Abstract
Background:
Patients with hepatocellular carcinoma (HCC) presenting with both portal vein tumor thrombus (PVTT) and pulmonary metastasis (PM) have a particularly poor prognosis and limited treatment options. Hepatic arterial infusion chemotherapy (HAIC)-based combination therapy has shown promising activity in advanced HCC, but evidence in this high-risk population remains limited.
Objectives:
To evaluate the efficacy and safety of HAIC combined with lenvatinib and tislelizumab (HLP) in patients with HCC with concurrent PVTT and PM.
Design:
Multicenter retrospective cohort study with propensity score matching (PSM).
Methods:
Treatment-naïve patients with HCC with concurrent PVTT and PM who received lenvatinib plus tislelizumab (LP), with or without HAIC, between June 2018 and June 2024 were included. PSM was performed to balance baseline characteristics between treatment groups. The primary endpoint was overall survival (OS). Secondary endpoints included progression-free survival (PFS), objective response rate (ORR), disease control rate (DCR), and adverse events (AEs).
Results:
A total of 262 patients were enrolled, including 114 patients in the HLP group and 148 patients in the LP group. After 1:1 PSM, 96 matched pairs were analyzed. The HLP group achieved significantly longer median OS than the LP group (15.1 vs 7.0 months; hazard ratio (HR), 0.55; 95% confidence interval (CI), 0.40–0.76; p < 0.001) and longer median PFS (7.3 vs 3.7 months; HR, 0.77; 95% CI, 0.56–0.94; p = 0.033). The ORR (47.7% vs 19.8%; p < 0.001) and DCR (86.1% vs 49.0%; p < 0.001) were also significantly higher in the HLP group. Although both all-grade and grade 3/4 treatment-related AEs were more frequent in the HLP group, toxicities were manageable, and no grade 5 AEs were observed.
Conclusion:
In patients with HCC with concurrent PVTT and PM, HLP was associated with improved survival outcomes and tumor response compared with lenvatinib plus tislelizumab alone, with a manageable safety profile.
Keywords: hepatic arterial infusion chemotherapy, hepatocellular carcinoma, lenvatinib, portal vein tumor thrombus, pulmonary metastasis, tislelizumab
Introduction
Hepatocellular carcinoma (HCC) remains one of the leading causes of cancer-related mortality worldwide, and prognosis is particularly poor among patients presenting with advanced disease.1,2 Among these patients, the coexistence of portal vein tumor thrombus (PVTT) and pulmonary metastasis (PM) represents an especially aggressive clinical phenotype characterized by extensive intrahepatic tumor burden, systemic dissemination, and limited therapeutic options. 3 PVTT is observed in approximately 30%–60% of HCC patients in China and is associated with impaired portal venous flow, rapid disease progression, and poor survival outcomes. 4 The presence of PM further worsens prognosis and complicates treatment decision-making because both intrahepatic and extrahepatic disease control are required. 5
Although systemic therapies, including tyrosine kinase inhibitors (TKIs) and immune checkpoint inhibitors (ICIs), have improved outcomes in advanced HCC, treatment responses remain suboptimal in patients with PVTT and extrahepatic metastases. 6 Consequently, locoregional approaches have been explored to enhance therapeutic efficacy. Hepatic arterial infusion chemotherapy (HAIC), particularly FOLFOX-based regimens, enables the delivery of high concentrations of chemotherapeutic agents directly to intrahepatic tumors and has demonstrated favorable local tumor control and survival benefits in patients with advanced HCC, especially those with PVTT.7,8 Furthermore, accumulating evidence suggests that HAIC may synergize with TKIs and ICIs by modulating the tumor microenvironment and enhancing antitumor immune responses. 9
Based on this rationale, triple-combination therapy consisting of HAIC, lenvatinib, and tislelizumab has attracted increasing attention. Several retrospective and early-phase studies have reported encouraging objective response rates (ORRs) and survival outcomes with acceptable safety profiles in patients with unresectable or advanced HCC.10–12 Emerging evidence also suggests potential activity against extrahepatic metastatic lesions, including PMs. 13 However, the available evidence remains limited by retrospective study designs, relatively small sample sizes, and substantial heterogeneity in patient characteristics. Moreover, most published cohorts have predominantly enrolled patients with preserved liver function (Child–Pugh class A), which may restrict the generalizability of the reported outcomes to broader real-world populations. Consequently, robust multicenter data focusing specifically on patients with concurrent PVTT and PM remain scarce.
Therefore, we conducted a multicenter propensity score matching (PSM) analysis to evaluate the efficacy and safety of HAIC combined with lenvatinib and tislelizumab compared with lenvatinib plus tislelizumab (LP) in patients with advanced HCC complicated by both PVTT and PM. By focusing on this particularly high-risk population, we aimed to provide real-world evidence to inform treatment selection and optimize therapeutic strategies.
Methods
Participants
In this retrospective, multicenter study, we reviewed patients with concurrent portal vein tumor thrombosis (PVTT) and PM who received lenvatinib and tislelizumab between June 2018 and June 2024 at three institutions in China: The First Affiliated Hospital of Sun Yat-sen University, Guangdong Provincial People’s Hospital, and the Memorial Hospital of Sun Yat-sen University. Patients were stratified according to treatment regimen into two groups: the HAIC plus lenvatinib and tislelizumab (HLP) group, which received a triple combination of HAIC, lenvatinib, and tislelizumab, and the LP group, which received LP. The distribution of patients across participating centers is summarized in Table S1.
