Abstract
Objective
This study aimed to identify risk factors associated with early mortality (within 2 years) following hepatectomy in patients with hepatocellular carcinoma (HCC), with the goal of informing more rational treatment strategies.
Methods
A retrospective analysis was conducted on clinical data from 236 HCC patients who underwent hepatectomy at the University of Hong Kong-Shenzhen Hospital between 2013 and 2021. Patients were categorized into early mortality (EM) and non-early mortality (NEM) groups based on survival status at 24 months post-surgery. Univariate and multivariable logistic regression analyses were employed to identify independent risk factors for early mortality. Based on the availability of clinical information, preoperative and early postoperative prediction models were developed, and corresponding nomograms were constructed. The predictive performance of the model was assessed using receiver operating characteristic (ROC) curves, and patients were stratified into different risk groups based on the predicted probabilities for further evaluation.
Results
Independent risk factors for EM in the preoperative assessment included Child–Pugh grade B (P < .001), advanced BCLC staging (P = .010), and high Tumor burden score (TBS) (P < .001). For the early postoperative assessment, independent risk factors were Child–Pugh grade B (P < .001), high TBS (P < .001), non-Textbook Outcome (P < .001), and presence of Microvascular invasion (MVI) (P = .014). The preoperative model achieved an area under the ROC curve (AUC) of 0.832, while the postoperative model yielded an AUC of 0.865, indicating strong predictive performance. Stratification of patients based on predicted probabilities into distinct risk groups revealed statistically significant differences in both 2-year and 5-year survival rates, demonstrating robust risk-discriminatory ability.
Conclusion
Based on the availability of clinical information, this study identified independent risk factors associated with EM in both the preoperative and early postoperative phases, and constructed two nomograms to visualize the corresponding risks. Patients with multiple risk factors—particularly those deemed high-risk for EM (predicted probability > 66%) according to the nomogram—faced a significantly higher risk of early death.
Keywords: Hepatocellular carcinoma, Hepatectomy, Early mortality, Textbook Outcome, Tumor Burden Score
Introduction
According to the Global Cancer Statistics 2022, liver cancer ranks sixth in global cancer incidence and third in cancer-related mortality. In 2022, approximately 866,000 new cases of liver cancer were diagnosed, resulting in 759,000 deaths. Hepatocellular carcinoma (HCC) accounts for 75% to 85% of all liver cancer cases [1]. Due to the scarcity of liver transplant donors, radical surgical resection remains the most critical intervention for achieving long-term survival in HCC patients.
In the past, stringent quality control in liver surgery prioritized the reduction of perioperative mortality within 30 or 90 days after hepatectomy as a key surgical outcome. However, early mortality occurring within one or two years postoperatively, although not directly attributable to the surgery itself, remains a non-negligible aspect affecting the overall treatment outcomes for HCC. Several previous studies have attempted to identify risk factors associated with one-year mortality after hepatectomy in HCC patients [2–4]. Nevertheless, due to considerable variations in the variables included across these studies, consistent conclusions have yet to be established.
Advances in systemic therapies, particularly targeted therapies combined with immunotherapy and dual immunotherapy regimens, have demonstrated median survival times of 20–24 months in patients with intermediate to advanced HCC undergoing non-surgical treatments [5, 6]. Notably, a subset of HCC patients who underwent hepatectomy experienced mortality within 24 months post-surgery, indicating suboptimal surgical outcomes. Identifying, even before surgery or in the early postoperative period, those patients whose survival time may fall short of the median survival expected with palliative care is of critical importance for clinicians to develop individualized treatment plans and for patients to make informed medical decisions. In this retrospective study, we aimed to identify risk factors associated with early postoperative mortality (within 2 years) in patients with HCC undergoing hepatectomy, and to develop early risk assessments—both preoperatively and shortly after surgery—to support clinical decision-making regarding treatment strategies.
Materials and methods
Study population
Clinical data from HCC patients who underwent hepatectomy at the University of Hong Kong-Shenzhen Hospital between 2013 and 2021 were collected. Inclusion criteria were: (1) cardiopulmonary function sufficient to tolerate general anesthesia; (2) preoperative imaging showing no portal vein tumor thrombus and absence of extrahepatic metastases (e.g., lung, bone); (3) age between 18 and 85 years; and (4) pathological confirmation of HCC. Exclusion criteria included: (1) concurrent malignancies at other sites; (2) incomplete clinical-pathological data or prognostic information; and (3) tumor rupture.
