Abstract
Background
Hepatocellular carcinoma (HCC) imposes a substantial global burden. Triple therapy including transarterial chemoembolization, immune checkpoint inhibitors, and targeted therapy offers survival benefits but increases toxicity risk. We aimed to evaluate the impact of age or its related comorbidities on the efficacy and safety of triple therapy.
Methods
This multicenter retrospective study included 627 HCC patients receiving triple therapy from CLEAP database (2019–2023). Overall survival (OS) and progression-free survival (PFS) were the primary endpoints. Treatment response and treatment-related adverse events (TRAEs) were the secondary endpoints. Cox/logistic regression models were used to assess the risk factors of outcomes.
Results
Of 627 patients, 177 were classified as elderly (≥65 years) and 450 as younger (<65 years), as defined according to World Health Organization criteria. Elderly patients present with lower tumor burden compared with younger patients. After propensity score matching (1:2), two cohorts were comparable in objective response rate, disease control rate, PFS, OS, and grade ≥3 TRAEs. Age, either continuous or categorical data, was not associated with outcomes. In contrast, stratification by Charlson Comorbidity Index (CCI), there were significant outcome disparities for PFS and OS among low- (0–1), intermediate- (2–3), and high-risk (≥4) groups. The age-adjusted CCI (aCCI) could not only discriminate survival outcomes but also predict the risk of TRAEs among low- (0–3), intermediate- (4–5), and high-risk (≥6) groups. Multivariable analyses demonstrated that high aCCI score (≥6) independently predicted shorter PFS (hazard ratio: 1.36, p = 0.02) and OS (hazard ratio: 1.75, p = 0.001) and increased grade ≥ 3 TRAEs (odds ratio: 1.96, p = 0.005).
Conclusion
aCCI was a potent predictor of efficacy and safety in triple therapy for HCC.
Keywords: Age, Comorbidity, Hepatocellular carcinoma, Prognosis, Immunotherapy
Introduction
Hepatocellular carcinoma (HCC) represents a formidable global health challenge, ranking as the sixth most commonly malignancy and the third leading cause of cancer-related deaths worldwide [1]. Epidemiological studies have shown that more than 70% of patients are diagnosed at Barcelona Clinic Liver Cancer (BCLC) stage B or C [2, 3], meaning that a substantial proportion require nonsurgical treatment at the time of diagnosis. Meanwhile, population aging is markedly shifting the age structure of HCC patients [4]. Analysis of UK HCC incidence data (2015–2017; https://www.cancerresearchuk.org) reveals that 73% of newly diagnosed patients are aged ≥65 years, with over 43% being ≥75 years [5]. The rising proportion of elderly patients makes treatment decision-making at advanced stages increasingly complex.
For unresectable HCC, the European Association for the Study of the Liver (EASL), China Liver Cancer Staging (CNLC), and American Association for the Study of Liver Diseases (AASLD) guidelines recommend locoregional and/or systemic therapy [6–8]. In recent years, triple therapy including transarterial chemoembolization (TACE), tyrosine kinase inhibitors (TKIs), and immune checkpoint inhibitors (ICIs) has demonstrated superior survival benefits compared with TACE alone or systemic therapy alone. TACE combined with lenvatinib and pembrolizumab achieved a median progression-free survival (PFS) of 14.6 months – a 46% improvement over TACE alone (10 months) [9]. CHANCE2201 real-world trial demonstrated significantly improved overall survival (OS) (22.6 vs. 15.9 months) and PFS (9.9 vs. 7.4 months) with TACE-ICI-TKI versus ICI-TKI alone [10]. However, the incidence of grade ≥3 treatment-related adverse events (TRAEs) is substantially higher with triple therapy compared to monotherapy [11], raising particular concerns in elderly patients with reduced physiological reserves.
Elderly HCC patients frequently have reduced physical and nutritional reserves, multiple comorbidities, and age-related immune alterations [12, 13], which may affect responses to immunotherapy. Previous studies have yielded conflicting results on whether age predicts efficacy and toxicity in triple therapy. Some studies report higher treatment discontinuation rates in elderly patients, whereas others suggest similar or even better outcomes compared with younger patients [14, 15]. This observation suggests that chronological age may be only a crude surrogate for biological status, and its ability to accurately reflect individual risk remains inconclusive. In contrast, comorbidity burden can directly influence organ reserve, immune function, and treatment tolerance. The Charlson Comorbidity Index (CCI) is a widely used tool to quantify this burden. Its extension, the age-adjusted Charlson Comorbidity Index (aCCI), integrates age and comorbidities into a quantitative score and has been validated as a prognostic factor in other treatment modalities and tumor types, such as HCC surgery [16] and gallbladder cancer [17]. However, its predictive value in HCC patients undergoing triple therapy has not been systematically investigated.
This study aimed to evaluate the impact of chronological age and comorbidities on the treatment efficacy and safety among HCC patients receiving triple therapy.
Methods
Study Design and Patients
This multicenter study retrieved data between 2019 and 2023 from the China Liver Cancer Study Group Young Investigators (CLEAP) database. The eligibility criteria were as follows: (1) histologically or clinically confirmed HCC; (2) adult patients received triple therapy as first-line treatment for conversion; (3) at least two cycles of ICIs and targeted therapy; (4) Eastern Cooperative Oncology Group Performance Status (ECOG PS) <2; (5) Child-Pugh score ≤7. The exclusion criteria were (1) presence of additional malignancies other than HCC and (2) missing follow-up data. Oncologic outcomes, along with clinical and radiological information, were obtained from patients’ medical records. Baseline was defined as the date of initiation of TACE or the start of ICIs and targeted therapy. The study flowchart is shown in Figure 1.
Fig. 1.
Flowchart of patient selection and analytical methods in the study.
Data Collection and Endpoints
Patients aged ≥65 years were classified as elderly by World Health Organization (WHO) [18]. Baseline variables included age, sex, ECOG-PS, BCLC stage, alpha-fetoprotein (AFP) level, Child-Pugh class, hepatitis B virus (HBV), protein induced by vitamin K absence or antagonist-II (PIVKA-II) level, the presence of distant metastasis.