The inclusion criteria were as follows: (1) age 18–75 years; (2) a diagnosis of HCC established by contrast-enhanced imaging in high-risk individuals or by histopathology, in accordance with the American Association for the Study of Liver Diseases guidelines; (3) presence of PVTT and PM confirmed by imaging; (4) an Eastern Cooperative Oncology Group performance status (ECOG PS) score ⩽ 2 and Child-Pugh class A or B liver function; (5) no prior treatment for HCC; (6) no history of other malignancies within the preceding 5 years; (7) receipt of at least six cycles of tislelizumab and a minimum of 2 months of lenvatinib in both groups, and at least two cycles of HAIC in the HLP group; and (8) availability of complete clinical and follow-up data. Key exclusion criteria included: (1) histologic subtypes of fibrolamellar HCC, sarcomatoid HCC, or combined hepatocellular-cholangiocarcinoma (HCC-CC); (2) evidence of hepatic decompensation, including hepatic encephalopathy or variceal gastrointestinal bleeding; and (3) discontinuation or modification of treatment without a documented medical indication.
Baseline laboratory tests and imaging assessments, including contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI), were performed within 1 week before treatment initiation.
Treatment procedures
HAIC was performed by experienced interventional radiologists at each participating center in accordance with established protocols using the FOLFOX regimen. The catheter tip was superselectively advanced into the tumor-feeding branch of the hepatic artery based on tumor size, location, and vascular anatomy. The chemotherapy regimen consisted of oxaliplatin (130 mg/m2 administered over 2 h on day 1), leucovorin (200 mg/m2 administered from hour 2 to 4 on day 1), and fluorouracil (400 mg/m2 as a 15-min bolus, followed by 2400 mg/m2 as a continuous infusion over 46 h on days 1 and 2). Treatment was repeated every 3 weeks for up to 8 cycles, at the discretion of the treating physician.
In both groups, patients received oral lenvatinib at a dose of 8 mg daily (body weight ⩽60 kg) or 12 mg daily (body weight >60 kg), in combination with an intravenous tislelizumab (200 mg every 3 weeks). In the HLP group, lenvatinib was initiated on day 1 of HAIC, and tislelizumab was administered on day 2 immediately after completion of HAIC. In the LP group, lenvatinib and tislelizumab were administered on the same day.
In the event of treatment-related adverse events (AEs), the dose of lenvatinib or tislelizumab was reduced or temporarily withheld until resolution or improvement of symptoms.
Efficacy and safety assessment
Chest CT and abdominal contrast-enhanced CT or MRI were performed every two treatment cycles. Imaging assessments were independently reviewed by two radiologists with expertise in liver diseases. In cases of discrepancy, a senior radiologist adjudicated the findings, and consensus was reached.
To ensure consistency across participating institutions, tumor response was evaluated according to the Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST 1.1). The primary end point was overall survival (OS), defined as the time from initial admission to death from any cause. Secondary end points included progression-free survival (PFS), ORR, and disease control rate (DCR). PFS was defined as the time from admission to radiologic progressive disease or death, whichever occurred first. ORR was defined as the proportion of patients achieving a complete response (CR) or partial response (PR), and DCR as the proportion achieving CR, PR, or stable disease (SD). AEs occurring during treatment were recorded and graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE), version 5.0. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement. 14 A completed STROBE checklist is provided in the Supplemental File.
PSM analysis
PSM was performed to reduce selection bias and balance baseline characteristics between treatment groups. Propensity scores were estimated using logistic regression based on variables associated with treatment allocation, including age, sex, ECOG performance status, etiology, number of intrahepatic lesions, tumor size, number of PMs, PVTT type, alpha-fetoprotein level, albumin–bilirubin grade, and Child-Pugh class. Patients were matched in a 1:1 ratio using a nearest-neighbor algorithm with a caliper width of 0.2.
Statistical analysis
Statistical analyses were performed using R software (version 4.0.3; R Foundation for Statistical Computing, Vienna, Austria) and SPSS (version 27.0; IBM Corp., Armonk, NY, USA). The normality of continuous variables was assessed using the Shapiro–Wilk test. Continuous variables are presented as mean ± standard deviation or median (interquartile range), as appropriate, and were compared using the Student’s t test or the Mann–Whitney U test. Categorical variables are presented as frequencies and percentages and were compared using the χ2 test or Fisher’s exact test, as appropriate.
OS and PFS were estimated using the Kaplan–Meier method and compared between groups with the log-rank test. Univariable and multivariable Cox proportional hazards regression analyses were performed to identify factors associated with survival outcomes. Variables with a p-value < 0.10 in univariable analysis were included in the multivariable model. All tests were two-sided, and a p-value < 0.05 was considered statistically significant.
Results
Participants
A total of 262 patients with concomitant PVTT and PM were included (Figure 1). Of these, 114 received HAIC combined with lenvatinib and tislelizumab (HLP group), and 148 received LP (LP group). At the data cutoff (March 31, 2026), the median follow-up was 26.8 months (95% confidence interval (CI), 17.4–53.2) in the HLP group and 24.3 months (95% CI, 14.4–42.1) in the LP group. PSM (1:1) yielded 96 matched pairs. Baseline characteristics were well balanced after matching (Table 1). Before matching, patients in the HLP group received a median of 4 HAIC cycles (range, 2–8), with a median lenvatinib exposure of 6.5 months (range, 3.1–13.3) and 12 cycles of tislelizumab (range, 4–22). In the LP group, the median duration of lenvatinib treatment was 5.4 months (range, 2.3–12.5), and the median number of tislelizumab cycles was 10 (range, 4–18).