Follow-up
Postoperative follow-up included alpha-fetoprotein (AFP) testing and contrast-enhanced CT of the chest and abdomen at 1 month, followed by AFP testing and contrast-enhanced CT or MRI every 3 months. Patient survival status was tracked through outpatient visits and telephone follow-ups. Patients who died within 24 months post-surgery were classified into the early mortality (EM) group, while those who survived beyond 24 months were classified into the non-early mortality (NEM) group.
Definitions
Using the Textbook Outcome in Liver Surgery(TOLS) definitions established by the IHPBA [7], non-TO was defined by the occurrence of any of the following: intraoperative events classified as grade ≥ 2 (according to the Oslo classification [8]), postoperative bile leak classified as grade B or C (based on the severity grading by the International Liver Surgery Study Group [9]), postoperative liver failure classified as grade B or C (according to the same severity grading [10]), occurrence of major complications within 90 days post-surgery (Clavien-Dindo classification grade III or higher [11]), readmission due to major surgical complications within 90 days (Clavien-Dindo classification grade III or higher), in-hospital mortality or death within 90 days post-surgery, and failure to achieve R0 resection margins (i.e., tumor-free margins of ≤ 1 mm). The Barcelona Clinic Liver Cancer (BCLC) staging was conducted according to the 2022 BCLC prognostic prediction and treatment recommendations [12]. The Tumor Burden Score (TBS) was calculated based on the Pythagorean theorem, considering tumor size and number (α2 + β2 = γ2, where α = maximum tumor diameter, β = number of tumors, and γ = TBS) [13].
Statistical analysis
Statistical analyses were conducted using R Studio version 2023.12.0 + 369 (R 4.3.2). Categorical variables are expressed as frequencies (%), and differences between groups were assessed using either the Pearson chi-squared test or Fisher’s exact test. Continuous variables are presented as means ± standard deviations or medians (interquartile ranges), with differences between groups analyzed using the Mann–Whitney U test. Univariate logistic regression was performed to identify potential risk factors for EM among patients. Variables with a p-value less than 0.1 were included in both preoperative and postoperative multivariable logistic regression models based on their preoperative availability, to determine independent risk factors for EM. Based on these independent risk factors, two separate nomograms were constructed. Receiver operating characteristic (ROC) curves were plotted, and the area under the curve (AUC) was calculated to evaluate the predictive ability of the model. Patients were stratified into high-, intermediate-, and low-risk groups based on equally spaced intervals of predicted probabilities, and overall survival (OS) was assessed using Kaplan–Meier survival curves, with the log-rank test employed to compare the significance of differences between survival curves. A p-value of less than 0.05 was considered statistically significant.
Results
Baseline characteristics
The study included a total of 236 patients. Among them, 41 patients experienced EM, including two cases of perioperative mortality and 39 who developed tumor recurrence postoperatively and ultimately succumbed to tumor-related causes. These 41 patients were categorized into the EM group. The remaining 195 patients were assigned to the non-EM group. Baseline characteristics are summarized in Table 1. Both groups were predominantly male (82.9% vs. 82.1%, P = 1.000). Patients in the EM group were younger (49.8 ± 13.5 years vs. 56.5 ± 12.3 years, P = 0.002), but no significant differences were observed in preoperative comorbidities, including the Charlson Comorbidity Index (P = 0.590) and ASA classification (P = 0.835). The majority of patients had concurrent HBV infection (90.2% vs. 85.1%, P = 0.541), with a higher proportion of Child–Pugh grade B patients in the EM group (17.1% vs. 5.6%). The EM group exhibited more advanced BCLC staging (P < 0.001) and higher Tumor Burden Scores (8.2 vs. 4.3, P < 0.001), with similar trends observed for tumor number and maximum tumor diameter. A greater proportion of patients in the EM group had AFP levels > 400 μg/L (51.2% vs. 25.1%, P = 0.002). Additionally, the EM group had a higher rate of anatomical hepatectomy (85.4% vs. 67.7%, P = 0.038) and major hepatectomy (80.5% vs. 38.5%, P < 0.001). Operative duration was longer in the EM group (372.0 vs. 306.0 min, P < 0.001), with greater intraoperative blood loss (700.0 vs. 500.0 mL, P = 0.037) and a higher proportion of patients failing to achieve Textbook Outcome (75.6% vs. 31.3%, P < 0.001). Microvascular invasion (MVI) was more prevalent in the EM group (75.6% vs. 34.4%, P < 0.001), and postoperative hospital stays were longer (15.0 vs. 10.0 days, P < 0.001).
