The value of prealbumin in predicting post hepatectomy liver failure for patients with hepatocellular carcinoma undergoing major hepatectomy

Abstract

Post hepatectomy liver failure (PHLF) poses a significant threat to the prognosis of patients with hepatocellular carcinoma (HCC), particularly those undergoing major hepatectomy. The present research endeavors to clarify the influence of prealbumin on the development of PHLF in HCC patients who have undergone major hepatectomy. Patients with HCC who have undergone major hepatectomy were included. Based on the ROC curve, the optimal cut-off value for prealbumin was determined and patients were divided into two groups. Univariate and multivariate logistic analyses were conducted to identify risk factors for PHLF in HCC patients. Furthermore, the predictive ability of PHLF was also evaluated. 466 patients were included, among whom 98 (21%) developed PHLF. Compared with the high prealbumin group, patients in the low prealbumin group had significantly higher proportions of cirrhosis, portal hypertension, intraoperative blood loss, and transfusion, as well as a higher incidence of PHLF (12.3% vs. 23.5%, P = 0.011). Multivariate analysis revealed that prealbumin is a risk factor for PHLF (HR 1.446, 95%CI 1.091–2.369, P = 0.015), but it is not a risk factor for severe PHLF (HR 1.183, 95%CI 0.584–2.692, P = 0.289). However, the comprehensive indicator, prealbumin-bilirubin (preALBI), is not only a risk factor for PHLF but also severe PHLF. Furthermore, its predictive performance is significantly higher than that of other related indicators (all P < 0.05). Patients with low prealbumin levels require perioperative protocols: precise resection control, liver volume assessment, and PHLF prevention monitoring.

Introduction

Hepatocellular carcinoma (HCC), is a major global health burden with high incidence and mortality rates^1,2. Given the shortage of liver donors, coupled with the advanced stage of the tumor at diagnosis, hepatectomy remains the cornerstone of treatment for HCC, offering the best chance for long-term survival and potential cure^3,4. However, the challenge lies in the fact that the majority of patients with HCC are diagnosed at an advanced or intermediate stage, where tumors are often large or multifocal, necessitating major hepatectomy, a procedure that involves the removal of three or more Couinaud liver segments⁵. Such extensive surgical interventions, while crucial for tumor eradication, inadvertently exacerbate the risk of postoperative complications, particularly post-hepatectomy liver failure (PHLF)⁶, which emerges as a pivotal determinant of perioperative mortality and long-term prognosis⁷. Meanwhile, patients with HCC frequently harbor a background of chronic liver diseases, most notably hepatitis B virus (HBV) or HCV infection, with a significant proportion also exhibiting cirrhosis^{8, 9–10}. This underlying liver pathology further complicates the surgical landscape, as cirrhotic livers are inherently compromised in their regenerative capacity and metabolic functions. Consequently, these patients are particularly susceptible to PHLF, a devastating complication that can rapidly progress to multi-organ failure and death if not promptly recognized and managed¹¹.

Given the gravity of PHLF as a postoperative sequela, early and accurate prediction of its occurrence is paramount. Such predictions can facilitate preoperative risk stratification, optimize patient selection for surgery, and guide personalized perioperative management strategies aimed at mitigating the risk of PHLF. Traditional risk assessment models, while informative, often lack the precision necessary to capture the intricate interplay between patient-specific factors, tumor characteristics, and underlying liver disease severity^{12, 13–14}. Albumin, a well-recognized marker of liver synthetic function, has been extensively studied in the context of postoperative outcomes following hepatectomy^{15, 16–17}. Its depletion has been implicated as a predictor of impaired liver regeneration and increased susceptibility to PHLF. However, the clinical utility of serum albumin as a standalone predictor is limited by its labile nature and susceptibility to external perturbations, particularly in the perioperative setting where blood transfusions and albumin infusions are commonplace. These interventions can artificially elevate albumin levels, thus obscuring the true status of liver synthetic function and undermining the accuracy of predictive models. In the search for a more robust and reliable biomarker, prealbumin, a protein that is more stable and reflective of liver synthetic function than albumin, has garnered widespread attention. Unlike albumin, prealbumin has a shorter half-life and a more rapid turnover rate, reflecting changes in liver function with greater sensitivity and specificity¹⁸. Preliminary studies have hinted at its potential as an early indicator of hepatic dysfunction¹⁹, suggesting that it might offer superior predictive performance for PHLF compared to albumin.

The present study aims to delve deeper into the relationship between prealbumin levels and the risk of PHLF in patients undergoing major hepatectomy for HCC. Furthermore, we explored a comprehensive indicator based on prealbumin, specifically the impact of prealbumin-bilirubin (preALBI) levels on PHLF, and compared its predictive performance of PHLF with existing indicators of liver function, inlcuding albumin, albumin-bilirubin (ALBI)²⁰, and Child-Pugh classification²¹.