Figure 1.
Patient selection flowchart. A patient might meet several exclusion criteria, but they were excluded only once from the uppermost criteria.
BCLC, Barcelona Clinic Liver Cancer; ECOG, Eastern Cooperative Oncology Group; HAIC, hepatic arterial infusion chemotherapy; HLP, HAIC combined with lenvatinib and tislelizumab; LP, lenvatinib combined with tislelizumab; PS, performance score; PSM, propensity score matching.
Table 1.
Baseline characteristics of the study patients before and after PSM.
| Characteristics | Before matching | After matching | ||||
|---|---|---|---|---|---|---|
| LP, n = 148 | HLP, n = 114 | p | LP, n = 96 | HLP, n = 96 | p | |
| Age (mean ± SD, years) | 48 ± 12 | 49 ± 12 | 0.578 | 50 ± 11 | 49 ± 12 | 0.551 |
| Age (years) | 0.590 | 0.553 | ||||
| ⩽46 | 62 (42%) | 44 (39%) | 35 (36%) | 39 (41%) | ||
| >46 | 86 (58%) | 70 (61%) | 61 (64%) | 57 (59%) | ||
| Sex | 0.232 | 0.488 | ||||
| Male | 128 (86%) | 104 (91%) | 84 (88%) | 87 (91%) | ||
| Female | 20 (14%) | 10 (9%) | 12 (13%) | 9 (9%) | ||
| ECOG-PS | 0.003 | >0.999 | ||||
| 0 | 132 (89%) | 88 (77%) | 82 (85%) | 82 (85%) | ||
| 1 | 14 (9%) | 26 (23%) | 14 (15%) | 14 (15%) | ||
| 2 | 2 (1%) | 0 (0) | 0 (0) | 0 (0) | ||
| Etiology | 0.429 | 0.551 | ||||
| HBV | 141 (95%) | 106 (93%) | 89 (93%) | 91 (95%) | ||
| Others | 7 (5%) | 8 (7%) | 7 (7%) | 5 (5%) | ||
| Intrahepatic lesion number | 0.087 | 0.873 | ||||
| Single | 34 (23%) | 37(32%) | 27 (28%) | 28 (29%) | ||
| Multiple | 114 (77%) | 77 (68%) | 69 (72%) | 68 (71%) | ||
| Size (mean ± SD, cm) | 12.4 ± 3.5 | 12.5 ± 3.4 | 0.779 | 12.1 ± 3.2 | 12.5 ± 3.6 | 0.466 |
| Size (cm) | 0.209 | 0.414 | ||||
| ⩽10 | 44 (30%) | 26 (23%) | 28 (29%) | 23 (24%) | ||
| >10 | 104 (70%) | 88 (77%) | 68 (71%) | 73 (76%) | ||
| PVTT | 0.622 | 0.463 | ||||
| Vp1/2 | 85 (57%) | 62 (54%) | 59 (61%) | 54 (56%) | ||
| Vp3/4 | 63 (43%) | 52 (46%) | 37 (39%) | 42 (44%) | ||
| AFP (μg/L) | 0.931 | 0.636 | ||||
| ⩽400 | 46 (31%) | 36 (32%) | 30 (31%) | 27 (28%) | ||
| >400 | 102 (69%) | 78 (68%) | 66 (69%) | 69 (72%) | ||
| ALBI | 0.175 | 0.644 | ||||
| 1 | 49 (33%) | 36 (32%) | 29 (30%) | 33 (34%) | ||
| 2 | 99 (67%) | 75 (66%) | 67 (70%) | 63 (66%) | ||
| 3 | 0 (0) | 3 (3%) | 0 (0) | 0 (0) | ||
| Child-Pugh | 0.580 | 0.520 | ||||
| A | 127 (86%) | 95 (83%) | 82 (85%) | 85 (89%) | ||
| B | 21 (14%) | 19 (17%) | 14 (15%) | 11 (11%) | ||
Continuous variables are presented as mean ± SD, and categorical variables are presented as n (%). p Values were calculated using a two-sided Welch t-test and Pearson’s χ2 test.
AFP, alpha-fetoprotein; ALBI, albumin-bilirubin; ECOG, Eastern Cooperative Oncology Group; HAIC, hepatic arterial infusion chemotherapy; HLP, HAIC combined with lenvatinib and tislelizumab; LP, lenvatinib combined with tislelizumab; PS, performance score; PSM, propensity score matching; PVTT, portal vein tumor thrombus.
Tumor response
Tumor responses were evaluated according to RECIST version 1.1 before and after PSM (Table 2). Before matching, the HLP group achieved significantly higher ORR (43.0% vs 23.7%; p < 0.001) and DCR (79.8% vs 56.8%; p < 0.001) than the LP group. These differences remained significant after matching, with higher ORR (47.7% vs 19.8%; p < 0.001) and DCR (86.1% vs 49.0%; p < 0.001) observed in the HLP group.
Table 2.
