioperative morbidity and mortality after extended hepatectomy for hepatocellular carcinoma. Br J Surg. 2003;90(1):33–41. doi:10.1002/bjs.4018

8. Donadon M, Costa G, Cimino M, et al. Safe hepatectomy selection criteria for hepatocellular carcinoma patients: a validation of 336 consecutive hepatectomies. The BILCHE score. World J Surg. 2015;39(1):237–243. doi:10.1007/s00268-014-2786-6

9. Durand F, Valla D. Assessment of the prognosis of cirrhosis: Child-Pugh versus MELD. J Hepatol. 2005;42(1):S100–S107. doi:10.1016/j.jhep.2004.11.015

10. 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. doi:10.1200/JCO.2014.57.9151

11. Wiesner R, Edwards E, Freeman R, et al. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology. 2003;124(1):91–96. doi:10.1053/gast.2003.50016

12. Wang YY, Zhong JH, Su ZY, et al. Albumin-bilirubin versus Child-Pugh score as a predictor of outcome after liver resection for hepatocellular carcinoma. Br J Surg. 2016;103(6):725–734. doi:10.1002/bjs.10095

13. Cherrick GR, Stein SW, Leevy CM, Davidson CS. Indocyanine green: observations on its physical properties, plasma decay, and hepatic extraction. J Clin Invest. 1960;39:592–600. doi:10.1172/JCI104072

14. Burczynski FJ, Pushka KL, Sitar DS, Greenway CV. Hepatic plasma flow: accuracy of estimation from bolus injections of indocyanine green. Am J Physiol. 1987;252(5 Pt 2):H953–62. doi:10.1152/ajpheart.1987.252.5.H953

15. Kokudo T, Hasegawa K, Kokudo N. Assessment of preoperative liver function based on indocyanine green clearance. Hepatology. 2017;66(2):675–676. doi:10.1002/hep.29232

16. Yamada A, Hara T, Li F, et al. Quantitative evaluation of liver function with use of gadoxetate disodium-enhanced MR imaging. Radiology. 2011;260(3):727–733. doi:10.1148/radiol.11100586

17. Ibis C, Albayrak D, Sahiner T, Soytas Y, Gurtekin B, Sivrikoz N. Value of preoperative indocyanine green clearance test for predicting post-hepatectomy liver failure in noncirrhotic patients. Med Sci Monit. 2017;23:4973–4980. doi:10.12659/msm.907306

18. Schwarz C, Plass I, Fitschek F, et al. The value of indocyanine green clearance assessment to predict postoperative liver dysfunction in patients undergoing liver resection. Sci Rep. 2019;9(1):8421. doi:10.1038/s41598-019-44815-x

19. Berzigotti A, Reig M, Abraldes JG, Bosch J, Bruix J. Portal hypertension and the outcome of surgery for hepatocellular carcinoma in compensated cirrhosis: a systematic review and meta-analysis. Hepatology. 2015;61(2):526–536. doi:10.1002/hep.27431

20. Wang YY, Xiang BD, Ma L, et al. Development and validation of a nomogram to preoperatively estimate post-hepatectomy liver dysfunction risk and long-term survival in patients with hepatocellular carcinoma. Ann Surg. 2020;274(6):e1209–e1217. doi:10.1097/SLA.0000000000003803

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. doi:10.1097/01.sla.0000133083.54934.ae

22. Pol B, Campan P, Hardwigsen J, Botti G, Pons J, Le TYP. Morbidity of major hepatic resections: a 100-case prospective study. Eur J Surg. 1999;165(5):446–453. doi:10.1080/110241599750006686

23. Mai RY, Ye JZ, Long ZR, et al. Preoperative aspartate aminotransferase-to-platelet-ratio index as a predictor of posthepatectomy liver failure for resectable hepatocellular carcinoma. Cancer Manag Res. 2019;11:1401–1414. doi:10.2147/CMAR.S186114

24. Vickers AJ, Elkin EB. Decision curve analysis: a novel method for evaluating prediction models. Med Decis Making. 2006;26(6):565–574. doi:10.1177/0272989X06295361

25. Gurusamy KS, Li J, Sharma D, Davidson BR, Pharmacological interventions to decrease blood loss and blood transfusion requirements for liver resection. Cochrane Database Syst Rev. 2009;(4):CD008085. doi:10.1002/14651858.CD008085