Methods

Patients

From January 2017 and January 2024, consecutive patients with HCC after curative major hepatectomy were identified at Zhejiang Provincial People’s Hospital, Hangzhou, Chian. Major hepatectomy was defined as a procedure that involves the removal of three or more Couinaud liver segments⁵. The patient selection process incorporated the following exclusion criteria: (1) patients who were either 18 years old or younger, or 80 years old or older at the time of diagnosis; (2) patients with recurrent HCC, or mixed with intrahepatic cholangiocarcinoma; (3) patients suffering from nutritional-related metabolic diseases, such as renal insufficiency; (4) patients received anti-tumor therapy preoperatively; (5) patients who died from other causes during the perioperative period, such as massive hemorrhage or cardiocerebrovascular events; and (6) patients received emergency surgery without rigorous liver function assessment, such as in cases of ruptured and bleeding HCC, (7) patients with incomplete or missing medical records. The informed consent from all included patients was obtained before surgery. All procedures adhered to the ethical standards set by the institutional and national research committees, in accordance with the 1964 Helsinki Declaration and its subsequent amendments, or equivalent ethical guidelines. The present research had obtained approval from the Institutional Ethics Committees of Zhejiang Provincial People’s Hospital.

Data collection

The study variables were retrospectively collected from the medical record system of Zhejiang Provincial People’s Hospital, including patients’ baseline characteristics, tumor-related variables, and intraoperative variables. Open and laparoscopic hepatectomies adhere to the same surgical principles, including assessments of tumor resectability, adequacy of future liver remnant, liver function, and the patient’s general physical condition²². Laparoscopic surgeries are performed in accordance with the Chinese Expert Consensus on Laparoscopic Hepatectomy and relevant surgical guidelines²³. The definition and severity grading (A, B, and C) of PHLF was established based on the guidelines set forth by the International Study Group of Liver Surgery (ISGLS)⁶. Co-morbidities included diabetes mellitus, hypertension, cardiovascular disease, and chronic obstructive pulmonary disease. Portal hypertension is diagnosed when either splenomegaly accompanied by a platelet count decrease to ≤ 100 × 10⁹/L or gastroesophageal varices is present²⁴. Cirrhosis was confirmed by histopathological examination of the resected specimens. According to previous studies, ALBI score = (log10 bilirubin*0.66) +(albumin *-0.085), where bilirubin is in µmol/L and albumin in g/L²⁵, and preALBI score = (log10 bilirubin*0.042) +(prealbumin *− 0.397), where bilirubin is in µmol/L and prealbumin in g/dL²⁶. ALBI score was stratified into three groups (ALBI grade 1: ≤ − 2.60, ALBI grade 2: more than − 2.60 to ≤ − 1.39, and ALBI grade 3: more than − 1.39), and preALBI score was stratified into two groups (low preALBI: ≤ − 0.19, high preALBI: more than − 0.19).

Statistical analysis

To depict the distribution of each categorical variable, frequencies and percentages were employed, while differences among the variables were assessed using either Pearson’s chi-square test or Fisher’s exact test, depending on the suitability of each method. The optimal cut-off value for prealbumin is determined based on the corresponding value of the Youden index derived from the reveiver operating characteristic (ROC) curve. Variables that demonstrated significance at the level of P < 0.1 in the univariate logistic analysis were incorporated into the forward stepwise multivariate logistic regression model. The predictive capabilities of prealbumin, albumin, the ALBI score, the preALBI score, and the Child-Pugh score were assessed through the utilization of ROC curves. The areas under these curves (AUC) were computed and subsequently compared using the DeLong test. A p-value threshold of < 0.05 was considered statistically significant. The statistical computations for this study were conducted using R software, version 4.4.0 (available at https://cran.r-project.org/src/base/R-4/).

Results

Baseline characteristics

A cohort comprising 466 patients diagnosed with HCC and initially treated with major hepatectomy was enrolled in this study (Table 1). Within this group, a preponderance of 386 patients (82.8%) were male, and a notable majority (90.7%) had a history of HBV infection. Additionally, 171 patients (36.7%) presented with large HCC tumors exceeding 5 cm in diameter, while 94 patients (20.2%) had two or more tumors. Furthermore, 307 patients (65.9%) underwent laparoscopic hepatectomy. Using the Youden index derived from the ROC curve, an optimal cut-off value for prealbumin of 1.32 g/dL was established. Subsequently, patients were dichotomized into two groups: the high prealbumin group (n = 226, 48.5%) and the low prealbumin group (n = 240, 51.5%). Notably, patients in the low prealbumin group exhibited significantly worsened liver function compared to those in the high prealbumin group, evident from a higher prevalence of cirrhosis, portal hypertension, low platelet counts, and elevated transaminase levels (all P < 0.05). Additionally, patients within the low prealbumin group were more susceptible to intraoperative blood loss (> 600 mL) and required blood transfusion treatment.

Table 1. Comparisons of clinical characteristics between the two groups according to the prealbumin value.