Overall efficacy evaluated by RECIST 1.1 criteria before and after PSM.
| Response | Before PSM | After PSM | ||||
|---|---|---|---|---|---|---|
| LP, n = 148 | HLP, n = 114 | p | LP, n = 96 | HLP, n = 96 | p | |
| CR | 0 (0) | 0 (0) | 0 (0) | 0 (0) | ||
| PR | 35 (23.7%) | 49 (43.0%) | 19 (19.8%) | 43 (47.7%) | ||
| SD | 49 (33.1%) | 42 (36.8%) | 28 (29.2%) | 35 (38.4%) | ||
| PD | 64 (43.2%) | 23 (20.2%) | 49 (51.0%) | 18 (13.9%) | ||
| ORR | 35 (23.7%) | 49 (43.0%) | <0.001 | 19 (19.8%) | 43 (47.7%) | <0.001 |
| DCR | 84 (56.8%) | 91 (79.8%) | <0.001 | 47 (49.0%) | 78 (86.1%) | <0.001 |
Summary of best response. Values are presented as n (%). p Values were calculated using a two-sided χ2 test.
CR, complete response; DCR, disease control rate; HAIC, hepatic arterial infusion chemotherapy; HLP, HAIC combined with lenvatinib and tislelizumab; LP, lenvatinib combined with tislelizumab; ORR, objective response rate; PD, progressive disease; PR, partial response; PSM, propensity score matching; RECIST, response evaluation criteria in solid tumors; SD, stable disease.
Subgroup analyses of intrahepatic lesions and pulmonary lesions showed consistent results. For intrahepatic tumors, the HLP group demonstrated significantly higher ORR and DCR both before matching (ORR, 50.9% vs 26.3%; p < 0.001; DCR, 83.3% vs 65.5%; p < 0.001) and after matching (ORR, 53.1% vs 25.0%; p < 0.001; DCR, 87.5% vs 64.6%; p < 0.001). Similarly, for PMs, the HLP group showed improved response rates compared with the LP group before matching (ORR, 30.7% vs 18.3%; p = 0.014; DCR, 81.6% vs 69.1%; p = 0.002) and after matching (ORR, 28.1% vs 12.5%; p = 0.006; DCR, 83.3% vs 64.6%; p = 0.005; Tables S2 and S3).
Survival outcomes
Before PSM, 130 patients (87.8%) in the LP group and 92 patients (80.7%) in the HLP group had died at the time of data cutoff. The HLP group demonstrated significantly longer OS compared with the LP group, with a median OS of 13.0 months (95% CI, 11.2–16.8) versus 7.0 months (95% CI, 5.9–8.6; HR, 0.56; 95% CI, 0.43–0.73; p < 0.001). PFS was also significantly improved in the HLP group, with a median PFS of 7.1 months (95% CI, 6.0–8.0) compared with 3.7 months (95% CI, 3.1–5.5) in the LP group (HR, 0.73; 95% CI, 0.56–0.94; p = 0.015). After matching (96 pairs), the survival benefit associated with HLP remained consistent. The median OS was 15.1 months (95% CI, 11.6–18.6) in the HLP group and 7.0 months (95% CI, 5.0–8.6) in the LP group (HR, 0.55; 95% CI, 0.40–0.76; p < 0.001). Median PFS likewise favored the HLP group (7.3 months (95% CI, 6.1–9.0) vs 3.7 months (95% CI, 2.9–5.8); HR, 0.76; 95% CI, 0.56–0.94; p = 0.033; Figure 2).
Figure 2.
Kaplan–Meier curves comparing OS and PFS among patients who underwent HAIC plus lenvatinib and tislelizumab (HLP) or LP before (a–b) and after(c–d) PSM. p Values were calculated using the Log-rank test.
CI, confidence interval; HAIC, hepatic arterial infusion chemotherapy; HLP, HAIC combined with lenvatinib and tislelizumab; HR, hazard ratio; LP, lenvatinib combined with tislelizumab; OS, overall survival; PFS, progression-free survival; PSM, propensity score matching.
Univariate and multivariate analysis
Univariate and multivariate Cox regression analyses were performed to identify factors associated with OS and PFS (Table 3). In the multivariable model, treatment option and PVTT type remained significantly associated with OS. For PFS, treatment option and PVTT type were also significantly associated with outcome; however, the association between treatment option and PFS should be interpreted with appropriate caution, as the upper bound of the CI approached unity.
Table 3.
Univariate and multivariate analyses of predictors of survival after treatment.
| Characteristics | OS | PFS | ||||||
|---|---|---|---|---|---|---|---|---|
| Univariate analysis | Multivariate analysis | Univariate analysis | Multivariate analysis | |||||
| HR (95% CI) | p | HR (95% CI) | p | HR (95% CI) | p | HR (95% CI) | p | |
| Age (>46) | 0.82 (0.60, 1.13) | 0.220 | 0.76 (0.56, 1.04) | 0.088 | 0.63 (0.45, 0.88) | 0.007 | ||
| Sex (female) | 1.19 (0.74, 1.90) | 0.477 | 1.12 (0.71, 1.77) | 0.625 | ||||
| ECOG-PS (1) | 1.24 (0.81, 1.90) | 0.328 | 1.19 (0.78, 1.83) | 0.420 | ||||
| Etiology (HBV) | 0.67 (0.31, 1.46) | 0.316 | 0.62 (0.32, 1.18) | 0.144 | ||||
| Intrahepatic lesion number (multiple) | 1.08 (0.77, 1.52) | 0.669 | 1.10 (0.79, 1.54) | 0.572 | ||||
| Size (>10 cm) | 1.04 (0.74, 1.52) | 0.669 | 1.12 (0.79, 1.57) | 0.527 | ||||
| PVTT (Vp3/4) | 2.74 (1.93, 3.89) | <0.001 | 2.89 (2.04, 4.10) | <0.001 | 1.71 (1.24, 2.36) | 0.001 | 1.84 (1.32, 2.56) | <0.001 |
| AFP (>400) | 0.96 (0.67, 1.37) | 0.822 | 1.34 (0.94, 1.92) | 0.109 | ||||
| ALBI (2) | 1.29 (0.92, 1.81) | 0.139 | 1.01 (0.73, 1.40) | 0.961 | ||||
| Child-Pugh (B) | 1.39 (0.85, 2.28) | 0.194 | 1.38 (0.85, 2.24) | 0.194 | ||||
| Treatment (HLP) | 0.55 (0.40, 0.75) | <0.001 | 0.51 (0.37, 0.70) | <0.001 | 0.76 (0.56, 1.03) | 0.074 | 0.63 (0.45, 0.98) | 0.005 |
Univariable and multivariable Cox regression analyses were performed to identify the factors associated with survival. Factors with p < 0.1 in univariate analysis were included in multivariate analysis. Two-sided p < 0.05 was defined as statistically significant.