The primary endpoint of the study was OS and PFS. Secondary endpoints included objective response rate (ORR), disease control rate (DCR), and TRAEs. The tumor best overall response (BOR) was evaluated according to the modified Response Evaluation Criteria in Solid Tumors (mRECIST) [19], with computed tomography scans performed every 8–10 weeks. ORR was calculated as the proportion of patients achieving complete response or partial response, while DCR included complete response, partial response, or stable disease. PFS was defined as the time from treatment initiation to tumor progression or death from any cause and OS as the time from treatment initiation to death from any cause. TRAEs were graded according to the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. The CCI score was calculated according to Charlson et al. [20, 21], and aCCI was calculated by further adding one point for each decade above 40 years of age [22]. Specifically, myocardial infarct, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, connective tissue disease, ulcer disease, mild liver disease, hypertension, and diabetes are assigned 1 point each. Hemiplegia, moderate or severe renal disease, diabetes with end organ damage, any tumor, leukemia, and lymphoma are assigned 2 points each. Moderate or severe liver disease is assigned 3 points. Metastatic solid tumor and AIDS are assigned 6 points each. Given all patients had HCC, HCC is excluded from the aCCI score calculation. This approach, consistent with other studies [20, 23, 24], ensures that the score exclusively reflects the burden of comorbidities other than the primary disease under investigation, thereby providing an unbiased assessment of its independent prognostic value. For risk stratification, the CCI and aCCI were stratified into three groups: low, medium, and high. The cutoffs for this stratification were determined based on the distribution of the scores in our cohort. Specifically, the lower cutoff aligns with the first quartile (Q1), and the upper cutoff aligns with the third quartile (Q3) of the CCI and aCCI distribution.
Statistical Analysis
The Shapiro-Wilk test was used to assess the normality of continuous variables. Normally distributed variables were presented as mean ± standard deviation, while non-normally distributed variables were expressed as median and interquartile range. Categorical variables were summarized as frequencies and percentages. Between-group comparisons were performed using independent t tests for normally distributed continuous variables, Mann-Whitney U tests or Kruskal-Wallis tests for non-normally distributed variables, chi-square tests or Fisher’s exact tests for categorical variables, and Wilcoxon rank-sum tests for ordinal categorical variables. PFS and OS were estimated by the Kaplan-Meier method and compared with the log-rank test. Propensity score matching (PSM) was applied to balance baseline differences between groups. Multivariate Cox proportional-hazards regression models were employed to identify independent prognostic factors for survival, with results reported as hazard ratios (HRs) and 95% confidence intervals (CIs). Logistic regression analysis was used to identify factors associated with grade ≥3 TRAEs. All statistical tests were two sided, and p value <0.05 was considered statistically significant. Analyses were performed using R software, version 4.5.1.
Results
Comparable Treatment Efficacy and Safety between Elderly and Younger Patients
Demographic and clinical characteristics of the 627 eligible patients, stratified by age-group (<65 years vs. ≥65 years) according to WHO criteria [18], are shown in online supplementary Table 1 (for all online suppl. material, see https://doi.org/10.1159/000550356). The median age of elderly patients was 70 years (interquartile range: 67–73). Compared with younger patients, elderly patients had significantly lower proportions of BCLC stage C disease, lower AFP and PIVKA-II levels, and fewer HBV-related cases (all p < 0.05).
To minimize baseline imbalances of demographic and clinical characteristics, PSM was performed, yielding a cohort of 422 patients with well-balanced covariates (standardized mean difference <0.1) (shown in Table 1). After PSM, treatment efficacy was comparable between elderly and younger patients (shown in online suppl. Table 2). No significant differences were observed in best response (p = 0.421), ORR (70.9% vs. 72.7%, p = 0.737), or DCR (95.6% vs. 95.1%, p = 0.999). The surgical conversion rate was numerically lower in the elderly group (36.7% vs. 45.8%) but did not reach statistical significance (p = 0.068). Kaplan-Meier analysis demonstrated no significant difference in median OS (20.1 vs. 23.4 months; HR: 0.80; 95% CI: 0.58–1.10; p = 0.172, log rank) (shown in Fig. 2a) or median PFS (17.4 vs. 16.7 months; HR: 0.96; 95% CI: 0.74–1.24; p = 0.752, log rank) (shown in Fig. 2b).
Table 1.
Comparison of baseline characteristics based on age (post-PSM)
| | Overall (N = 422) | Aged ≥65 years (n = 158) | Aged <65 years (n = 264) | p value |
|---|---|---|---|---|
| Age, median (IQR), years | 60.5 (54.0, 68.0) | 69.5 (67.0, 73.0) | 56.0 (50.0, 60.0) | <0.001 |
| Sex, n (%) | | | | 0.200 |
| Male | 375 (89) | 136 (86) | 239 (91) | |
| Female | 47 (11) | 22 (14) | 25 (9.5) | |
| BMI, median (IQR) | 22.9 (20.7, 25.2) | 22.5 (20.4, 25.4) | 23.1 (20.8, 25.2) | 0.307 |
| ECOG, n (%) | | | | 0.532 |
| 0 | 266 (63) | 103 (65) | 163 (62) | |
| 1 | 156 (37) | 55 (35) | 101 (38) | |
| BCLC, n (%) | | | | 0.911 |
| A | 116 (27) | 42 (27) | 74 (28) | |
| B | 134 (32) | 52 (33) | 82 (31) | |
| C | 172 (41) | 64 (41) | 108 (41) | |
| AFP, median (IQR), ng/mL | 36.0 (4.6, 528.0) | 32.8 (4.4, 309.0) | 38.1 (4.8, 813.2) | 0.174 |
| Child-Pugh classification, n (%) | | | | 0.404 |
| A | 380 (90) | 145 (92) | 235 (89) | |
| B | 42 (10.0) | 13 (8.2) | 29 (11) | |
| HBV, n (%) | 341 (81) | 121 (77) | 220 (83) | 0.098 |
| PIVKA-II, median (IQR), ng/mL | 1,176.0 (113.0, 10,024.0) | 936.9 (94.0, 9,164.0) | 1,233.0 (159.5, 11,078.1) | 0.436 |
| Metastasis, n (%) | 31 (7.3) | 9 (5.7) | 22 (8.3) | 0.343 |
| Number of TACE, n (%) | | | | 0.551 |
| 1 | 194 (46) | 76 (48) | 118 (45) | |
| 2 | 120 (28) | 40 (25) | 80 (30) | |
| ≥3 | 108 (26) | 42 (27) | 66 (25) | |
| Target agents, n (%) | | | | 0.671 |
| Bevacizumab | 145 (34) | 53 (34) | 92 (35) | |
| Donafenib | 105 (25) | 39 (25) | 66 (25) | |
| Lenvatinib | 159 (38) | 59 (37) | 100 (38) | |
| Other | 13 (3.1) | 7 (4.4) | 6 (2.3) | |
| ICIs, n (%) | | | | 0.587 |
| Camrelizumab | 50 (12) | 15 (9.5) | 35 (13) | |
| Sintilimab | 193 (46) | 71 (45) | 122 (46) | |
| Tislelizumab | 156 (37) | 62 (39) | 94 (36) | |
| Other | 23 (5.5) | 10 (6.3) | 13 (4.9) | |
ECOG, Eastern Cooperative Oncology Group; BCLC, Barcelona Clinic Liver Cancer; AFP, alpha-fetoprotein; HBV, hepatitis B virus; PIVKA-II, protein induced by vitamin K absence or antagonist-II; ICIs, immune checkpoint inhibitors; IQR, interquartile range; n, number.