Table 1.
Comparative analysis of clinical data between the EM group and the NEM group
| Variable | Description | Total (N = 236) | EM group (N = 41) | NEM group (N = 195) | P Value |
|---|---|---|---|---|---|
| Sex | Male | 194 (82.2%) | 34 (82.9%) | 160 (82.1%) | 1.000 |
| Female | 42 (17.8%) | 7 (17.1%) | 35 (17.9%) | ||
| Age(years) | 55.4 ± 12.8 | 49.8 ± 13.5 | 56.5 ± 12.3 | .002 | |
| Charlson comorbidity index | 2.4 ± 1.5 | 3.0 (1.0 to 3.0) | 2.0 (1.0 to 3.0) | .590 | |
| ASA classification | 1 | 56 (23.7%) | 10 (24.4%) | 46 (23.6%) | .835 |
| 2 | 162 (68.6%) | 27 (65.9%) | 135 (69.2%) | ||
| 3 | 18 (7.6%) | 4 (9.8%) | 14 (7.2%) | ||
| Abdominal surgery history | Yes | 34 (14.4%) | 4 (9.8%) | 30 (15.4%) | .491 |
| No | 202 (85.6%) | 37 (90.2%) | 165 (84.6%) | ||
| HBV infection | Yes | 203 (86.0%) | 37 (90.2%) | 166 (85.1%) | .541 |
| No | 33 (14.0%) | 4 (9.8%) | 29 (14.9%) | ||
| Child–Pugh grade | A | 218 (92.4%) | 34 (82.9%) | 184 (94.4%) | .029 |
| B | 18 (7.6%) | 7 (17.1%) | 11 (5.6%) | ||
| BCLC staging | 0 | 37 (15.7%) | 2 (4.9%) | 35 (17.9%) | <.001 |
| A | 136 (57.6%) | 19 (46.3%) | 117 (60%) | ||
| B | 41 (17.4%) | 6 (14.6%) | 35 (17.9%) | ||
| C | 22 (9.3%) | 14 (34.1%) | 8 (4.1%) | ||
| Tumor burden score | 5.8 ± 4.0 | 8.2 (5.1 to 13.0) | 4.3 (2.7 to 6.3) | <.001 | |
| Tumor number | Single | 177 (75.0%) | 25 (61%) | 152 (77.9%) | .001 |
| Double | 30 (12.7%) | 4 (9.8%) | 26 (13.3%) | ||
| Multiple | 29 (12.3%) | 12 (29.3%) | 17 (8.7%) | ||
| Maximum tumor diameter | 5.5 ± 4.1 | 8.0 (5.0 to 13.0) | 4.0 (2.3 to 6.0) | <.001 | |
| AFP | ≤ 400 μg/L | 166 (70.3%) | 20 (48.8%) | 146 (74.9%) | .002 |
| > 400 μg/L | 70 (29.7%) | 21 (51.2%) | 49 (25.1%) | ||
| NLR | 3.0 ± 1.5 | 3.0 (2.2 to 4.1) | 2.6 (1.9 to 3.5) | .117 | |
| Anatomical hepatectomy | Yes | 167 (70.8%) | 35 (85.4%) | 132 (67.7%) | .038 |
| No | 69 (29.2%) | 6 (14.6%) | 63 (32.3%) | ||
| Resection extent | Major | 108 (45.8%) | 33 (80.5%) | 75 (38.5%) | <.001 |
| Minor | 128 (54.2%) | 8 (19.5%) | 120 (61.5%) | ||
| Operation duration (min) | 342.6 ± 159.7 | 372.0 (318.0 to 498.0) | 306.0 (220.5 to 387.0) | <.001 | |
| Operation blood loss (ml) | 861.4 ± 1065.8 | 700.0 (500.0 to 1000.0) | 500.0 (300.0 to 1000.0) | .037 | |
| Textbook Outcome | Yes | 144 (61.0%) | 10 (24.4%) | 134 (68.7%) | <.001 |
| No | 92 (39.0%) | 31 (75.6%) | 61 (31.3%) | ||
| Microvascular Invasion | Yes | 98 (41.5%) | 31 (75.6%) | 67 (34.4%) | <.001 |
| No | 138 (58.5%) | 10 (24.4%) | 128 (65.6%) | ||
| Degree of inflammation and necrosis(G) | 0/1 | 58 (24.6%) | 8 (19.5%) | 50 (25.6%) | .529 |
| 2/3/4 | 178 (75.4%) | 33 (80.5%) | 145 (74.4%) | ||
| Degree of fibrosis(S) | 0/1 | 52 (22.0%) | 7 (17.1%) | 45 (23.1%) | .525 |
| 2/3/4 | 184 (78.0%) | 34 (82.9%) | 150 (76.9%) | ||
| Postoperative hospital stay | 13.2 ± 8.9 | 15.0 (11.0 to 20.0) | 10.0 (7.0 to 15.0) | <.001 |
ASA classification, American Society of Anesthesiologists Physical Status Classification; HBV infection, Hepatitis B Virus infection; BCLC staging, Barcelona Clinic Liver Cancer staging system; NLR, Neutrophil-to-Lymphocyte Ratio
Risk factor analysis