Variables (N, %)	High prealbumin (≥ 1.32 g/mL, n = 226)	Low prealbumin (< 1.32 g/mL,n = 240)	P value
Sex, Male	182 (80.5)	204 (82.9)	0.551
Age, > 60 years	114 (50.4)	117 (47.6)	0.580
Co-morbid illness, yes	48 (21.2)	47 (19.1)	0.568
ASA, > 2	46 (20.4)	62 (25.2)	0.229
Performance status, ≥ 1	51 (22.6)	83 (33.7)	0.002
HBV ( +)	197 (87.2)	226 (91.9)	0.099
Cirrhosis	154 (68.1)	190 (77.2)	0.030
Portal hypertension, yes	45 (19.9)	93 (37.8)	< 0.001
PT, > 13 S	39 (17.3)	50 (20.3)	0.412
PLT, < 100*109/L	40 (17.7)	64 (26.0)	0.035
ALT, > 40 IU/L	53 (23.5)	70 (28.5)	0.248
AST, > 40 IU/L	66 (29.2)	94 (38.2)	0.041
TB, ≥ 34 μmol/L	52 (23.0)	59 (24.0)	0.829
ALB, < 35 g/L	57 (25.2)	91 (37.0)	0.007
ALBI*, ≥ 2	133 (58.8)	173 (70.3)	0.009
preALBI#, ≥—0.19	120 (53.1)	238 (96.7)	< 0.001
Child–Pugh, B	14 (6.2)	34 (13.8)	0.006
AFP, ≥ 400 ng/mL	52 (23.0)	68 (27.6)	0.290
Tumor size, > 5 cm	75 (33.2)	96 (39.0)	0.213
Multiple tumors	40 (17.7)	54 (22.0)	0.252
Primary tumor location			0.447
Left hepatic lobe	51 (22.6)	48 (19.5)
Right hepatic lobe	150 (66.4)	168 (68.3)
Median hepatic lobe	19 (8.4)	27 (11.0)
Caudate lobe	6 (2.7)	3 (1.2)
Macrovascular invasion ( +)	19 (8.4)	29 (11.8)	0.286
Type of surgery, laparoscope	136 (60.2)	141 (57.3)	0.575
Anatomical resection	81 (35.8)	74 (30.1)	0.203
Resection margin, < 1 cm	160 (70.8)	172 (69.9)	0.841
Blood loss, > 600 mL	41 (18.1)	79 (32.1)	0.001
Blood transfusion, yes	54 (23.9)	99 (40.2)	< 0.001
Operation time, > 180 min	146 (64.6)	161 (65.4)	0.848
MVI ( +)	97 (42.9)	128 (52.0)	0.053
Differentiation, poor	79 (35.0)	65 (26.4)	0.046

ASA, physical status classification system; HBV, hepatitis B virus; PT, prothrombin time; PLT, platelet count; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ALB, albumin; ALBI, albumin-bilirubin; preALBI, prealbumin-bilirubin; AFP, alpha-fetoprotein; MVI, microvascular invasion.

Post-hepatectomy liver failure

According to the definition of PHLF by the ISGLS, 98 patients (21.0%) developed PHLF following major hepatectomy. Among them, 69 patients exhibited mild liver failure (grade A), while the remaining 29 patients suffered from severe PHLF, specifically 20 patients classified as grade B, and 9 patients as grade C. In the group of patients with low prealbumin levels, 78 patients (31.7%) developed PHLF, categorized as follows: 52 patients in grade A, 18 patients in grade B, and 8 patients in grade C. In contrast, among patients with high prealbumin levels, only 20 patients (8.8%) developed PHLF, with a distribution of 17 patients in grade A, two patients in grade B, and only one patient in grade C (P = 0.001).

Multivariable analysis for risk factors

To address potential multicollinearity, albumin, ALBI, preALBI and the Child-Pugh grade were assessed independently from the prealbumin. In the context of overall PHLF, the multivariable analysis identified prealbumin (OR 1.446, 95%CI 1.091–2.369, P = 0.015), and preALBI (OR 2.190, 95%CI 1.189–4.032, P = 0.012) as significant independent risk factors (Table 2). What’s more, for severe PHLF, only preALBI (OR: 1.285, 95% CI: 1.021–3.446, P = 0.011) emerged as a significant independent risk factor. And, the prealbumin (OR 1.183, 95%CI 0.584–2.692, P = 0.289) was found to be insignificant (Table 3). Furthermore, neither albumin nor Child-Pugh grade is a risk factor for either overall PHLF or severe PHLF, whereas ALBI is a risk factor for overall PHLF but not for severe PHLF.

Table 2. Univariable and multivariable logistic regression analyses of risk factors associated with post-hepatectomy liver failure for patients with hepatocellular carcinoma after major hepatectomy.