Fig. 2.
Survival outcomes of HCC patients treated with triple therapy stratified by age and treatment agents. a Kaplan-Meier curves for overall survival (OS) stratified by age. There was no significant difference between the patients aged ≥65 years and <65 years. b Kaplan-Meier curves for progression-free survival (PFS) stratified by age. There was no significant difference between the two groups. c Kaplan-Meier curves for OS stratified by targeted agents. There was no significant difference among patients receiving bevacizumab, donafenib, and lenvatinib. d Kaplan-Meier curves for PFS stratified by targeted agents. There was no significant difference among the three groups. e Kaplan-Meier curves for OS stratified by immune checkpoint inhibitors (ICIs). There was no significant difference among patients receiving camrelizumab, sintilimab, and tislelizumab. f Kaplan-Meier curves for PFS stratified by ICIs. There was no significant difference among the three groups.
As shown in online supplementary Table 3, the incidence of grade ≥3 TRAEs was similar between elderly and younger patients (30% vs. 22%, p = 0.104, Fisher’s exact test). In elderly patients, the most frequent TRAEs of any grade were liver function impairment (66%), renal function impairment (32%), and skin reaction (24%). Notably, renal impairment (p = 0.008) and pulmonary impairment (p = 0.021) occurred more frequently in elderly patients. However, these differences did not remain statistically significant after correction for multiple comparisons using the false discovery rate method (FDR p = 0.068 and 0.093, respectively).
In addition, different targeted agents and ICIs were included in this cohort. To eliminate the impact of agent heterogeneity on the efficacy and toxicity analyses, we performed subgroup analyses. For ORR, significant differences were observed among the three groups (p = 0.007) (shown in online suppl. Table 4), and this difference was specifically driven by the comparison between patients treated with bevacizumab and lenvatinib (64.0% vs. 77.4%, FDR-adjusted p = 0.006). In contrast, as shown in Figure 2c–f, PFS (p = 0.987 and 0.624) and OS (p = 0.119 and 0.485) did not differ significantly among treatments, and the incidence of grade ≥3 TRAEs was also comparable across different targeted agents and ICIs.
CCI Stratifies Survival but Fails to Predict Toxicity
We assessed the prevalence of individual comorbidities of CCI in this HCC cohort. The most prevalent conditions were mild liver disease (32.9%), chronic pulmonary disease (21.5%), and diabetes (15.3%). Patients were stratified into low-risk (0–1), intermediate-risk (2–3), and high-risk (≥4) groups based on the Q1 and Q3 of the CCI distribution. Treatment responses did not differ significantly among the three groups (shown in online suppl. Table 5). Kaplan-Meier survival curves demonstrated significant differences in both OS (p < 0.001) and PFS (p = 0.005) across the three groups (shown in Fig. 3a, b). Pairwise analyses showed worse OS and PFS in high- versus low-risk group (p < 0.001 and p = 0.011) and high- versus intermediate-risk groups (p = 0.013 and p = 0.011). However, the incidence of TRAEs did not differ significantly among the three groups (shown in online suppl. Table 6).
Fig. 3.
Survival outcomes of HCC patients treated with triple therapy stratified by CCI and aCCI score: low, intermediate, and high. a Kaplan-Meier curves for overall survival (OS) stratified by Charlson Comorbidity Index (CCI). OS was significantly reduced in the high CCI group relative to the intermediate and low CCI groups. b Kaplan-Meier curves for progression-free survival (PFS) stratified by CCI. The high CCI group exhibited significantly shorter PFS compared with the intermediate and low CCI groups. c Kaplan-Meier curves for overall survival (OS) stratified by age-adjusted Charlson Comorbidity Index (aCCI). OS was significantly reduced in the high aCCI group relative to the intermediate and low aCCI groups. d Kaplan-Meier curves for progression-free survival (PFS) stratified by aCCI. The high aCCI group exhibited significantly shorter PFS compared with the intermediate and low aCCI groups.
Next, the association between individual comorbidities and prognosis was analyzed. Comorbidities with prevalence <1% were excluded to ensure adequate sample size for robust statistical inference. Univariate Cox regression analysis identified chronic pulmonary disease (HR: 1.44, 95% CI: 1.08–1.90, p = 0.011) and mild liver disease (HR 1.36, 95% CI: 1.05–1.77, p = 0.021) as risk factors for poor OS. Both conditions were also associated with worse PFS, with chronic pulmonary disease showing HR: 1.48 (95% CI: 1.17–1.86, p = 0.001) and mild liver disease showing HR: 1.49 (95% CI: 1.21–1.84, p < 0.001). After adjustment for other variables, these two comorbidities remained independent risk factors for poor OS and PFS (all p < 0.05). In contrast, univariate logistic regression analysis for grade ≥3 TRAEs revealed no individual comorbidity as an independent risk factor.
As a further exploration, we conducted subgroup analyses by different agents. Analysis within each ICIs subgroup showed a consistent trend toward poorer PFS and OS with higher CCI grade. A statistically significant differential effect of CCI stratification was confirmed in the tislelizumab subgroup but not in the camrelizumab or sintilimab subgroups (shown in online suppl. Fig. 1). A similar pattern was observed among targeted agents, where the association reached significance in the donafenib subgroup, but not in the lenvatinib or bevacizumab subgroups (shown in online suppl. Fig. 2).