As shown in Table 2, univariate analysis identified several potential risk factors for EM following hepatectomy in HCC patients, including age < 60 years, Child–Pugh grade B, BCLC staging, Tumor Burden Score, AFP > 400 μg/L, anatomical hepatectomy, major hepatectomy, operative duration, intraoperative blood loss, non-TO, MVI, and postoperative hospital stay. Incorporating the preoperative available potential risk factors (Age < 60 years, Child–Pugh grade B, BCLC staging, Tumor burden score, AFP > 400 μg/L) into the multivariable logistic regression model revealed the following independent risk factors affecting EM: Child–Pugh grade B (P < 0.001), BCLC staging (P = 0.010), and Tumor burden score (P < 0.001). Similarly, including the postoperative early-stage available potential risk factors in the multivariable logistic regression model identified the following independent risk factors for EM: Child–Pugh grade B (P < 0.001), Tumor burden score (P < 0.001), non-TO (P < 0.001), and MVI (P = 0.040).
Table 2.
Univariate and multivariate risk factor analysis of EM after hepatectomy in patients with HCC
| Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| All | Preoperative | Postoperative | ||||
| Variable | HR (95% CI) | P value | HR (95% CI) | P value | HR (95% CI) | P value |
| Sex, Female vs Male | 0.941(0.386,2.296) | .894 | ||||
| Age, ≥ 60 vs < 60 | 0.426(0.198,0.918) | .029 | 0.639(0.265,1.540) | .319 | 0.637(0.243,1.670) | .359 |
| Charlson Comorbidity Index | 1.034(0.833,1.284) | .762 | ||||
| ASA classification | 1.064(0.567,1.995) | .847 | ||||
| Abdominal Surgery History, Yes vs No | 1.682(0.558,5.065) | .355 | ||||
| HBV Infection, Yes vs No | 0.619(0.205,1.867) | .394 | ||||
|
Child–Pugh g rade, B vs A |
3.444(1.247,9.511) | .017 | 7.892(2.318,26.865) | <.001 | 5.419(1.422,20.651) | .013 |
| BCLC staging | 2.736(1.796,4.166) | <.001 | 1.955(1.172,3.261) | .010 | 1.660(0.948,2.906) | .076 |
| Tumor burden score | 1.294(1.179,1.419) | <.001 | 1.256(1.127,1.400) | <.001 | 1.174(1.034,1.333) | .014 |
| AFP, > 400 μg/L vs ≤ 400 μg/L | 3.129(1.565,6.254) | .001 | 1.233(0.525,2.897) | .631 | 0.884(0.353,2.213) | .792 |
| NLR | 1.163(0.948,1.426) | .148 | ||||
| Anatomical hepatectomy, Yes vs No | 2.784(1.113,6.962) | .029 | 0.943(0.266,3.348) | .928 | ||
| Resection extent, Major vs minor | 6.600(2.894,15.053) | <.001 | 1.894(0.621,5.773) | .262 | ||
| Operation duration (min) | 1.004(1.002,1.005) | <.001 | 1.002(0.999,1.005) | .169 | ||
| Operation blood loss (ml) | 1.000(1.000,1.001) | .032 | 1.000(0.999,1.000) | .301 | ||
| Textbook Outcome, No vs Yes | 6.810(3.139,14.773) | <.001 | 5.124(1.924,13.643) | .001 | ||
| Microvascular Invasion, Yes vs No | 5.922(2.738,12.812) | <.001 | 2.858(1.050,7.785) | .040 | ||
| Degree of Inflammation and Necrosis(G), 2/3/4 vs 0/1 | 1.422(0.616,3.284) | .409 | ||||
| Degree of Fibrosis(S), 2/3/4 vs 0/1 | 1.457(0.605,3.510) | .401 | ||||
| Postoperative hospital stays | 1.050(1.015,1.086) | .005 | 0.981(0.937,1.028) | .423 | ||
ASA classification, American Society of Anesthesiologists Physical Status Classification; HBV infection, Hepatitis B Virus infection; BCLC staging, Barcelona Clinic Liver Cancer staging system; NLR, Neutrophil-to-Lymphocyte Ratio
Development of the nomogram
Based on the multivariable logistic regression model, nomograms were constructed for both preoperative (Fig. 1A) and postoperative (Fig. 1B) assessments by assigning weighted points to each risk factor. The higher the total score from all risk factors, the greater the risk of EM. For instance, consider a patient assessed for EM preoperatively (Fig. 1A) with a Child–Pugh grade B (32 points), BCLC staging B (22 points), and a Tumor burden score of 14 (50 points), resulting in a total score of 104 points, which corresponds to an approximately 90% probability of EM. In contrast, for a patient evaluated for EM in the early postoperative phase (Fig. 1B), the scores were as follows: Child–Pugh grade A (0 points), Tumor burden score of 6 (22 points), failure to achieve Textbook Outcome (26 points), and the presence of MVI (19 points), leading to a total score of 67 points, which indicates an approximately 25% probability of EM.
