1.1 Study Design and Patients

This retrospective study encompassed patients who initiated second-line systemic treatment with TKI or a combination of TKI and ICI (TKI-ICI) after HCC progressed on LEN between May 2018 and December 2023 at Beijing Ditan hospitals. Eligible participants met the following criteria: (1) age > 18 years; (2) presence of at least one measurable target lesion as defined by modified Response Evaluation Criteria in Solid Tumors (mRECIST) [16]; (3) Eastern Cooperative Oncology Group Performance Status (ECOG PS) of 0–1; (4) Child-Pugh class A or B (score ≤ 7). Patients were excluded if they had: (1) received any prior systemic therapy other than LEN monotherapy; (2) severe dysfunction of the heart, brain, liver, or kidneys; or (3) coagulation disorders or a high risk of hemorrhage (gastroesophageal varices of grade III, gastric ulcer, and duodenal ulcer). Notably, before the start of the second-line treatment, all the patients performed upper gastrointestinal endoscopy, and the hemorrhagic risk was assessed. Subsequently, patients were classified based on the type of second-line treatment administered after HCC progression or discontinuation of LEN due to serious adverse events (AEs).

1.2 Treatment Regimens and Assessments

Patients previously treated with LEN (12 mg/day for body weight ≥ 60 kg or 8 mg/day for body weight < 60 kg) received subsequent systemic therapy as determined by their attending specialists upon confirmation of tumor progression or unacceptable toxicity. REG was administered at an initial dose of 80 mg/day in 4-week cycles (3 weeks on, 1 week off). SOR (400 mg twice daily), donafenib (200 mg twice daily), or apatinib (250 mg three times daily) was administered in 4-week cycles. Patients also received SIN, camrelizumab, or tislelizumab at 200 mg intravenously on day 1 of a 21-day cycle following the first REG, SOR, donafenib, or apatinib dose. Treatment continued until disease progression, unacceptable toxicity, or withdrawal by the patient or physician.

Patients received second-line systemic treatment for 4 weeks per cycle, with treatment response and safety evaluated every 8–12 weeks. Radiological response was assessed according to mRECIST criteria based on liver dynamic computed tomography (CT) or magnetic resonance imaging (MRI) [16]. OS and PFS were calculated from the initiation of second-line treatment. ORR and disease control rate (DCR) were based on the best radiographic response observed during treatment. ORR was defined as the proportion of patients achieving complete response (CR) and partial response (PR), while DCR included stable disease (SD) in addition to CR and PR. Treatment-related AEs were assessed according to the Common Terminology Criteria for Adverse Events version 5.0 (CTC-AE 5.0). The primary endpoint was PFS, and the secondary endpoints included OS, ORR, DCR, and safety.

1.3 Variable Collection

Baseline data collected included age, sex, etiology, previous treatment history, ALBI score, and BCLC stage. BCLC stage information encompassed tumor size, number, ECOG PS, portal vein tumor thrombus (PVTT), extrahepatic metastases, and Child-Pugh class. Besides, the baseline alpha-fetoprotein (AFP) levels and AFP levels after 8 weeks of treatment were collected.

1.4 Inflammation Indices

The analysis included the examination of various inflammation indices, including neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), prognostic nutritional index (PNI), albumin-toto-neutrophil ratio index (ANRI), lymphocyte-toto-monocyte ratio (LMR), symptom number (SN), systemic immune-inflammation response index (SIRI), and C-reactive protein (CRP). Considering the varying threshold values outlined in the existing literature, we use X-tile software to determine the cut-off values (NLR: 5.7; PLR: 197.6; PNI: 403; ANRI: 5.4; LMR: 2.3; SN: 1314.3; SIRI: 4.0; CRP: 46.1 mg/L) and subsequently undertook the process of dichotomizing the index-related data.

1.5 Statistical Analysis

Categorical variables were expressed as numbers (percentages), while continuous variables with skewed distributions were presented as medians [interquartile range]. Baseline characteristics, treatment responses, and AEs between groups were compared using Pearson's χ² test or Fisher's exact test, as appropriate. Survival curves were generated by the Kaplan–Meier method. Factors independently predictive of OS and PFS were evaluated using univariate and multivariate Cox proportional hazards models, incorporating variables with p < 0.1 from the univariate analysis into the multivariate analysis. Statistical analyses were conducted using SPSS (version 26.0) and R software (version 4.1.3, http://www.rproject.org), with p < 0.05 considered statistically significant.

2.1 Baseline Characteristics

Between May 2018 and December 2023, a total of 85 HCC patients received LEN as first-line treatment and subsequently discontinued treatment. Of these, 76 patients who proceeded to second-line systemic therapy were considered eligible， while 6 patients were excluded for the following reasons: Child-Pugh score > B7 (n = 1), ECOG PS > 1 (n = 2), or loss to follow-up (n = 3). Ultimately, a total of 70 patients were enrolled in the study, of whom 21 patients received mono-TKI treatment and 49 patients received TKI-ICI combination therapy after LEN failure (Figure 1).

Baseline characteristics are summarized in Table 1. The median age was 56 years (IQR, 51–53), with 59 (84.3%) patients being male. The majority of patients, 63 (90.0%) had HBV-related HCC, 52 (74.3%) had BCLC stage C tumors, and 47 (67.1%) had Child-Pugh A liver function. At the start of second-line therapy, 26 (37.1%) and 34 (48.6%) of patients exhibited the presence of PVTT and extrahepatic metastases, respectively. With regard to prior tumor treatments, 9 (12.9%) patients had undergone hepatectomy, 66 (94.3%) patients had received trans-arterial chemoembolization (TACE), and 35 (50.0%) patients had undergone ablation. Although the proportion of patients with AFP > 400 ng/mL was higher in the TKI group than in the TKI-ICI group (52.4% vs. 34.7%, p = 0.166), the demographic and disease characteristics of the patients at baseline were balanced between the treatment groups.