High aCCI Associated with Poor Survival and Increased Toxicity
The entire cohort was also stratified according to aCCI score into low-risk (0–3), intermediate-risk (4–5), and high-risk (≥6) groups, with baseline characteristics balanced among groups (shown in Table 2). The surgical conversion rate was lowest in the high aCCI group (p = 0.032), while ORR and DCR did not differ significantly among the groups (shown in online suppl. Table 7). The 2-year OS rates were 46.7% in the low aCCI group, 41.9% in the intermediate group, and 23.0% in the high group (p = 0.016). Kaplan-Meier curves revealed significant differences in OS (p = 0.002) (shown in Fig. 3c) and PFS (p = 0.040) (shown in Fig. 3d) among the three groups. Pairwise comparisons showed significantly worse OS in the high aCCI group compared with the low (p = 0.005, FDR corrected) and intermediate (p = 0.006) groups, while no significant difference was observed between the low and intermediate groups. Similar patterns were found for PFS (p = 0.044 for both comparisons).
Table 2.
Comparison of baseline characteristics based on aCCI score
| | Low aCCI (n = 222) | Intermediate aCCI (n = 232) | High aCCI (n = 173) | p value |
|---|---|---|---|---|
| Age, median (IQR), years | 53 (10) | 61 (10) | 60 (12) | <0.001 |
| BMI, median (IQR) | 22.7 (3.2) | 23.0 (3.2) | 22.9 (3.4) | 0.874 |
| Sex, n (%) | | | | 0.202 |
| Male | 200 (90) | 206 (89) | 146 (84) | |
| Female | 22 (9.9) | 26 (11) | 27 (16) | |
| ECOG, n (%) | | | | 0.682 |
| 0 | 137 (62) | 148 (64) | 103 (60) | |
| 1 | 85 (38) | 84 (36) | 70 (40) | |
| Child-Pugh classification, n (%) | | | | 0.570 |
| A | 199 (90) | 210 (91) | 151 (87) | |
| B | 23 (10) | 22 (9.5) | 22 (13) | |
| HBV, n (%) | 186 (84) | 188 (81) | 140 (81) | 0.684 |
| AFP, median (IQR), ng/mL | 14,836 (30,441) | 11,815 (29,583) | 10,318 (25,610) | 0.237 |
| PIVKA-II, median (IQR), ng/mL | 17,922 (25,764) | 13,701 (21,950) | 14,190 (22,629) | 0.080 |
| BCLC, n (%) | | | | 0.251 |
| A | 44 (20) | 64 (28) | 38 (22) | |
| B | 58 (26) | 63 (27) | 43 (25) | |
| C | 120 (54) | 105 (45) | 92 (53) | |
| Metastasis, n (%) | 24 (11) | 22 (9.5) | 19 (11) | 0.855 |
| Number of TACE, n (%) | | | | 0.197 |
| 1 | 97 (44) | 107 (46) | 70 (40) | |
| 2 | 73 (33) | 56 (24) | 56 (32) | |
| ≥3 | 52 (23) | 69 (30) | 47 (27) | |
| Target agents, n (%) | | | | 0.314 |
| Bevacizumab | 62 (28) | 70 (30) | 68 (39) | |
| Donafenib | 60 (27) | 65 (28) | 40 (23) | |
| Lenvatinib | 93 (42) | 91 (39) | 59 (34) | |
| Other | 7 (3.2) | 6 (2.6) | 6 (3.5) | |
| ICIs, n (%) | | | | 0.443 |
| Camrelizumab | 28 (13) | 25 (11) | 21 (12) | |
| Sintilimab | 102 (46) | 97 (42) | 82 (47) | |
| Tislelizumab | 83 (37) | 102 (44) | 59 (34) | |
| Other | 9 (4.1) | 8 (3.4) | 11 (6.4) | |
aCCI, age-adjusted Charlson Comorbidity Index; ECOG, Eastern Cooperative Oncology Group; BCLC, Barcelona Clinic Liver Cancer; AFP, alpha-fetoprotein; HBV, hepatitis B virus; PIVKA-II, protein induced by vitamin K absence or antagonist-II; ICIs, immune checkpoint inhibitors; IQR, interquartile range; n, number.
Furthermore, the incidence of grade ≥3 TRAEs was highest in the high aCCI group (34%, p = 0.017) (shown in Table 3), with gastrointestinal, pulmonary, and musculoskeletal events more frequent in this group. Given that neither age nor CCI showed association with TRAEs, we specifically examined their interaction. The results showed that the age-CCI interaction term significantly affected the incidence of grade ≥3 TRAEs (p = 0.005). As shown in Figure 4, in younger patients, higher CCI did not significantly elevate the risk of grade ≥3 TRAEs, whereas high CCI was notably associated with a marked increase in grade ≥3 TRAEs in older patients.
Table 3.
Comparison of TRAEs based on aCCI score
| | Low aCCI (n = 222) | Intermediate aCCI (n = 232) | High aCCI (n = 173) | p value |
|---|---|---|---|---|
| Any grade TRAEs, n (%) | 189 (85) | 197 (85) | 158 (91) | 0.114 |
| Grade ≥3 TRAEs, n (%) | 48 (22) | 58 (25) | 59 (34) | 0.017 |
| Liver, n (%) | 152 (68) | 142 (61) | 119 (69) | 0.168 |
| Renal, n (%) | 42 (19) | 61 (26) | 41 (24) | 0.169 |
| Skin, n (%) | 43 (19) | 62 (27) | 43 (25) | 0.164 |
| Endocrine, n (%) | 18 (8.1) | 18 (7.8) | 19 (11) | 0.478 |
| Gastrointestinal, n (%) | 17 (7.7) | 40 (17) | 29 (17) | 0.005 |
| Pulmonary, n (%) | 4 (1.8) | 3 (1.3) | 9 (5.2) | 0.040 |
| Musculoskeletal, n (%) | 5 (2.3) | 5 (2.2) | 11 (6.4) | 0.035 |
| Hematological, n (%) | 32 (14) | 24 (10) | 27 (16) | 0.246 |
| Cardiac, n (%) | 1 (0.5) | 2 (0.9) | 5 (2.9) | 0.114 |
aCCI, age-adjusted Charlson Comorbidity Index; TRAEs, treatment-related adverse events; n, number.
Fig. 4.