Fig. 1.
Development of Nomograms Based on Preoperative and Early Postoperative Models. A Nomogram Developed from the Preoperative Model; B Nomogram Developed from the Early Postoperative Model
Predictive performance
To evaluate the discriminative ability of the models, internal validation was performed on the dataset, and receiver operating characteristic (ROC) curves for both the preoperative and postoperative models were plotted (Fig. 2). The area under the ROC curve (AUC) for the preoperative model was 0.832, while for the postoperative model it was 0.865, indicating good predictive capability.
Fig. 2.
ROC Curves of Preoperative and Early Postoperative Models for Predicting EM. The red curve represents the ROC curve for predicting EM using the preoperative model, while the blue curve corresponds to the ROC curve of the early postoperative model for EM prediction
Based on the preoperative model, the predicted probability of EM was calculated for each patient. Patients were then divided into three equally spaced risk groups: low (< 33%), intermediate (≥ 33% and < 66%), and high (≥ 66%). The 2- and 5-year postoperative survival curves for these groups are shown in Fig. 3A. The 2-year survival rates were 90%, 50%, and 27% for the low-, intermediate-, and high-risk groups, respectively, while the 5-year survival rates were 73%, 33%, and not calculable. Similarly, patients were stratified into corresponding risk groups based on the predicted probabilities from the postoperative model, and their 2- and 5-year survival curves are presented in Fig. 3B. The 2-year survival rates were 90%, 59%, and 23%, and the 5-year survival rates were 75%, 24%, and 7.7% for the low-, intermediate-, and high-risk groups, respectively.
Fig. 3.
Postoperative Survival Outcomes in Patients with Different Risk Stratifications. A Postoperative Survival Outcomes in Patients Risk-Stratified Using the Preoperative Model; B Postoperative Survival Outcomes in Patients Risk-Stratified Using the Early Postoperative Model
Discussion
In this study, preoperative and early postoperative models were developed based on the availability of clinical information. Additionally, two nomograms were constructed to visually represent the predictive power of risk factors and facilitate clinical evaluation. Internal validation demonstrated that both models exhibited good predictive performance, with the early postoperative model showing slightly better results. This improvement may be attributed to the inclusion of two strong prognostic indicators in the postoperative model: non-TO and MVI. In summary, the preoperative model can help identify patients whose expected survival time may fall below the median survival (compared with palliative care) at an early stage, providing crucial reference information for clinicians to develop individualized treatment plans and for patients to make informed medical decisions. For early postoperative patients who have already undergone surgery, the early postoperative model can alert high-risk EM patients to shorten the interval for postoperative follow-up, enabling earlier detection of recurrence and prompt initiation of aggressive treatment.