AFP, alpha-fetoprotein; ALBI, albumin-bilirubin; ECOG, Eastern Cooperative Oncology Group; HAIC, hepatic arterial infusion chemotherapy; HLP, HAIC combined with lenvatinib and tislelizumab; LP, lenvatinib combined with tislelizumab; OS, overall survival; PFS, progression-free survival; PS, performance score; PVTT, portal vein tumor thrombus.
Subgroup analysis
Forest plots were constructed to evaluate treatment effects across predefined subgroups after PSM (Figure 3). For both OS and PFS, the HLP group consistently demonstrated greater benefit than the LP group across nearly all subgroups, suggesting that HAIC combined with lenvatinib and tislelizumab provides clinical benefit in diverse populations of patients with HCC with PVTT and PM.
Figure 3.
Forest plots based on OS (a) and PFS (b) of each subgroup. p Values were calculated using the log-rank test.
AFP, alpha-fetoprotein; ALBI, albumin-bilirubin; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HAIC, hepatic arterial infusion chemotherapy; HLP, HAIC combined with lenvatinib and tislelizumab; HR, hazard ratio; LP, lenvatinib combined with tislelizumab; OS, overall survival; PFS, progression-free survival; PS, performance score; PSM, propensity score matching; PVTT, portal vein tumor thrombus.
Patterns of progressive disease
Five patterns of progressive disease were identified: intrahepatic lesion progression (ILP), new intrahepatic lesions (NIL), extrahepatic lesion progression (ELP), new extrahepatic lesions (NEL), and death. At data cutoff, progressive disease had occurred in 99 patients in the HLP group and 132 patients in the LP group; some patients exhibited more than one progressive pattern. The distribution of progressive patterns in the HLP versus LP groups was as follows: ILP (15.1% vs 39.7%), NIL (10.9% vs 25.0%), ELP (27.7% vs 14.0%), NEL (39.6% vs 19.5%), and death (6.7% vs 1.8%). The proportions of ILP and NIL were notably lower in the HLP group than in the LP group (Figure S1).
Safety
As shown in Table 4, the overall incidence of AEs was higher in the HLP group than in the LP group (71.9% vs 58.1%). In the HLP group, the most common AEs were decreased appetite (74.4%), abdominal pain (56.1%), and nausea (51.8%). The most frequent grade 3–4 AEs were nausea (13.2%), abdominal pain (11.4%), and diarrhea (9.6%), most of which were attributed to HAIC. In the LP group, decreased appetite (31.1%), increased AST (28.4%), and hypertension (25.0%) were the most common AEs, whereas grade 3–4 AEs most frequently included hypertension (7.4%), proteinuria (7.4%), and fatigue (4.7%). Although both any-grade and grade 3–4 AEs were more frequent in the HLP group, these events were generally manageable, and no treatment-related deaths were observed.
Table 4.
Treatment-related AEs.
| AEs | Any grade | Grade 3/4 | ||||
|---|---|---|---|---|---|---|
| LP (n = 148) | HLP (n = 114) | p | LP (n = 148) | HLP (n = 114) | p | |
| Overall incidence | 86 (58.1%) | 82 (71.9%) | 0.027 | 19 (12.8%) | 22 (19.3%) | 0.172 |
| Abdominal pain | 6 (4.1%) | 64 (56.1%) | <0.001 | 1 (0.7%) | 13 (11.4%) | <0.001 |
| Nausea | 18 (12.2%) | 59 (51.8%) | <0.001 | 3 (2.0%) | 15 (13.2%) | <0.001 |
| Diarrhea | 27 (18.2%) | 49 (43.0%) | <0.001 | 5 (3.4%) | 11 (9.6%) | 0.041 |
| Decreased appetite | 46 (31.1%) | 61 (74.4%) | <0.001 | 3 (2.0%) | 5 (4.4%) | 0.300 |
| Rash | 17 (11.5%) | 12 (10.5%) | 0.845 | 3 (2.0%) | 2 (1.8%) | 0.998 |
| Fatigue | 22 (14.9%) | 27 (23.7%) | 0.080 | 7 (4.7%) | 5 (4.4%) | 0.999 |
| Hypoproteinemia | 31 (20.9%) | 25 (21.9%) | 0.880 | 4 (2.7%) | 3 (1.8%) | 0.999 |
| Elevated bilirubin | 15 (10.1%) | 14 (12.3%) | 0.692 | 5 (3.4%) | 3 (2.6%) | 0.999 |
| Elevated ALT | 31 (20.9%) | 26 (16.6%) | 0.764 | 2 (1.4%) | 1 (0.9%) | 0.999 |
| Elevated AST | 42 (28.4%) | 29 (22.8%) | 0.674 | 1 (0.7%) | 1 (0.9%) | 0.999 |
| Hypothyroidism | 8 (5.4%) | 5 (4.4%) | 0.781 | 2 (1.4%) | 2 (1.8%) | 0.999 |
| Leukopenia | 7 (4.7%) | 19 (16.7%) | 0.002 | 1 (0.7%) | 3 (2.6%) | 0.320 |
| Thrombocytopenia | 24 (16.2%) | 21 (18.4%) | 0.741 | 3 (2.0%) | 4 (3.5%) | 0.473 |
| Hypertension | 37 (25.0%) | 29 (25.4%) | 0.999 | 11 (7.4%) | 9 (7.9%) | 0.999 |
| Hand-foot skin reaction | 14 (9.5%) | 11 (9.6%) | 0.999 | 5 (3.4%) | 4 (3.5%) | 0.999 |
| Proteinuria | 31 (20.9%) | 23 (20.2%) | 0.999 | 11 (7.4%) | 7 (6.1%) | 0.807 |
| Immunity-related AEs | 16 (10.8%) | 11 (9.6%) | 0.839 | 6 (4.1%) | 4 (3.5%) | 0.999 |
Values are presented as n (%). p Values were calculated using a two-sided χ2 test.