Variables	UV OR (95% CI)	P	MV OR (95% CI)#	P
Sex, male vs. female	1.228 (0.685–2.200)	0.490
Age, > 65 vs. ≤ 65 years	1.622 (0.806–2.615)	0.347
Co-morbid illness, yes vs. no	1.469 (0.838–2.575)	0.180
ASA, > 2 vs. ≤ 2	1.264 (0.710–2.251)	0.426
Performance status, ≥ 1 vs. < 1	2.218 (1.370–3.590)	0.001	NS
HBV ( +)	2.012 (0.805–5.023)	0.134
Cirrhosis, yes vs. no	1.089 (0.620–1.910)	0.768
Portal hypertension, yes vs. no	1.625 (1.027–2.618)	0.038	1.383 (1.079–2.612)	0.040
PT, > 13 vs. ≤ 13 S	1.044 (0.594–1.836)	0.880
PLT, ≥ 100*109/L vs. < 100*109/L	1.468 (0.870–2.477)	0.150
ALT, > 40 vs. ≤ 40 IU/L	1.060 (0.561–2.001)	0.858
AST, > 40 vs. ≤ 40 IU/L	1.802 (0.886–3.293)	0.256
TB, ≥ 34 vs. < 34 μmol/L	1.149 (0.661–1.998)	0.623
*ALB, < 35 vs. ≥ 35 g/L	1.217 (0.769–1.926)	0.401	1.132 (0.866–1.815)	0.260
preALB, < 1.70 vs. ≥ 1.70 g/dL	1.425 (1.102–2.854)	0.007	1.446 (1.091–2.369)	0.015
ALBI, < 2 vs. ≥ 2	2.003 (1.267–3.167)	0.003	1.790 (1.070–2.996)	0.027
preALBI, < -0.19 vs. ≥ -0.19	4.782 (2.810–8.139)	0.001	2.190 (1.189–4.032)	0.012
Child–Pugh, B vs. A	1.310 (1.054–2.625)	0.006	1.025 (0.834–2.086)	0.404
AFP, > 400 vs ≤ 400 ng/mL	1.514 (0.874–2.308)	0.214
Tumor size, > 5 vs ≤ 5 cm	1.122 (0.672–1.875)	0.659
Multiple tumors, ≥ 2 vs. 1	1.695 (0.924–3.108)	0.088	NS
Tumor location, Left hepatic lobe	Ref
Right hepatic lobe	1.376 (0.745–2.540)	0.308
Median hepatic lobe	1.036 (0.383–2.804)	0.944
Caudate lobe	1.690 (0.790–6.016)	0.377
Type of surgery, laparoscopic vs. open	1.325 (0.821–2.139)	0.250
Anatomical resection, yes vs. no	1.012 (0.618–1.657)	0.961
Resection margin, < 1 vs ≥ 1 cm	1.168 (0.698–1.657)	0.420
Macrovascular invasion, yes vs. no	2.057 (1.023–4.138)	0.043	NS
Blood loss, > 600 vs ≤ 600 ml	1.879 (1.096–3.222)	0.022	1.452 (1.034–3.086)	0.004
Blood transfusion, yes vs. no	1.613 (0.950–2.738)	0.077	NS
Operation time, > 180 vs. ≤ 180 min	1.712 (0.883–2.984)	0.158
MVI, yes vs. no	1.025 (0.849–1.547)	0.651
Differentiation, poor vs. moderate/well	1.208 (0.487–2.321)	0.385

^# Those variables found significant at P < 0.1 in univariable analyses were entered into multivariable logistic analyses. * To avoid collinearity with ALB, prealbumin, ALBI, preALBI, and Child–Pugh were analyzed separately with other variables. ASA, physical status classification system; HBV, hepatitis B virus; PT, prothrombin time; PLT, platelet count; ALT, alanine aminotransferase; AST, aspartate aminotransferase; TB, total bilirubin; ALB, albumin; ALBI, albumin-bilirubin; preALBI, prealbumin-bilirubin; AFP, alpha-fetoprotein; MVI, microvascular invasion; UV, univariable; MV, multivariable; NS, no significance; OR, odds ratio.

Table 3. Univariable and multivariable logistic regression analyses of risk factors associated with severe post-hepatectomy liver failure for patients with hepatocellular carcinoma after major hepatectomy.

Variables	UV OR (95% CI)	P	MV OR (95% CI)#	P
Sex, male vs. female	1.162 (0.352–3.834)	0.805
Age, > 65 vs. ≤ 65 years	1.215 (0.461–3.202)	0.694
Co-morbid illness, yes vs. no	1.593 (0.659–2.722)	0.300
ASA, > 2 vs. ≤ 2	1.275 (0.441–3.684)	0.653
Performance status, ≥ 1 vs. < 1	1.632 (0.713–2.527)	0.350
HBV ( +)	2.407 (0.354–4.809)	0.249
Cirrhosis, yes vs. no	1.729 (0.281–3.650)	0.555
Portal hypertension, yes vs. no	1.639 (1.013–6.184)	0.001	1.328 (1.041–5.348)	0.031
PT, > 13 vs. ≤ 13 S	1.259 (0.716–1.989)	0.396
PLT, ≥ 100*109/L vs. < 100*109/L	1.519 (0.643–2.612)	0.229
ALT, > 40 vs. ≤ 40 IU/L	1.292 (0.770–1.432)	0.608
AST, > 40 vs. ≤ 40 IU/L	1.172 (0.753–1.764)	0.127
TB, ≥ 34 vs. < 34 μmol/L	1.087 (0.190–1.626)	0.393
*ALB, < 35 vs. ≥ 35 g/L	1.165 (0.550–3.899)	0.445	1.147 (0.610–2.374)	0.183
preALB, < 1.70 vs. ≥ 1.70 g/dL	1.255 (0.351–2.621)	0.471	1.183 (0.584–2.692)	0.289
ALBI, < 2 vs. ≥ 2	2.419 (1.060–6.460)	0.016	1.259 (0.855–1.367)	0.218
preALBI, < -0.19 vs. ≥ -0.19	3.227 (1.407–7.339)	0.006	1.285 (1.021–3.446)	0.011
Child–Pugh, B vs. A	1.762 (0.960–4.601)	0.073	1.028 (0.370–2.860)	0.958
AFP, > 400 vs ≤ 400 ng/mL	1.327 (0.705–2.349)	0.246
Tumor size, > 5 vs ≤ 5 cm	1.068 (0.277–1.667)	0.399
Multiple tumors, ≥ 2 vs. 1	1.402 (0.532–3.695)	0.494
Tumor location, Left hepatic lobe	Ref
Right hepatic lobe	1.229 (0.437–3.455)	0.696
Median hepatic lobe	1.101 (0.190–5.257)	0.999
Caudate lobe	1.057 (0.845–7.968)	0.577
Type of surgery, laparoscopic vs. open	1.445 (0.625–3.342)	0.390
Anatomical resection, yes vs. no	1.014 (0.446–2.303)	0.974
Resection margin, < 1 vs ≥ 1 cm	1.629 (0.282–2.402)	0.257
Macrovascular invasion, yes vs. no	2.362 (0.847–6.588)	0.101
Blood loss, > 600 vs ≤ 600 ml	2.860 (1.146–7.136)	0.024	1.517 (1.100–5.760)	0.029
Blood transfusion, yes vs. no	1.809 (0.764–4.287)	0.178
Operation time, > 180 vs. ≤ 180 min	1.141 (0.421–3.092)	0.795
MVI, yes vs. no	1.517 (0.767–5.938)	0.305
Differentiation, poor vs. moderate/well	1.084 (0.339–1.909)	0.621