The Child–Pugh classification serves as a scoring system for liver function and is currently adopted by many treatment guidelines for HCC [14]. Child–Pugh grade B encompasses varying degrees of hepatic impairment, and the actual clinical status of patients can differ significantly. As a result, most surgeons believe that such patients should not be excluded from surgical options from the outset [15]. Our study not only identifies Child–Pugh grade B as an independent risk factor for early EM but also demonstrates that, when combined with other risk factors, it can effectively predict the likelihood of EM in these patients.
The BCLC staging integrates tumor burden, liver function, and performance status, playing a crucial role in guiding treatment decisions for HCC [16]. Given that most HCC patients are diagnosed at intermediate to advanced stages, less than 30% are eligible for surgical intervention, and the overall prognosis for unresectable advanced HCC patients is extremely poor [17]. For some patients with relatively stable conditions in BCLC staging B and C, liver resection was attempted out of necessity. In our cohort, the incidence of EM among these patients was 31.75%. In this group (especially for patients in BCLC staging B), early identification of patients at low risk for EM is crucial for medical decision-making.
The TBS, introduced by Sasaki et al. [13] in 2018, is a prognostic tool initially developed for predicting outcomes in patients with colorectal liver metastases. Subsequent studies [18, 19] have extended the application of TBS to HCC patients undergoing hepatectomy, demonstrating its robust predictive capability for patient outcomes. Our study similarly supports the view that TBS is an independent risk factor for EM in patients; as TBS increases, the risk of EM also rises. On one hand, TBS minimizes the heterogeneity of tumor nodule size and quantity in HCC through the application of the Pythagorean theorem. On the other hand, unlike the binary classifications of the Milan criteria or the up-to-seven criteria, TBS serves as a simple and continuous variable that reflects the extent of tumor involvement in the liver. This approach mitigates inaccuracies in predictions that may arise from arbitrary cutoff values [20].
TO is a recently developed comprehensive metric for evaluating surgical quality, initially applied to gastrointestinal tumor resections and cardiovascular surgeries [21]. Studies have consistently demonstrated that achieving TO is associated with improved patient outcomes. In 2022, the IHPBA published an international expert consensus defining TOLS [7]. Our study confirms non-TO is an independent risk factor for early mortality following hepatectomy in HCC patients. Consistent with our findings, studies by Wu et al. [22] and Görgec et al. [23] have also demonstrated a strong correlation between TO and survival outcomes in patients undergoing liver resection. On one hand, the TO criteria encompass multiple dimensions of surgical quality, including intraoperative safety (e.g., absence of severe adverse events), postoperative recovery (e.g., absence of major complications), and oncological objectives (e.g., R0 resection), thereby enhancing the accuracy of prognostic predictions. On the other hand, the strong association between TO and patient outcomes underscores the importance for surgeons to improve TO rates across these dimensions, ultimately benefiting patients.
MVI is widely recognized as a critical factor associated with tumor recurrence and poor prognosis in HCC [24]. Our results corroborate this, identifying MVI as an independent risk factor for early postoperative mortality in HCC patients. While the predictive value of MVI is well-established, its limitation lies in the fact that it can only be definitively determined through postoperative pathological examination, rendering it impractical for preoperative surgical planning. However, recent advancements in artificial intelligence have enabled researchers to explore preoperative prediction of MVI using radiomics approaches, demonstrating promising predictive performance [25]. As research in this area progresses, preoperative assessment of MVI in HCC patients may become feasible, potentially providing more reliable guidance for treatment decision-making.
This study has several limitations that should be acknowledged. First, it is a single-center retrospective study, which may introduce selection bias. Second, due to the limited sample size, only internal validation was conducted, potentially leading to an overestimation of the model’s predictive performance. Third, the early mortality of HCC patients after liver resection is also influenced by the treatment methods employed after recurrence; however, this aspect was not included in our study due to practical constraints. Further validation of our findings through multicenter studies and prospective investigations is warranted.