Assessment of the interaction between age and the Charlson Comorbidity Index (CCI) on the incidence of grade ≥3 treatment-related adverse events (TRAEs).
aCCI Was an Independent Risk Factor for Prognosis
In univariate Cox analysis, when age was analyzed as a continuous variable, categorized (≤40, 40–49, 50–59, 60–69, ≥70), or dichotomized (<65 vs. ≥65 years), it was not significantly associated with OS, PFS, or grade ≥3 TRAEs (shown in Table 4). In contrast, CCI (as both continuous variable and stratification) was significantly associated with worse PFS and OS (all p < 0.05). The aCCI stratification, together with BCLC stage, and AFP level were significantly associated with OS (shown in Table 5), while aCCI stratification, ECOG performance status, BCLC stage, and AFP level were significantly associated with PFS (shown in online suppl. Table 8). When analyzed as a continuous variable, the aCCI remained significantly associated with OS (HR: 1.10; 95% CI: 1.04–1.16; p = 0.002), PFS (HR: 1.06; 95% CI: 1.02–1.11; p = 0.006), and grade ≥3 TRAEs (HR: 1.11; 95% CI: 1.03–1.20; p = 0.006). Furthermore, spline modeling demonstrated that the aCCI was a significant and predominantly linear predictor across all endpoints, including OS (p = 0.019), PFS (p = 0.013), and grade ≥3 TRAEs (p = 0.018). The aCCI showed higher statistical power than CCI in predicting prognosis.
Table 4.
Prognostic performance of age for PFS, OS, and TRAEs
| | OS | PFS | Grade ≥3 TRAEs | |||
|---|---|---|---|---|---|---|
| HR (95% CI) | p value | HR (95% CI) | p value | OR (95% CI) | p value | |
| Age, years | 1.01 (0.99, 1.02) | 0.428 | 1.01 (0.99, 1.02) | 0.389 | 0.99 (0.96, 1.01) | 0.154 |
| Age group | ||||||
| ≥65 years | 1.00 | | 1.00 | | 1.00 | |
| <65 years | 0.80 (0.58, 1.10) | 0.173 | 0.96 (0.74–1.24) | 0.750 | 1.47 (0.94–2.3) | 0.091 |
| Age stratification | ||||||
| <40 years | 1.00 | | 1.00 | | | |
| 40–49 years | 1.39 (0.53, 3.59) | 0.498 | 0.84 (0.40, 1.76) | 0.636 | 0.45 (0.12, 1.51) | 0.210 |
| 50–59 years | 1.42 (0.64, 3.12) | 0.389 | 1.04 (0.60, 1.82) | 0.882 | 0.91 (0.28, 2.46) | 0.860 |
| 60–69 years | 0.94 (0.42, 2.08) | 0.877 | 1.05 (0.61, 1.82) | 0.852 | 0.65 (0.21, 1.72) | 0.420 |
| ≥70 years | 1.64 (0.73, 3.66) | 0.229 | 1.18 (0.67, 2.10) | 0.567 | 0.49 (0.15, 1.36) | 0.197 |
PFS, progression-free survival; OS, overall survival; TRAEs, treatment-related adverse events; HR, hazard ratio; CI, confidence interval.
Table 5.
Univariate and multivariable analysis of OS by Cox model
| | Univariate analysis | Multivariable analysis | ||
|---|---|---|---|---|
| HR (95% CI) | p value | HR (95% CI) | p value | |
| aCCI score | ||||
| Low | 1 | | | |
| Intermediate | 1.10 (0.80–1.51) | 0.559 | | |
| High | 1.70 (1.23–2.36) | 0.001 | 1.73 (1.23–2.41) | 0.001 |
| Child-Pugh classification | ||||
| A | 1 | | | |
| B | 1.08 (0.73–1.60) | 0.704 | | |
| ECOG | ||||
| 0 | 1 | | | |
| 1 | 0.94 (0.72–1.23) | 0.653 | | |
| BCLC | ||||
| A | 1 | | | |
| B | 1.39 (0.95–2.05) | 0.091 | | |
| C | 1.47 (1.04–2.06) | 0.027 | 1.43 (1.00–2.06) | 0.049 |
| AFP | ||||
| <400 ng/mL | 1 | | | |
| ≥400 ng/mL | 1.26 (0.97–1.64) | 0.079 | | |
| TACE, n | ||||
| 1 | 1 | | | |
| 2 | 0.76 (0.55–1.06) | 0.109 | | |
| ≥3 | 0.98 (0.73–1.32) | 0.903 | | |
| Target agents | ||||
| Bevacizumab | 1 | | | |
| Donafenib | 0.86 (0.61–1.20) | 0.368 | | |
| Lenvatinib | 1.20 (0.88–1.62) | 0.249 | | |
| Other | 0.87 (0.35–2.16) | 0.767 | | |
| ICIs | ||||
| Camrelizumab | 1 | | | |
| Sintilimab | 0.87 (0.57–1.31) | 0.491 | | |
| Tislelizumab | 0.78 (0.51–1.19) | 0.253 | | |
| Other | 0.75 (0.35–1.58) | 0.447 | | |
OS, overall survival; aCCI, age-adjusted Charlson Comorbidity Index; ECOG, Eastern Cooperative Oncology Group; BCLC, Barcelona Clinic Liver Cancer; AFP, alpha-fetoprotein; ICIs, immune checkpoint inhibitors; HR, hazard ratio; CI, confidence interval.
In multivariate Cox analysis, high aCCI was identified as an independent risk factor for shorter OS (HR: 1.73; 95% CI: 1.24–2.41; p = 0.001) (shown in Table 5) and PFS (HR: 1.36; 95% CI: 1.04–1.78; p = 0.023) (shown in online suppl. Table 8). In addition, the univariate and multivariate logistic regression analyses indicated that high aCCI was independently associated with a higher risk of grade ≥3 TRAEs (odds ratio [OR]: 1.96; 95% CI: 1.23–3.11; p = 0.005) (shown in Table 6). Consistently, as a continuous variable, the aCCI remained an independent risk factor for shorter OS (HR: 1.10; 95% CI: 1.04–1.17; p = 0.001), worse PFS (HR: 1.06; 95% CI: 1.02–1.11; p = 0.006), and grade ≥3 TRAEs (OR: 1.12; 95% CI: 1.04–1.21; p = 0.004).
Table 6.