2.2 Treatment Profiles of First-Line LEN and Second-Line Therapy

Of 70 patients who received first-line LEN treatment, 55 patients experienced disease progression, and 15 patients discontinued treatment due to adverse events (grade 3/4 bilirubin increase, grade 4 hypertension, grade 3 ALT/AST elevation, and proteinuria). All patients were subsequently transitioned to a second-line therapeutic regimen. In the second-line therapy, the TKI group comprised patients receiving REG (n = 12) or SOR (n = 9). The TKI-ICI group was divided into three subgroups: REG plus ICI (n = 23), SOR plus ICI (n = 10), and Oth plus ICI (n = 16) (Figure 1).

2.3 Survival Analysis

During the follow-up duration, 72.9% (n = 51) of patients had died, and 72.9% (n = 51) of patients had experienced disease progression. The median PFS in the whole population was 4.44 months (95% CI, 3.25–5.55 months), and the median OS was 10.93 months (95% CI, 9.14–12.72 months). The median PFS in the TKI–ICI group was 5.27 months (95% CI, 4.09–6.44 months), and that in the TKI group was 3.10 months (95% CI, 2.04–4.16 months) (Figure 2A). The median OS results were similar to the PFS result, with a median OS of 12.53 months (95% CI, 10.79–14.26 months) for the TKI–ICI group and 7.50 months (95% CI, 4.06–10.94 months) for the TKI group (Figure 2B). The survival curves for the subdivided treatment subgroup are presented in Figure 2C,D, with the respective median PFS and OS detailed in Table 2.

After 8 weeks of second-line therapy, 51.0% (n = 25) of patients in the TKI–ICI group and 57.1% (n = 12) of patients in the TKI group developed elevated AFP levels. When stratified by AFP levels, patients with baseline AFP ≤ 400 ng/mL had a median PFS of 4.73 months (95% CI, 2.80–6.66) and a median OS of 12.17 months (95% CI, 10.34–14.0 months). Patients with AFP > 400 ng/mL had a median PFS of 4.13 months (95% CI, 2.21–6.05) and a median OS of 9.20 months (95% CI, 6.33–12.07 months) (Figure 3A,B). Furthermore, a decrease in AFP levels was not associated with improved survival outcomes (Figure 3C,D).

2.4 Therapeutic Responses

Therapeutic responses to second-line treatments are summarized in Table 3, as assessed by mRECIST criteria. Four (19.1%) patients in the TKI group had a confirmed ORR, including 1 (4.8%) patient with CR and 3 (14.3%) with PR. In the TKI-ICI group, the ORR was 16.3%, including 1 (2%) patient who achieved CR and 7 (14.3%) patients who achieved PR. The DCR was 85.7% in the TKI-ICI group and 61.9% in the TKI group. After 8 weeks of second-line treatment, patients with decreased AFP levels exhibited an ORR of 24.2% and a DCR of 78.8%, while those with elevated AFP levels showed an ORR of 16.2% and a DCR of 73.0% (Figure 4). This result demonstrated that the pattern of serologic response was not associated with the mRECIST response.

2.5 Predictors for Survival Outcomes

Predictors for survival outcomes are depicted in Table 4.

Univariable and multivariate Cox regression analyses identified AFP > 400 ng/mL (HR 2.04; 95% CI 1.10–3.79; p = 0.03) as the sole independent predictor for OS. Significant predictors for PFS in univariable analysis included second-line treatment modality, CRP, ANRI, LMR, NLR, PLR, PNI, SN, and SIRI. Multivariable analysis revealed that TKI-ICI treatment (HR 0.49; 95% CI 0.26–0.95; p = 0.033) and high ANRI (HR 0.23; 95% CI 0.08–0.26; p = 0.004) were protective predictors of HCC progression. Furthermore, in the TKI-ICI population, AFP (HR 2.68; 95% CI 1.26–5.70; p = 0.01) remained an independent predictor for OS, and high ANRI (HR 0.27; 95% CI 0.08–0.87; p = 0.029) was also associated with improved PFS (Table 5).

2.6 Safety Analysis

Treatment-related adverse events of any grade occurred in 79.6% of patients in the TKI-ICI arm and 85.7% in the TKI arm, the most common of which were hypertension, diarrhea, nausea, and vomiting. A total of 9 patients (18.4%) in the TKI-ICI group and 5 (23.8%) in the TKI group experienced at least one grade ≥ 3 AE. It is noteworthy that immune-related AEs in the TKI-ICI group included two cases of immune hepatitis, one case of immune myocarditis, and one case of interstitial pneumonia. Additionally, 7 patients (14.3%) in the TKI-ICI group and 3 (14.3%) in the TKI group experienced AEs leading to dose reduction. The proportion of patients who discontinued treatment due to AEs was 8.2% in the TKI-ICI group and 4.8% in the TKI group (Table 6).

2.7 Subsequent Therapy

Among the TKI group, 5 (25%) patients received subsequent systemic treatment, including apatinib (n = 1), REG plus SIN/camrelizumab (n = 3), and LEN plus SIN (n = 1). 15 (75%) patients in the TKI-ICI group performed a 3rd line systemic treatment, with options including apatinib (n = 2), apatinib plus SIN/cadonilimab/tislelizumab (n = 5), LEN (n = 1), LEN plus SIN/Tislelizumab (n = 2), REG plus camrelizumab (n = 3), and atezolizumab plus bevacizumab (n = 2).