In summary, our study indicates that independent preoperative risk factors for ED include Child–Pugh grade B, higher TBS, and more advanced BCLC staging. Early postoperative risk factors associated with ED comprise Child–Pugh grade B, elevated TBS, non-TO, and the presence of MVI. Patients with multiple risk factors—particularly those identified by the nomogram as high-risk for EM (predicted probability > 66%)—face a significantly elevated risk of early death.
Acknowledgements
Not applicable.
Clinical trial number
Not applicable.
Abbreviations
- HCC
Hepatocellular Carcinoma
- EM Group
Early Mortality Group
- NEM Group
Non-Early Mortality Group
- TBS
Tumor Burden Score
- MVI
Microvascular Invasion
- AFP
Alpha-Fetoprotein
- OS
Overall Survival
- BCLC staging
Barcelona Clinic Liver Cancer Stage
- HBV
Hepatitis B Virus
- CT
Computed Tomography
- TO
Textbook Outcome
- IHPBA
International Hepato-Pancreato-Biliary Association
- TOLS
Textbook Outcome in Liver Surgery
Authors’ contributions
CWK and RJ conceived and designed the study. HTZ and WTF collected the clinical data. CWK, HTZ, and LMF drafted the manuscript. CWK and LMF analyzed and interpreted the data. TTC and RJ critically revised the initial draft. All authors read and approved the final manuscript.
Funding
Not applicable.
Data availability
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
We confirm that all methods were performed in accordance declaration of Helsinki with the relevant guidelines and regulations and our research has been approved by The University of Hong Kong-Shenzhen Hospital Ethics Committee (hkuszh2024121). Since waiving the requirement for obtaining informed consent would not adversely affect participants’ rights, the Ethics Committee of The University of Hong Kong-Shenzhen Hospital approved the application for exemption from informed consent.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2024;74(3):229–63. [DOI] [PubMed] [Google Scholar]
- 2.Kuo MJ, Mo LR, Chen CL. Factors predicting long-term outcomes of early-stage hepatocellular carcinoma after primary curative treatment: the role of surgical or nonsurgical methods. BMC Cancer. 2021;21(1):250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kamiyama T, Nakanishi K, Yokoo H, Kamachi H, Tahara M, Kakisaka T, et al. Analysis of the risk factors for early death due to disease recurrence or progression within 1 year after hepatectomy in patients with hepatocellular carcinoma. World J Surg Oncol. 2012;10:107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sheriff S, Madhavan S, Lei GY, Chan YH, Junnarkar SP, Huey CW, et al. Predictors of mortality within the first year post-hepatectomy for hepatocellular carcinoma. J Egypt Natl Cancer Inst. 2022;34(1):14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Dan Y, Meng W, Li W, Chen Z, Lyu Y, Yu T. Transarterial chemoembolization combined with radiofrequency ablation versus hepatectomy for hepatocellular carcinoma: a meta-analysis. Front Surg. 2022;9:948355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Zheng L, Fang S, Wu F, Chen W, Chen M, Weng Q, et al. Efficacy and safety of TACE combined with Sorafenib plus immune checkpoint inhibitors for the treatment of intermediate and advanced TACE-refractory hepatocellular carcinoma: a retrospective study. Front Mol Biosci. 2020;7:609322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Görgec B, Benedetti Cacciaguerra A, Pawlik TM, Aldrighetti LA, Alseidi AA, Cillo U, et al. An international expert Delphi consensus on defining textbook outcome in liver surgery (TOLS). Ann Surg. 2023;277(5):821–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kazaryan AM, Røsok BI, Edwin B. Morbidity assessment in surgery: refinement proposal based on a concept of perioperative adverse events. ISRN surgery. 2013;2013:625093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Koch M, Garden OJ, Padbury R, Rahbari NN, Adam R, Capussotti L, et al. Bile leakage after hepatobiliary and pancreatic surgery: a definition and grading of severity by the International Study Group of Liver Surgery. Surgery. 2011;149(5):680–8. [DOI] [PubMed] [Google Scholar]
- 10.Rahbari NN, Garden OJ, Padbury R, Brooke-Smith M, Crawford M, Adam R, et al. Posthepatectomy liver failure: a definition and grading by the International Study Group of Liver Surgery (ISGLS). Surgery. 2011;149(5):713–24. [DOI] [PubMed] [Google Scholar]
- 11.Dindo D. The Clavien–Dindo Classification of Surgical Complications. In: Cuesta MA, Bonjer HJ, editors. Treatment of Postoperative Complications After Digestive Surgery. London: Springer; 2014. p. 13–7.