Univariate and multivariable analysis of grade ≥3 TRAEs by logistic regression
| | Univariate analysis | Multivariable analysis | ||
|---|---|---|---|---|
| OR (95% CI) | p value | OR (95% CI) | p value | |
| aCCI score | ||||
| Low | 1 | | | |
| Intermediate | 1.21 (0.78–1.87) | 0.395 | | |
| High | 1.88 (1.20–2.95) | 0.006 | 1.96 (1.23–3.11) | 0.005 |
| Child-Pugh classification | ||||
| A | 1 | | | |
| B | 2.38 (1.41–4.00) | 0.001 | 2.25 (1.29–3.91) | 0.004 |
| ECOG | ||||
| 0 | 1 | | | |
| 1 | 1.15 (0.8–1.66) | 0.443 | | |
| BCLC | ||||
| A | 1 | | | |
| B | 1.77 (1.03–3.12) | 0.042 | 1.85 (1.05–3.26) | 0.032 |
| C | 2.10 (1.30–3.50) | 0.003 | 1.97 (1.17–3.33) | 0.011 |
| AFP | ||||
| <400 ng/mL | 1 | | | |
| ≥400 ng/mL | 1.28 (0.89–1.83) | 0.175 | | |
| TACE | ||||
| 1 | 1 | | | |
| 2 | 0.89 (0.58–1.37) | 0.608 | | |
| ≥3 | 0.99 (0.64–1.52) | 0.959 | | |
| Target agents | ||||
| Bevacizumab | 1 | | | |
| Donafenib | 1.19 (0.73–1.90) | 0.464 | | |
| Lenvatinib | 1.06 (0.69–1.63) | 0.806 | | |
| Other | 2.24 (0.83–5.86) | 0.102 | | |
| ICIs | ||||
| Camrelizumab | 1 | | | |
| Sintilimab | 0.52 (0.31–0.91) | 0.020 | | |
| Tislelizumab | 0.55 (0.32–0.96) | 0.032 | 0.49 (0.26–0.94) | 0.030 |
| Other | 0.66 (0.24–1.65) | 0.384 | | |
TRAEs, treatment-related adverse events; aCCI, age-adjusted Charlson Comorbidity Index; ECOG, Eastern Cooperative Oncology Group; BCLC, Barcelona Clinic Liver Cancer; AFP, alpha-fetoprotein; ICIs, immune checkpoint inhibitors; OR, odds ratio; CI, confidence interval.
Discussion
This study demonstrates that aCCI is a stronger predictor of prognosis and treatment safety in HCC patients receiving triple therapy than chronological age and CCI. In this multicenter retrospective analysis, patients aged ≥65 years showed no significant differences from younger patients in ORR, DCR, PFS, or OS, and the overall incidence of adverse events was comparable between the two groups. CCI could stratify patients with different PFS and OS but not the risk of TRAEs. In contrast, patients with higher aCCI scores had shorter PFS and OS as well as a higher incidence of severe TRAEs.
Notably, these findings were based on the PSM cohort, in which baseline comparability had been ensured. Within this matched population, whether age was analyzed as a categorical variable (≥65 years vs. <65 years), a continuous variable, or stratified in 10-year intervals, the results consistently indicated that chronological age was not an independent predictor of the efficacy or safety of triple therapy. Such findings are consistent with previous studies. For example, a study with a median age of 65 years in HCC patients receiving triple therapy also found no significant differences in PFS or safety between older and younger patients, with median PFS of 15.0 months and 8.2 months, respectively [11]. Similar findings have been reported in other cancer types, such as head and neck squamous cell carcinoma, where age was not associated with treatment efficacy or toxicity risk [25], and meta-analyses have likewise shown comparable efficacy of ICIs across age-groups [26]. Collectively, these results further support that advanced age per se should not preclude the use of triple therapy in clinical practice.
To further dissect the roles of age and comorbidity, we subsequently analyzed the CCI. This analysis confirmed that comorbidity burden per se was a powerful, independent predictor of survival, with the high CCI group showing significantly worse PFS and OS. Analysis across all regimen subgroups demonstrated a consistent trend associating higher CCI scores with poorer PFS and OS. This association reached statistical significance, specifically in the tislelizumab and donafenib subgroups. Due to its unique Fc engineering, tislelizumab exhibits therapeutic efficacy that is more purely dependent on the host’s immune status. Consequently, it may be more sensitive to the CCI score, which reflects immunosenescence [27]. Similarly, as a deuterated derivative, donafenib possesses unique pharmacokinetic properties that render its efficacy and toxicity more dependent on organ reserve function [28], for which the CCI score is a classic assessment tool. However, the CCI alone did not predict the risk of TRAEs, indicating limitations in its ability to capture toxicity risk. A deeper investigation into the interaction between age and CCI revealed a critical finding: while higher CCI did not increase the risk of grade ≥3 TRAEs in younger patients, it was associated with a marked elevation in risk among older patients. Older patients exhibit a decline in the functional capacity of key organs responsible for drug metabolism and clearance [29]. When a high comorbidity burden is superimposed on this already compromised organ functional baseline, the cumulative strain likely exceeds a critical threshold [30]. This synergy leads to altered pharmacokinetics, increased drug toxicity, and a significantly elevated risk of severe TRAEs [31]. This underscores that the confluence of advanced age and comorbidity burden, rather than either factor alone, predisposes patients to severe organ injury. The imperative to move beyond composite scores and identify specific organ vulnerabilities prompted a granular analysis of the CCI’s components. We investigated the association of individual comorbidities (prevalence >1%) with survival and toxicity. This deconstruction revealed that the prognostic power of the CCI is not uniformly distributed but is primarily driven by specific organ dysfunctions. Mild liver disease and chronic pulmonary disease were key independent risk factors for worse OS and PFS. Previous studies suggest that chronic pulmonary disease probably influences the efficacy of triple therapy by causing T-cell imbalance and suppressing immune cell function, which may weaken the therapeutic effect [32, 33]. Similarly, evidence has shown that hepatic dysfunction and immune microenvironment disruption associated with liver cirrhosis jointly led to poorer prognosis [34, 35]. Conversely, none of the individual comorbidities were an independent risk factor for grade ≥3 TRAEs. This reinforces that the survival is heavily influenced by the burden of specific comorbidities that compromise vital organ function, whereas severe toxicity risk arises from a more complex interplay, notably the interaction between age and overall comorbidity burden.
To capture this complexity, the aCCI incorporates chronological age and comorbidities closely related to systemic function and organ tolerance and has been applied in Chinese populations [36], potentially making it more relevant to the biological characteristics of triple therapy and providing a robust basis for risk stratification. In our study, a higher aCCI was associated with a lower conversion rate, worse survival outcomes, and an increased risk of severe TRAEs. External evidence provides clinical support for these observations. In surgical cohorts of HCC, patients with higher aCCI experienced significantly worse OS and, at recurrence, were less likely to undergo repeat resection and more likely to receive palliative treatment [16]. In patients treated with TKIs, a higher Charlson Index independently predicted severe and dose-limiting toxicities, reflecting poorer tolerance and a greater likelihood of dose modification or early discontinuation [37]. Furthermore, an analysis of the USA National Cancer Database (NCDB) showed that HCC patients with Charlson Index ≥3 were significantly more likely to receive no cancer-directed therapy at all, underscoring the restrictive impact of comorbidity burden on treatment deliverability [38]. Taken together, these findings suggest that a higher aCCI reflects diminished systemic reserve and organ tolerance due to heavier comorbidity burden, thereby reducing patients’ ability to withstand intensive multimodal therapy and damaging survival outcome. It also helps identify those who require closer monitoring, early supportive care, or treatment modification. Overall, these findings reinforce the value of aCCI in clinical risk stratification beyond chronological age and CCI.