^# Those variables found significant at P < 0.1 in univariable analyses were entered into multivariable logistic analyses. * To avoid collinearity with ALB, prealbumin, ALBI, preALBI, and Child–Pugh were analyzed separately with other variables. ASA, physical status classification system; HBV, hepatitis B virus; PT, prothrombin time; PLT, platelet count; ALT, alanine aminotransferase; AST, aspartate aminotransferase; TB, total bilirubin; ALB, albumin; ALBI, albumin-bilirubin; preALBI, prealbumin-bilirubin; AFP, alpha-fetoprotein; MVI, microvascular invasion; UV, univariable; MV, multivariable; NS, no significance; OR, odds ratio.

Predictive performance

Subsequently, ROC curves were generated to discern which variable exhibits superior predictive capacity for postoperative overall and severe PHLF. The respective AUCs for prealbumin, albumin, ALBI score, preALBI score, and Child-Pugh grade in predicting overall PHLF were 0.613 (CI: 0.556–0.681), 0.558 (CI: 0.520–0.599), 0.695 (CI: 0.597–0.713), 0.664 (0.565–0.705), and 0.554 (CI: 0.510–0.593) (Fig. 1A). Similarly, for severe PHLF, the corresponding AUCs were 0.594 (CI: 0.517–0.681), 0.587 (CI: 0.511–0.654), 0.714 (CI: 0.584–0.757), 0.617 (CI: 0.529–0.697), and 0.547 (CI: 0.509–0.612) (Fig. 1B). A comparison of AUCs using the DeLong test revealed that preALBI significantly outperforms prealbumin, albumin, ALBI score, and Child-Pugh grade in prognostic ability (all P-values < 0.05).

Fig. 1 [Images not available. See PDF.]

Comparisons of the area under the receiver operating character curves for post-hepatectomy liver failure (A) and severe post-hepatectomy liver failure (B) stratified by the prealbumin, ALB, ALBI, preALBI and Child-Pugh grade. Abbreviations: ALB, albumin; preALB, prealbumin; ALBI, albumin-bilirubin; preALBI, prealbumin-bilirubin; C-P, Child-Pugh.

Discussion

This study aims to explore the value of prealbumin and its comprehensive indicator, the prealbumin-Bilirubin (preALBI) score, in predicting PHLF. A total of 466 patients were included in the study and divided into two groups based on the cut-off value of prealbumin (1.32 g/dL). The results showed that compared with the high prealbumin group, patients in the low prealbumin group exhibited significantly worsened liver function compared to those in the high prealbumin group, evident from a higher prevalence of cirrhosis, portal hypertension, low platelet counts, and elevated transaminase levels, and were more prone to intraoperative bleeding, and required blood transfusion therapy. However, there was no significant difference in oncologic indicators between the two groups. Additionally, patients in the low prealbumin group had a higher incidence of PHLF (31.7% vs. 8.8%). Multivariate analysis revealed that prealbumin was a risk factor for overall PHLF but not for severe PHLF. Meanwhile, the preALBI score was identified as a risk factor for both overall PHLF and severe PHLF. Furthermore, additional research indicated that compared to prealbumin alone and other commonly used liver function indicators (albumin, ALBI and Child-Pugh grade), the preALBI score demonstrated superior predictive performance for both overall and severe PHLF.

Due to variations in the study populations and definitions of PHLF^{27, 28–29}, the reported incidence of postoperative liver failure varies widely, ranging from 1.2 to 32%^{30, 31, 32–33}. Recently, the ISGLS proposed a unified and widely accepted definition for PHLF, categorizing its severity into three grades: A, B, and C⁶. In this study, the results indicate that the overall incidence of PHLF, based on the definition by the ISGLS, was 21.0%. The possible reasons are that patients have undergone major hepatectomy and that the majority of them suffer from cirrhosis, especially with portal hypertension. Among them, the majority (14.6%) are grade A, with severe PHLF (grades B and C) accounting for only 6.4%.