- 12.Reig M, Forner A, Rimola J, Ferrer-Fàbrega J, Burrel M, Garcia-Criado Á, et al. BCLC strategy for prognosis prediction and treatment recommendation: the 2022 update. J Hepatol. 2022;76(3):681–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sasaki K, Morioka D, Conci S, Margonis GA, Sawada Y, Ruzzenente A, et al. The Tumor Burden Score: A New, “Metro-ticket” Prognostic Tool For Colorectal Liver Metastases Based on Tumor Size and Number of Tumors. Ann Surg. 2018;267(1):132–41. [DOI] [PubMed] [Google Scholar]
- 14.Watanabe Y, Aikawa M, Kato T, Takase K, Watanabe Y, Okada K, et al. Influence of Child-Pugh B7 and B8/9 cirrhosis on laparoscopic liver resection for hepatocellular carcinoma: a retrospective cohort study. Surg Endosc. 2023;37(2):1316–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Cai X, Liang X, Yu T, Liang Y, Jing R, Jiang W, 等. Liver cirrhosis grading Child-Pugh class B: a Goliath to challenge in laparoscopic liver resection?-prior experience and matched comparisons. Hepatobiliary Surg Nutr. 2015;4(6): 391–397. [DOI] [PMC free article] [PubMed]
- 16.Wang MD, Xu XJ, Wang KC, Diao YK, Xu JH, Gu LH, et al. Conversion therapy for advanced hepatocellular carcinoma in the era of precision medicine: current status, challenges and opportunities. Cancer Sci. 2024;115(7):2159–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Li Q, Cao M, Lei L, Yang F, Li H, Yan X, et al. Burden of liver cancer: from epidemiology to prevention. Chin J Cancer Res. 2022;34(3):554–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Qiu ZC, Li C, Zhang Y, Xie F, Yu Y, Leng SS, et al. Tumor burden score-AFP-albumin-bilirubin grade score predicts the survival of patients with hepatocellular carcinoma after liver resection. Langenbecks Arch Surg. 2023;408(1):250. [DOI] [PubMed] [Google Scholar]
- 19.Wang J, Chen Z, Wang L, Feng S, Qiu Q, Chen D, et al. A new model based inflammatory index and tumor burden score (TBS) to predict the recurrence of hepatocellular carcinoma (HCC) after liver resection. Sci Rep. 2022;12(1):8670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ho S Y, Liu P H, Hsu C Y, Ko C C, Huang Y H, Su C W, 等. Tumor burden score as a new prognostic marker for patients with hepatocellular carcinoma undergoing transarterial chemoembolization. J Gastroenterol Hepatol. 2021;36(11): 3196–3203. [DOI] [PubMed]
- 21.Kolfschoten NE, Kievit J, Gooiker GA, van Leersum NJ, Snijders HS, Eddes EH, et al. Focusing on desired outcomes of care after colon cancer resections; hospital variations in “textbook outcome.” Eur J Surg Oncol. 2013;39(2):156–63. [DOI] [PubMed] [Google Scholar]
- 22.Wu DH, Liao CY, Wang DF, Huang L, Li G, Chen JZ, et al. Textbook outcomes of hepatocellular carcinoma patients with sarcopenia: a multicenter analysis. Eur J Surg Oncol. 2023;49(4):802–10. [DOI] [PubMed] [Google Scholar]
- 23.Görgec B, Benedetti Cacciaguerra A, Lanari J, Russolillo N, Cipriani F, Aghayan D, et al. Assessment of textbook outcome in laparoscopic and open liver surgery. JAMA Surg. 2021;156(8):e212064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kim KA, Kim MJ, Jeon HM, Kim KS, Choi JS, Ahn SH, et al. Prediction of microvascular invasion of hepatocellular carcinoma: usefulness of peritumoral hypointensity seen on gadoxetate disodium-enhanced hepatobiliary phase images. J Magn Reson Imaging. 2012;35(3):629–34. [DOI] [PubMed] [Google Scholar]
- 25.Liu P, Tan XZ, Zhang T, Gu QB, Mao XH, Li YC, et al. Prediction of microvascular invasion in solitary hepatocellular carcinoma ≤ 5 cm based on computed tomography radiomics. World J Gastroenterol. 2021;27(17):2015–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.