AEs, adverse events; ALT, alanine transaminase; AST, aspartate aminotransferase; HAIC, hepatic arterial infusion chemotherapy; HLP, HAIC combined with lenvatinib and tislelizumab; LP, lenvatinib combined with tislelizumab.
Discussion
In this multicenter propensity score-matched analysis, we focused on a particularly high-risk and understudied population of patients with HCC presenting concurrently with PVTT and PM. To our knowledge, this is the first study to specifically evaluate therapeutic strategies in this distinct subgroup. Our findings demonstrate that the addition of HAIC to LP is associated with significantly improved survival outcomes and tumor response compared with LP alone. Notably, the triple-modality regimen resulted in prolonged OS, improved PFS, and higher objective response and DCRs, with a manageable safety profile.
These results are consistent with and extend prior evidence supporting the synergistic effects of combining HAIC with systemic therapies in advanced HCC. Recent real-world studies in high-risk HCC populations have reported substantial survival benefits with HAIC-based triple therapy, with median OS approaching 19 months compared with approximately 10 months for systemic therapy alone. 15 Similarly, meta-analyses have shown that adding HAIC to lenvatinib and PD-1 inhibitors significantly improves OS, PFS, and tumor response rates. 16 However, most existing studies have included heterogeneous populations of advanced HCC, without specifically addressing patients with both macrovascular invasion and extrahepatic spread. By isolating patients with concomitant PVTT and PM, our study fills an important clinical evidence gap and provides more precise insight into treatment effects in this particularly poor-prognosis subgroup.
Importantly, our findings suggest that HAIC not only enhances intrahepatic tumor control but may also contribute to improved control of PMs when combined with systemic therapy. In our analysis, the addition of HAIC was associated with a marked reduction in intrahepatic progression patterns, while maintaining favorable control of extrahepatic disease. This observation supports the hypothesis that effective intrahepatic tumor debulking may reduce systemic tumor dissemination and enhance the overall efficacy of systemic agents, thereby indirectly contributing to improved control of metastatic lesions.17,18
From a mechanistic perspective, the observed benefits are biologically plausible. HAIC delivers high concentrations of cytotoxic agents directly to the liver, leading to rapid tumor shrinkage and increased tumor antigen release, which may enhance antitumor immune responses.19,20 In parallel, lenvatinib exerts anti-angiogenic effects and promotes vascular normalization, facilitating immune cell infiltration into the tumor microenvironment.21,22 These effects may synergize with PD-1 blockade to potentiate systemic antitumor immunity, thereby contributing to both intrahepatic and extrahepatic disease control. Such multimodal interactions highlight the rationale for integrating locoregional and systemic therapies in advanced HCC.23,24
Clinically, our findings support a treatment paradigm shift toward multimodal strategies for patients with high-risk HCC. In particular, for patients with preserved liver function and good performance status, HAIC combined with lenvatinib and PD-1 inhibition may represent a promising therapeutic option. Given the historically poor prognosis of patients with concurrent PVTT and PM, the magnitude of benefit observed in our study is clinically meaningful and may help inform treatment decision-making in real-world practice.
Several limitations should be acknowledged. First, the retrospective design introduces the potential for residual selection bias and unmeasured confounding despite the use of PSM. Besides, though standardized RECIST 1.1 criteria were used across all centers, the absence of central independent radiologic review may have introduced inter-institutional variability in response assessment. Therefore, causal relationships between treatment exposure and clinical outcomes should be interpreted with caution. Second, the follow-up duration was relatively limited, precluding comprehensive assessment of long-term survival outcomes and late treatment-related toxicities. Third, the study population predominantly consisted of patients with Child–Pugh A liver function. Consequently, the observed efficacy and safety outcomes may not be readily generalizable to patients with more advanced hepatic impairment, who are frequently encountered in routine clinical practice and may exhibit different treatment tolerability and outcomes. Prospective studies involving more diverse patient populations are warranted to validate our findings.