Given the shortage of liver donors, coupled with the advanced stage of the tumor at diagnosis, hepatectomy remains the main choice of curative treatment for HCC^3,4. Importantly, advancements in preoperative assessment, surgical techniques, and postoperative care have resulted in a significant increase in the scope of hepatectomy aims to achieve radical tumor resection and prolong patient survival. However, extensive liver resection results in a significant reduction in the remaining liver volume. Despite rigorous preoperative assessment of liver function, including the remaining liver volume²³, some patients still developed PHLF, especially those with HBV-related cirrhosis. In this study, 90.7% patients had a history of HBV infection and 73.8% patients were confirmed with cirrhosis. Despite the contradiction between achieving radical tumor resection through extensive liver resection and the need to maximize the remaining liver volume in cases of cirrhosis, it remains a question whether rigorous preoperative assessment can be employed to reduce the incidence of PHLF. Recently, Famularo et al.¹² developed an enhanced CT-radiomics-based model for predicting PHLF. However, the study utilized the Model for End-Stage Liver Disease (MELD) score, which is intended for assessing liver function in patients with advanced cirrhosis, and is not suitable for evaluating liver function in patients undergoing liver surgery. Additionally, Jeong et al.¹³ established a model for predicting PHLF through enhanced MRI-radiomics, but albumin was not even included in their model, resulting in the absence of an indicator for assessing liver synthetic function. Therefore, it is crucial to explore more accurate indicators for assessing liver function and to apply them to improve the prediction of PHLF.

Prior to surgery, liver function stands as a pivotal determinant influencing PHLF, as evidenced in studies. Currently, the clinical landscape primarily relies on several indicators to assess preoperative liver function, namely albumin levels, the ALBI grade^20,34, Child-Pugh grade²¹. The Child-Pugh system, which assigns scores based on conventional liver function tests, ascites severity, and hepatic encephalopathy grade, was initially developed to anticipate the outcomes of cirrhotic patients with portal hypertension undergoing surgery for variceal bleeding. Due to the excessive subjectivity of some indicators, the scores do not accurately reflect differences in patients’ liver function. Furthermore, since most surgical patients require relatively good liver function, this results in the majority being Child-Pugh A patients who are eligible for surgery. For instance, in this study, only 48 Child-Pugh B patients (10.3%) underwent surgery. Despite its prevalent use, this grading system is not devoid of shortcomings, including its complexity, subjectivity, and a notable lack of precision and accuracy²¹. Certainly, as indicated by the multivariate analysis results, Child-Pugh grade does not categorize the patient as being at risk for either overall PHLF or severe PHLF. In contrast, the ALBI grade, derived from albumin and bilirubin laboratory results, has demonstrated superiority over the Child-Pugh classification in assessing hepatic functional reserve^25,35. However, it is noteworthy that serum albumin levels can fluctuate and are susceptible to factors such as peripheral albumin infusion or blood transfusion, potentially compromising the accuracy and precision of the ALBI grade.

However, it is crucial to acknowledge that serum albumin levels can be influenced by various factors, including peripheral albumin infusions or blood transfusions, which may undermine the reliability of the ALBI grade. Therefore, even though ALBI is a risk factor for overall and severe PHLF in this study, this does not negate the potential inaccuracy of its predictions due to the influence of albumin. Therefore, this study adopted prealbumin, which is a more robust and reliable biomarker and less susceptible to external supplementation, as an indicator for assessing liver function. The result showed that prealbumin is a risk factor influencing PHLF, with an AUC value of 0.613. However, possibly due to the low incidence of complications from severe PHLF and the relatively small sample size, statistical results indicate that prealbumin is not a significant risk factor for severe PHLF, with an AUC value of 0.594. Subsequently, we analyzed the composite indicator based on prealbumin, known as preALBI, to more comprehensively reflect liver synthetic and metabolic functions. The results show that preALBI is a risk factor influencing both overall and severe PHLF.

Several considerations regarding study limitations are pertinent. Being a retrospective analysis, standardization and identification of certain factors were inevitably constrained. Furthermore, the study population was predominantly Chinese with a high prevalence of HBV infection (exceeding 90%), necessitating further investigation to ascertain the applicability of the findings to HCC primarily associated with HCV. Moreover, our study was confined to the assessment of prealbumin, albumin, and their combined indicators, overlooking other liver function metrics like ICG-R15, as the focus was specifically on blood-based markers. Additionally, we restricted enrollment to patients who underwent major hepatectomy, excluding other potential postoperative contributors to PHLF, given the heightened susceptibility of this patient cohort to PHLF due to liver function compromises. Conversely, the factors influencing PHLF following minor hepatectomy are more diverse, encompassing bleeding and infection among others. Hence, further validation is crucial to ascertain the relevance of our study results to patients undergoing minor hepatectomy. Ultimately, a multicenter randomized controlled trial (RCT) remains imperative.

Conclusion

Prealbumin is a risk factor for PHLF, and the comprehensive indicator based on prealbumin, preALBI, demonstrates better predictive performance. For patients with low prealbumin levels, strict control of surgical indications is required, involving accurate preoperative assessment of residual liver volume ratio and hepatic functional reserve; intraoperative precision in controlling resection extent, and postoperative close monitoring of liver function and systemic organ changes with timely targeted management to prevent PHLF.