While this study drew on a relatively large, multicenter cohort and, to our knowledge, represents the first systematic evaluation of the prognostic and safety relevance of aCCI in HCC patients undergoing triple therapy, several limitations should be noted. First, its retrospective design inevitably carries a risk of selection and information bias, and although PSM was applied, residual confounding cannot be completely ruled out. Second, all patients were recruited from Chinese centers, predominantly with HBV-related HCC, which may limit the applicability of the findings to populations in Japan and Western countries where other etiologies such as HCV, alcohol, or metabolic-associated liver disease are more common [39]. Given this limitation, the prognostic performance of aCCI observed in our study indicates that its utility should be validated across diverse etiological subgroups. Third, treatment allocation was non-randomized and may have been influenced by physician judgment, patient condition, or socioeconomic factors, and variations in the specific drug regimens used could also have affected outcomes. Fourth, our elderly group was representative of patients in their seventies but included very few octogenarians (n = 7). The generalizability of our conclusions to patients aged over 80 years remains uncertain and warrants further investigation. A further limitation is that in this triple therapy setting, the clinical manifestations of many TRAEs are highly overlapping and nonspecific. Combination therapies may induce “synergistic toxicities,” where toxic effects of different treatment modalities are mutually amplified. Thus, it is difficult to distinguish between immune-related and non-immune-related events for analysis. Finally, the aCCI stratification strategy requires future prospective, multicenter studies in elderly and etiologically diverse patient populations to confirm its utility in guiding personalized treatment decisions.
In summary, our findings indicated that aCCI, by capturing comorbidity-related systemic function and organ reserve, provided a comprehensive and clinically relevant risk assessment in HCC patients receiving triple therapy. This index might serve as a practical tool to bridge the gap between biological frailty and therapeutic intensity, guiding clinicians toward treatment strategies that maximize benefit while minimizing harm.
Statement of Ethics
This study involving human participants was reviewed and approved by Zhongshan Hospital, Fudan University (Approval No. B2022-195R). Written informed consent was obtained.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
The study was supported by the National Natural Science Foundation of China (82572036 and 82241215). The funding enabled the data collection, analysis, interpretation, and personnel support.
Author Contributions
X. Lin was responsible for the study design and conceptualization, statistical analysis and interpretation, and manuscript drafting and editing. S. Zhang was responsible for data curation and analysis. M. Ong was responsible for data conceptualization and writing – review and editing. J. Fang and A. Mu were responsible for data curation and validation. K. Wang, L. Xu, X. Bi, X. Liu, and Y. Chen were responsible for investigation, project administration, data collection, and data curation. H. Sun was responsible for writing – review and editing and supervision. Q. Ling was responsible for conceptualization, funding acquisition, supervision, and final manuscript editing. All authors read and approved the final manuscript.
Funding Statement
The study was supported by the National Natural Science Foundation of China (82572036 and 82241215). The funding enabled the data collection, analysis, interpretation, and personnel support.
Data Availability Statement
The datasets analyzed during the current study are not publicly available due to the datasets containing information that could compromise the privacy of research participants but are available from the corresponding authors on reasonable request.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
References
- 1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–49. [DOI] [PubMed] [Google Scholar]
- 2. Park J-W, Chen M, Colombo M, Roberts LR, Schwartz M, Chen P-J, et al. Global patterns of hepatocellular carcinoma management from diagnosis to death: the BRIDGE study. Liver Int. 2015;35(9):2155–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. 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]
- 4. Kulik L, El-Serag HB. Epidemiology and management of hepatocellular carcinoma. Gastroenterology. 2019;156(2):477–91.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Lyu N, Yi J-Z, Zhao M. Immunotherapy in older patients with hepatocellular carcinoma. Eur J Cancer. 2022;162:76–98. [DOI] [PubMed] [Google Scholar]
- 6. Galle PR, Forner A, Llovet JM, Mazzaferro V, Piscaglia F, Raoul J-L, et al. EASL clinical practice guidelines: management of hepatocellular carcinoma. J Hepatol. 2018;69(1):182–236. [DOI] [PubMed] [Google Scholar]
- 7. Chinese Society of Clinical Oncology (CSCO) . Clinical Practice Guidelines for Primary Liver Cancer 2022. People’s Medical Publishing House; 2022. [Google Scholar]
- 8. Singal AG, Llovet JM, Yarchoan M, Mehta N, Heimbach JK, Dawson LA, et al. AASLD practice guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatol Baltim Md. 2023;78(6):1922–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Kudo M, Ren Z, Guo Y, Han G, Lin H, Zheng J, et al. Transarterial chemoembolisation combined with lenvatinib plus pembrolizumab versus dual placebo for unresectable, non-metastatic hepatocellular carcinoma (LEAP-012): a multicentre, randomised, double-blind, phase 3 study. Lancet 2025;405(10474):203–15. [DOI] [PubMed] [Google Scholar]