In addition, although the HLP regimen demonstrated superior tumor response and survival outcomes, it was associated with a higher incidence of treatment-related AEs. The absence of patient-reported quality-of-life assessments and health-economic evaluations therefore represents an important limitation of the present study. Without these data, the overall clinical value of HLP therapy cannot be fully characterized, as treatment benefit must be balanced against toxicity burden, patient experience, and healthcare resource utilization. Future prospective studies incorporating quality-of-life measures and cost-effectiveness analyses are needed to determine the net clinical benefit of this treatment strategy.
Conclusion
In this multicenter propensity score-matched analysis, HAIC combined with lenvatinib and tislelizumab was associated with improved OS, PFS, and tumor response compared with LP alone in patients with HCC complicated by both PVTT and PM. Although treatment-related AEs were more frequent with triple therapy, the safety profile remained manageable. These findings support the potential role of HAIC-based triple therapy in this particularly high-risk population and warrant further validation in prospective randomized studies.
Supplemental Material
Supplemental material, sj-docx-1-tam-10.1177_17588359261472766 for HAIC combined with lenvatinib and tislelizumab for HCC with concurrent PVTT and pulmonary metastasis: a multicenter propensity score matching analysis by Song Chen, Zhiqiang Wu, Yi Xiao, Xiongying Jiang, Feng Shi, Ning Wang, Fangzeng Lin, Guosheng Tan, Wenquan Zhuang and Yonghui Huang in Therapeutic Advances in Medical Oncology
Supplemental material, sj-docx-2-tam-10.1177_17588359261472766 for HAIC combined with lenvatinib and tislelizumab for HCC with concurrent PVTT and pulmonary metastasis: a multicenter propensity score matching analysis by Song Chen, Zhiqiang Wu, Yi Xiao, Xiongying Jiang, Feng Shi, Ning Wang, Fangzeng Lin, Guosheng Tan, Wenquan Zhuang and Yonghui Huang in Therapeutic Advances in Medical Oncology
Acknowledgments
None.
Footnotes
ORCID iD: Song Chen 
https://orcid.org/0000-0002-7065-6609
Supplemental material: Supplemental material for this article is available online.
Contributor Information
Song Chen, Department of Interventional Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China.
Zhiqiang Wu, Department of Interventional Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China.
Yi Xiao, Center of Hepato-Pancreato-Biliary Surgery, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China.
Xiongying Jiang, Interventional Radiology Division, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China.
Feng Shi, Department of Interventional Radiology, Guangdong Provincial People’s Hospital, Guangzhou, China.
Ning Wang, Department of Radiology, The Sixth Affiliated Hospital of Sun Yat-sen University, Guangzhou, China.
Fangzeng Lin, Department of Interventional Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China.
Guosheng Tan, Department of Interventional Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China.
Wenquan Zhuang, Department of Interventional Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou 510080, Guangdong, China.
Yonghui Huang, Department of Interventional Radiology, The First Affiliated Hospital of Sun Yat-sen University, Guangzhou 510080, Guangdong, China.
Declarations
Ethics approval and consent to participate: This multicenter retrospective study was approved by the Ethics Committee of The First Affiliated Hospital of Sun Yat-sen University (Approval Number: 2021-782). Separate ethics approvals were not obtained from the other participating centers, as the study was coordinated through the lead institution and involved the collection and analysis of de-identified retrospective clinical data only. The requirement for informed consent was waived by the Ethics Committee due to the retrospective nature of the study.
Consent for publication: Not applicable.
Author contributions: Song Chen: Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Writing – original draft.
Zhiqiang Wu: Conceptualization; Formal analysis; Methodology; Resources; Writing – review & editing.
Yi Xiao: Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Writing – review & editing.
Xiongying Jiang: Investigation; Methodology; Project administration; Validation; Writing – review & editing.
Feng Shi: Conceptualization; Project administration; Validation.
Ning Wang: Investigation; Project administration; Validation.
Fangzeng Lin: Formal analysis; Project administration; Validation.
Guosheng Tan: Conceptualization; Investigation; Methodology.
Wenquan Zhuang: Project administration; Resources; Supervision; Validation; Writing – review & editing.
Yonghui Huang: Project administration; Resources; Supervision; Validation; Writing – review & editing.
Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Postdoctoral Science Foundation of China (Song Chen; General Funding: No. 2025M782389).
The authors declare that there is no conflict of interest.