Author contributions

Authors’ Contributions: Hang-Dong Jia, Zhe-Jin Shi, Jian-Yong Yuan, and Kai Wang contributed equally to this work. Dr. Lei Liang had full access to all the data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Hang-Dong Jia, Zhe-Jin Shi and Lei Liang. Acquisition, analysis, or interpretation of data: Hang-Dong Jia, Zhe-Jin Shi, Jian-Yong Yuan, Kai Wang, Yang Yu, Zheng-Kang Fang, Kai-Di Wang and Yi Lu. Drafting of the manuscript: Hang-Dong Jia, Zhe-Jin Shi, Jian-Yong Yuan, and Kai Wang. Critical revision of the manuscript for important intellectual content: Guo-Liang Shen, and Cheng-Wu Zhang. Statistical analysis: Yang Yu. Obtained funding: Lei Liang. Administrative, technical, or material support: Cheng-Wu Zhang, and Lei Liang. Study supervision: Lei Liang.

Funding

This work was supported by the National Natural Science Foundation of China (No. 82302915 and 82203403), the Zhejiang Provincial Natural Science Foundation of China (No. LQ23H160049), the fund of Public Welfare Technology Research Program of Zhejiang Provincial Natural Science Foundation (No. LGF22H030012), the General Research Program of Zhejiang Provincial Department of Education (No. Y202146104), the fund of Medical and Health Research Projects in Zhejiang Province (No. 2024KY764), and the Basic Research Funds for Hangzhou Medical College Basic Research Program (No. KYQN202113).

Data availability

The datasets used and analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Abbreviations

HCC

Hepatocellular carcinoma

PHLF

Post-hepatectomy liver failure

ASA

American Standards Association of physical status classification system

HBV

Hepatitis B virus

PT

Prothrombin time

PLT

Platelet count

ALT

Alanine aminotransferase

AST

Aspartate aminotransferase

TB

Total bilirubin

ALB

Albumin

ALBI

Albumin-bilirubin

preALBI

Prealbumin-bilirubin

AFP

Alpha-fetoprotein

MVI

Microvascular invasion

UV

Univariable

MV

Multivariable

NS

No significance

OR

Odds ratio

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Vogel, A; Meyer, T; Sapisochin, G; Salem, R; Saborowski, A. Hepatocellular carcinoma. Lancet; 2022; 400, 10360 pp. 1345-1362.1:CAS:528:DC%2BB38XisFWku73I [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36084663]

2. Yang, JD et al. A global view of hepatocellular carcinoma: trends, risk, prevention and management. Nat. Rev. Gastroenterol. Hepatol.; 2019; 16, 10 pp. 589-604. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31439937][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6813818]

3. Bruix, J; Gores, GJ; Mazzaferro, V. Hepatocellular carcinoma: clinical frontiers and perspectives. Gut; 2014; 63, 5 pp. 844-855.1:CAS:528:DC%2BC2cXpsVSlt70%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24531850]

4. Liang, L et al. Surgical resection versus transarterial chemoembolization for BCLC intermediate stage hepatocellular carcinoma: a systematic review and meta-analysis. HPB: Official J. Int. Hepato Pancreato Biliary Association; 2018; 20, 2 pp. 110-119.

5. Couinaud, C. Liver anatomy: portal (and suprahepatic) or biliary segmentation. Dig. Surg.; 1999; 16, 6 pp. 459-467.1:STN:280:DC%2BD3c3mtlaqtg%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/10805544]

6. Rahbari, NN et al. Posthepatectomy liver failure: a definition and grading by the international study group of liver surgery (ISGLS). Surgery; 2011; 149, 5 pp. 713-724. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21236455]

7. Sarkhampee, P; Ouransatien, W; Chansitthichok, S; Lertsawatvicha, N; Wattanarath, P. The impact of post-hepatectomy liver failure on long-term survival after liver resection for Perihilar cholangiocarcinoma. HPB: Official J. Int. Hepato Pancreato Biliary Association; 2024; 26, 6 pp. 808-817.

8. Gao, ZY et al. Survival benefit of adjuvant TACE for patients with hepatocellular carcinoma and child-pugh B7 or B8 after hepatectomy. BMC Cancer; 2024; 24, 1 1241. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39379833][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11462920]

9. Singh, SP; Madke, T; Chand, P. Global epidemiology of hepatocellular carcinoma. J. Clin. Exp. Hepatol.; 2025; 15, 2 102446.1:CAS:528:DC%2BB2cXisFyrurjM [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39659901]

10. Sagnelli, E; Macera, M; Russo, A; Coppola, N; Sagnelli, C. Epidemiological and etiological variations in hepatocellular carcinoma. Infection; 2020; 48, 1 pp. 7-17.1:CAS:528:DC%2BC1MXhsVChtbvE [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31347138]

11. Merath, K et al. Postoperative liver failure: definitions, risk factors, prediction models and prevention strategies. J. Gastrointest. Surg.; 2023; 27, 11 pp. 2640-2649. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37783906]

12. Famularo, S. et al. Preoperative prediction of post hepatectomy liver failure after surgery for hepatocellular carcinoma on CT-scan by machine learning and radiomics analyses. Eur. J. Surg. Oncol. J. Eur. Soc. Surg. Oncol. Br. Assoc. Surg. Oncology2024, 109462 .

13. Jeong, B. et al. Predicting post-hepatectomy liver failure in patients with hepatocellular carcinoma: Nomograms based on deep learning analysis of Gadoxetic acid-enhanced MRI. Eur Radiol (2024).