- 10. Jin Z-C, Chen J-J, Zhu X-L, Duan X-H, Xin Y-J, Zhong B-Y, et al. Immune checkpoint inhibitors and anti-vascular endothelial growth factor antibody/tyrosine kinase inhibitors with or without transarterial chemoembolization as first-line treatment for advanced hepatocellular carcinoma (CHANCE2201): a target trial emulation study. eClinicalMedicine. 2024;72:102622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Sangro B, Kudo M, Erinjeri JP, Qin S, Ren Z, Chan SL, et al. Durvalumab with or without bevacizumab with transarterial chemoembolisation in hepatocellular carcinoma (EMERALD-1): a multiregional, randomised, double-blind, placebo-controlled, phase 3 study. Lancet 2025;405(10474):216–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Lian J, Yue Y, Yu W, Zhang Y. Immunosenescence: a key player in cancer development. J Hematol Oncol. 2020;13(1):151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Chen ACY, Jaiswal S, Martinez D, Yerinde C, Ji K, Miranda V, et al. The aged tumor microenvironment limits T cell control of cancer. Nat Immunol. 2024;25(6):1033–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Hajiev S, Allara E, Motedayеn Aval L, Arizumi T, Bettinger D, Pirisi M, et al. Impact of age on sorafenib outcomes in hepatocellular carcinoma: an international cohort study. Br J Cancer. 2021;124(2):407–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Xiao L, Liao Y, Wang J, Li Q, Zhu H, Hong C, et al. Efficacy and safety of immune checkpoint inhibitors in elderly patients with primary liver cancer: a retrospective, multicenter, real-world cohort study. Cancer Immunol Immunother. 2023;72(7):2299–308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Shinkawa H, Tanaka S, Takemura S, Amano R, Kimura K, Nishioka T, et al. Predictive value of the age-adjusted charlson comorbidity index for outcomes after hepatic resection of hepatocellular carcinoma. World J Surg. 2020;44(11):3901–14. [DOI] [PubMed] [Google Scholar]
- 17. Deng K, Xing J, Xu G, Ma R, Jin B, Leng Z, et al. Novel multifactor predictive model for postoperative survival in gallbladder cancer: a multi-center study. World J Surg Oncol. 2024;22(1):263. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. World Health Organization (WHO) . World Health Statistics 2025. Geneva: World Health Organization. 2025. Available from: https://www.who.int/publications/i/item/9789240110496 [Google Scholar]
- 19. Llovet JM, Lencioni R. mRECIST for HCC: performance and novel refinements. J Hepatol. 2020;72(2):288–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–83. [DOI] [PubMed] [Google Scholar]
- 21. Qu W-F, Zhou P-Y, Liu W-R, Tian M-X, Jin L, Jiang X-F, et al. Age-adjusted charlson comorbidity index predicts survival in intrahepatic cholangiocarcinoma patients after curative resection. Ann Transl Med. 2020;8(7):487. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol. 1994;47(11):1245–51. [DOI] [PubMed] [Google Scholar]
- 23. Kobayashi Y, Miura K, Hojo A, Hatta Y, Tanaka T, Kurita D, et al. Charlson comorbidity index is an independent prognostic factor among elderly patients with diffuse large B-cell lymphoma. J Cancer Res Clin Oncol. 2011;137(7):1079–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45(6):613–9. [DOI] [PubMed] [Google Scholar]
- 25. Saleh K, Auperin A, Martin N, Borcoman E, Torossian N, Iacob M, et al. Efficacy and safety of immune checkpoint inhibitors in elderly patients (≥70 years) with squamous cell carcinoma of the head and neck. Eur J Cancer. 2021;157:190–7. [DOI] [PubMed] [Google Scholar]
- 26. Elias R, Giobbie-Hurder A, McCleary NJ, Ott P, Hodi FS, Rahma O. Efficacy of PD-1 & PD-L1 inhibitors in older adults: a meta-analysis. J Immunother Cancer. 2018;6(1):26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Song Y, Gao Q, Zhang H, Fan L, Zhou J, Zou D, et al. Tislelizumab for relapsed/refractory classical hodgkin lymphoma: 3-year follow-up and correlative biomarker analysis. Clin Cancer Res. 2022;28(6):1147–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Keam SJ, Duggan S. Donafenib: first approval. Drugs. 2021;81(16):1915–20. [DOI] [PubMed] [Google Scholar]
- 29. Vandel P. Antidepressant drugs in the elderly--role of the cytochrome P450 2D6. World J Biol Psychiatry. 2003;4(2):74–80. [DOI] [PubMed] [Google Scholar]
- 30. Cheng D, Kochar B, Cai T, Ritchie CS, Ananthakrishnan AN. Comorbidity influences the comparative safety of biologic therapy in older adults with inflammatory bowel diseases. Am J Gastroenterol. 2022;117(11):1845–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Nhean S, Tseng A, Back D. The intersection of drug interactions and adverse reactions in contemporary antiretroviral therapy. Curr Opin HIV AIDS. 2021;16(6):292–302. [DOI] [PubMed] [Google Scholar]
- 32. Jiang Y-H, Zhou M, Cheng M-D, Chen S, Guo Y-Q. CAR-engineered cytolytic tregs reverse pulmonary fibrosis and remodel the fibrotic niche with limited CRS. JCI Insight. 2025;10(15):e182050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Li X, Jing Z, Li S, Wang J, Liu J. Immunomodulatory effects of short-chain fatty acids in chronic obstructive pulmonary disease: rebalancing Th17/treg axis and enhancing epithelial repair. Cytotechnology. 2025;77(6):188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Shi Y, Chien N, Fong A, Nguyen VH, Gudapati ST, Chau A, et al. Differential characteristics and survival outcomes of patients with cirrhosis according to underlying liver aetiology. Aliment Pharmacol Ther. 2025;61(10):1622–34. [DOI] [PubMed] [Google Scholar]
- 35. Su Z, Ding Y, Xue J, Sun J, Ji C. Identifying SUMOylation-related genes in liver fibrosis with bioinformatics and experimental models for diagnostic insights. Sci Rep. 2025;15(1):39783. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Chen X, Gao F, Pan Q, Huang C, Luo R, Lu X, et al. aCCI-HBV-ACLF: a novel predictive model for hepatitis B virus-related acute-on-chronic liver failure. Aliment Pharmacol Ther. 2025;61(2):286–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Demircan NC, Alan Ö, Başoğlu Tüylü T, Akın Telli T, Arıkan R, Çiçek FC, et al. Impact of the charlson comorbidity index on dose-limiting toxicity and survival in locally advanced and metastatic renal cell carcinoma patients treated with first-line sunitinib or pazopanib. J Oncol Pharm Pract. 2020;26(5):1147–55. [DOI] [PubMed] [Google Scholar]
- 38. Govalan R, Luu M, Lauzon M, Kosari K, Ahn JC, Rich NE, et al. Therapeutic underuse and delay in hepatocellular carcinoma: prevalence, associated factors, and clinical impact. Hepatol Commun. 2022;6(1):223–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Singal AG, Kanwal F, Llovet JM. Global trends in hepatocellular carcinoma epidemiology: implications for screening, prevention and therapy. Nat Rev Clin Oncol. 2023;20(12):864–84. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets analyzed during the current study are not publicly available due to the datasets containing information that could compromise the privacy of research participants but are available from the corresponding authors on reasonable request.