Availability of data and materials: All inquiries can be directed to the corresponding authors.*
References
- 1. Bray F, Laversanne M, Sung H, 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: 229–263. [DOI] [PubMed] [Google Scholar]
- 2. Zhou J, Sun H, Wang Z, et al. China liver cancer guidelines for the diagnosis and treatment of hepatocellular carcinoma (2024 edition). Liver Cancer 2025; 14: 779–835. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Xie D, Shi J, Zhou J, et al. Clinical practice guidelines and real-life practice in hepatocellular carcinoma: a Chinese perspective. Clin Mol Hepatol 2023; 29: 206–216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Xie DY, Zhu K, Ren ZG, et al. A review of 2022 Chinese clinical guidelines on the management of hepatocellular carcinoma: updates and insights. Hepatobiliary Surg Nutr 2023; 12: 216–228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Xia F, Chen Q, Li C, et al. Refining prognosis and treatment strategies beyond the Barcelona Clinic Liver Cancer stage in hepatocellular carcinoma with lung metastases: a multicenter cohort study. MedComm (2020) 2025; 6: e70306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Cappuyns S, Corbett V, Yarchoan M, et al. Critical appraisal of guideline recommendations on systemic therapies for advanced hepatocellular carcinoma: a review. JAMA Oncol 2024; 10: 395–404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Iwamoto H, Shimose S, Shirono T, et al. Hepatic arterial infusion chemotherapy for advanced hepatocellular carcinoma in the era of chemo-diversity. Clin Mol Hepatol 2023; 29: 593–604. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Patel KR, Menon H, Patel RR, et al. Locoregional therapies for hepatocellular carcinoma: a systematic review and meta-analysis. JAMA Netw Open 2024; 7: e2447995. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Xing R, Mei J, Zuo Z, et al. Enhanced formation of tertiary lymphoid structures shapes the anti-tumor microenvironment in hepatocellular carcinoma after FOLFOX-HAIC therapy. Cell Rep Med 2025; 6: 102298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Tang S, Shi F, Xiao Y, et al. HAIC plus lenvatinib and tislelizumab for advanced hepatocellular carcinoma with Vp4 portal vein invasion. Hepatol Int 2025; 19: 106–117. [DOI] [PubMed] [Google Scholar]
- 11. Cai H, Chen S, Tang S, et al. Lenvatinib and tislelizumab versus atezolizumab and bevacizumab in combination with TAE-HAIC for unresectable hepatocellular carcinoma with high tumor burden: a multicenter retrospective cohort study. Cancer Immunol Immunother 2025; 74: 88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Zhao Z, Jiang X, Wen S, et al. Efficiency and safety of HAIC combined with lenvatinib and tislelizumab for advanced hepatocellular carcinoma with high tumor burden: a multicenter propensity score matching analysis. Front Pharmacol 2024; 15: 1499269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Chen S, Wang N, Xiao Y, et al. Efficiency and safety of HAIC combined with lenvatinib and PD-1 inhibitor for advanced hepatocellular carcinoma with lung metastasis: a multicenter propensity score matching analysis. Immunotargets Ther 2025; 14: 953–965. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007; 370: 1453–1457. [DOI] [PubMed] [Google Scholar]
- 15. Zhong S, Yi J, Chen S, et al. Combining immune checkpoint inhibitors and molecular-targeted agents with hepatic arterial infusion chemotherapy for hepatocellular carcinoma with inferior vena cava and/or right atrium tumor thrombus. Hepatol Int 2025; 19: 560–575. [DOI] [PubMed] [Google Scholar]
- 16. Lu D, Li H, Sun P, et al. Systemic therapy plus HAIC versus systemic therapy for hepatocellular carcinoma: a systematic review and meta-analysis. Int J Surg 2025; 111: 3494–3507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Huang Y, Du Z, Lai Z, et al. Single-nucleus and spatial transcriptome profiling delineates the multicellular ecosystem in hepatocellular carcinoma after hepatic arterial infusion chemotherapy. Adv Sci (Weinh) 2025; 12: e2405749. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Zhao M, Guo Z, Zou YH, et al. Arterial chemotherapy for hepatocellular carcinoma in China: consensus recommendations. Hepatol Int 2024; 18: 4–31. [DOI] [PubMed] [Google Scholar]
- 19. Singal AG, Kanwal F, Llovet JM. Global trends in hepatocellular carcinoma epidemiology: implications for screening, prevention and therapy. Nat Rev Clin Oncol 2023; 20: 864–884. [DOI] [PubMed] [Google Scholar]
- 20. Moris D, Martinino A, Schiltz S, et al. Advances in the treatment of hepatocellular carcinoma: an overview of the current and evolving therapeutic landscape for clinicians. CA Cancer J Clin 2025; 75: 498–527. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Chen Y, Dai S, Cheng CS, et al. Lenvatinib and immune-checkpoint inhibitors in hepatocellular carcinoma: mechanistic insights, clinical efficacy, and future perspectives. J Hematol Oncol 2024; 17: 130. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Torrens L, Montironi C, Puigvehí M, et al. Immunomodulatory effects of lenvatinib plus anti-programmed cell death protein 1 in mice and rationale for patient enrichment in hepatocellular carcinoma. Hepatology 2021; 74: 2652–2669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Yang C, Zhang H, Zhang L, et al. Evolving therapeutic landscape of advanced hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol 2023; 20: 203–222. [DOI] [PubMed] [Google Scholar]
- 24. Greten TF, Wang XW, Korangy F. Current concepts of immune based treatments for patients with HCC: from basic science to novel treatment approaches. Gut 2015; 64: 842–848. [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.
Supplementary Materials
Supplemental material, sj-docx-1-tam-10.1177_17588359261472766 for HAIC combined with lenvatinib and tislelizumab for HCC with concurrent PVTT and pulmonary metastasis: a multicenter propensity score matching analysis by Song Chen, Zhiqiang Wu, Yi Xiao, Xiongying Jiang, Feng Shi, Ning Wang, Fangzeng Lin, Guosheng Tan, Wenquan Zhuang and Yonghui Huang in Therapeutic Advances in Medical Oncology
Supplemental material, sj-docx-2-tam-10.1177_17588359261472766 for HAIC combined with lenvatinib and tislelizumab for HCC with concurrent PVTT and pulmonary metastasis: a multicenter propensity score matching analysis by Song Chen, Zhiqiang Wu, Yi Xiao, Xiongying Jiang, Feng Shi, Ning Wang, Fangzeng Lin, Guosheng Tan, Wenquan Zhuang and Yonghui Huang in Therapeutic Advances in Medical Oncology