14. Alaimo, L et al. A comprehensive preoperative predictive score for post-hepatectomy liver failure after hepatocellular carcinoma resection based on patient comorbidities, tumor burden, and liver function: the CTF score. J. Gastrointest. Surg.; 2022; 26, 12 pp. 2486-2495. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36100827]

15. Zhang, J et al. The APP score: A simple serum biomarker model to enhance prognostic prediction in hepatocellular carcinoma. Biosci. Trends; 2025; 18, 6 pp. 567-583. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39631885]

16. Huang, X; Peng, G; Kong, Y; Cao, X; Zhou, X. The prognostic value of crp/alb ratio in predicting overall survival for hepatocellular carcinoma treated with transcatheter Intra-Arterial therapy combined with Molecular-Targeted agents and PD-1/PD-L1 inhibitors. J. Inflamm. Res.; 2025; 18, pp. 203-217. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39802506][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11725233]

17. Xie, YM et al. Naples prognostic score is an independent prognostic factor in patients undergoing hepatectomy for hepatocellular carcinoma. J. Hepatocellular Carcinoma; 2023; 10, pp. 1423-1433.1:CAS:528:DC%2BB3sXhvFWktrrP

18. Yang, F et al. Effects of early postoperative enteral nutrition versus usual care on serum albumin, prealbumin, transferrin, time to first flatus and postoperative hospital stay for patients with colorectal cancer: A systematic review and meta-analysis. Contemp. Nurse; 2018; 54, 6 pp. 561-577. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30176764]

19. Li, L et al. The association of liver function and quality of life of patients with liver cancer. BMC Gastroenterol.; 2019; 19, 1 66. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31046687][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6498612]

20. Endo, Y et al. Machine learning models including preoperative and postoperative albumin-bilirubin score: short-term outcomes among patients with hepatocellular carcinoma. HPB: Official J. Int. Hepato Pancreato Biliary Association; 2024; 26, 11 pp. 1369-1378.

21. Song, S et al. Perioperative impact of liver cirrhosis on robotic liver resection for hepatocellular carcinoma: a retrospective cohort study. Surg. Endosc; 2024; 38, 9 pp. 4926-4938. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38977502]

22. Zhang, KJ et al. Short- and long-term outcomes of laparoscopic versus open liver resection for large hepatocellular carcinoma: a propensity score study. Surg. Today; 2023; 53, 3 pp. 322-331. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35986784]

23. Sun, HC et al. Chinese expert consensus on conversion therapy for hepatocellular carcinoma (2021 edition). Hepatobiliary Surg. Nutr.; 2022; 11, 2 pp. 227-252. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35464283][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9023831]

24. Turco, L; Garcia-Tsao, G. Portal hypertension: pathogenesis and diagnosis. Clin. Liver Dis.; 2019; 23, 4 pp. 573-587. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31563212]

25. Johnson, PJ 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 pp. 550-558. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25512453]

26. Li, C et al. Development and validation of prealbumin-bilirubin score (preALBI score) for predicting long-term survival after hepatectomy for hepatocellular carcinoma: A multicenter analysis versus ALBI score. Am. J. Surg.; 2024; 232, pp. 87-94. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38238192]

27. Mullen, JT et al. Hepatic insufficiency and mortality in 1,059 noncirrhotic patients undergoing major hepatectomy. J. Am. Coll. Surg.; 2007; 204, 5 pp. 854-862. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17481498]discussion 862 – 854

28. Balzan, S 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 pp. 824-828. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16327492][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1409891]discussion 828–829

29. Schindl, MJ et al. The value of residual liver volume as a predictor of hepatic dysfunction and infection after major liver resection. Gut; 2005; 54, 2 pp. 289-296.1:STN:280:DC%2BD2M%2FitFSrug%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15647196][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1774834]

30. Dinant, S et al. Risk assessment of posthepatectomy liver failure using hepatobiliary scintigraphy and CT volumetry. J. Nucl. Med.; 2007; 48, 5 pp. 685-692. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17475954]

31. Karoui, M et al. Influence of preoperative chemotherapy on the risk of major hepatectomy for colorectal liver metastases. Ann. Surg.; 2006; 243, 1 pp. 1-7.2769307 [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16371728][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1449955]

32. McCormack, L; Petrowsky, H; Jochum, W; Furrer, K; Clavien, PA. Hepatic steatosis is a risk factor for postoperative complications after major hepatectomy: a matched case-control study. Ann. Surg.; 2007; 245, 6 pp. 923-930. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17522518][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1876953]

33. Kawano, Y et al. Short- and long-term outcomes after hepatic resection for hepatocellular carcinoma with concomitant esophageal varices in patients with cirrhosis. Ann. Surg. Oncol.; 2008; 15, 6 pp. 1670-1676. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18368453]

34. Boubaddi, M et al. Comprehensive review of future liver remnant (FLR) assessment and hypertrophy techniques before major hepatectomy: how to assess and manage the FLR. Ann. Surg. Oncol.; 2024; 31, 13 pp. 9205-9220. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39230854]

35. Pan, Z; Ye, YS; Wang, ZP; Li, W. Predictive value of early-stage postoperative albumin-bilirubin grade on the overall survival of hepatocellular carcinoma patients undergoing resection. Eur. J. Gastroenterol. Hepatol.; 2024; 36, 12 pp. 1464-1469.1:CAS:528:DC%2BB2cXitlKiu7bN [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39475785]