Efficacy and Safety of Stereotactic Body Radiotherapy Combined with Camrelizumab and Apatinib in Patients with Hepatocellular Carcinoma with Portal Vein Tumor Thrombus

This article is featured in Selected Articles from This Issue, p. 4019

Hepatocellular carcinoma (HCC) is a major health problem and the third leading cause of cancer-related deaths worldwide (1). A substantial proportion of patients with HCC present with portal vein tumor thrombus (PVTT) at the time of diagnosis or recurrence after treatment (2). In the Barcelona Clinic Liver Cancer (BCLC) staging system, the presence of PVTT allocates the patient to an advanced stage beyond potentially curative treatments in most cases, such as resection, liver transplantation, or ablation, leading to a poor prognosis with only 2 to 4 months of survival time via best support care (3, 4).

Sorafenib, an antiangiogenic multikinase inhibitor, has been the only recommended standard systemic therapy for patients with advanced HCC for more than a decade (5, 6). However, the efficacy of sorafenib is modest with limited survival benefits. Over the past 2 years, several new targeted agents have demonstrated efficacy, including lenvatinib in the first-line setting and regorafenib and cabozantinib in the second-line setting (7, 8). Recently, growing clinical evidence has demonstrated the effectiveness of combination therapies (8). The IMbrave 150 study showed a meaningful survival benefit with atezolizumab plus bevacizumab over sorafenib in patients with unresectable HCC (9). On the basis of the results of this study, atezolizumab combined with bevacizumab was recently approved by the FDA for patients with unresectable HCC who had not received prior systemic therapy. Moreover, the phase Ib KEYNOTE-524 trial demonstrated that lenvatinib and pembrolizumab showed promising efficacy and acceptable safety in treatment-naïve unresectable HCC (10). The RESCUE study, a phase II trial conducted in a Chinese population, reported that camrelizumab combined with apatinib had encouraging antitumor effects with a tolerable safety profile in patients with advanced HCC in both first- and second-line settings (11). On the basis of these findings, combination therapy with an anti-PD-1 antibody and antiangiogenic agents may be a potential first-line treatment option for patients with advanced HCC.

In addition to systemic therapy, evidence on the prospect of locoregional treatment, such as radiotherapy (RT) and transcatheter arterial chemoembolization (TACE), in the management of HCC with PVTT has been mounting with favorable results (12). A randomized clinical trial comparing the efficacy and safety of radiotherapy plus TACE with sorafenib alone in patients with untreated HCC with macrovascular invasion (n = 90) indicated that the TACE-RT group had a significantly better prognosis than the sorafenib group, with a higher radiologic response rate at 24 weeks (33.3% vs. 2.2%; P < 0.001), longer median time to progression (TTP; 31.0 vs. 11.7 weeks; P < 0.001) and OS (55.0 vs. 43.0 weeks; P = 0.04; ref. 13).With recent technological advances, stereotactic body radiotherapy (SBRT) allows high doses of radiation to be applied to the targeted tumor area in a limited number of fractions with less damage to the surrounding healthy tissues (12, 14). It has been reported that SBRT can achieve high rates of local tumor control (14, 15). However, many patients would experience out-of-field progression, highlighting the need for concurrent systemic disease control (15). A retrospective study investigating the safety and feasibility of palliative radiotherapy combined with PD-1/PD-L1 inhibitors and angiogenesis agents in BCLC stage C HCC (n = 16) showed that the objective response and disease control rates were 40.0% [95% confidence interval (CI), 16.3–67.7] and 86.7% (95% CI, 59.5–98.3), respectively, with acceptable safety (16). Nevertheless, only 8 patients had PVTT in this study; thus, it was difficult to draw a valid conclusion about the efficacy of triple therapy.

This prospective study was conducted to determine the efficacy and safety of camrelizumab plus apatinib combined with or without SBRT in patients with HCC with PVTT.

This multicenter, open-label, noncomparative, randomized controlled trial was conducted at six hospitals in China. Our study was conducted in accordance with Declaration of Helsinki. The trial was performed after approval by the Institutional Review Board, and registered in the Chinese Clinical Trial Registry (ChiCTR1900027102). All patients provided written informed consent prior to enrollment.

The study was first registered on the Chinese Clinical Trial registry (chiCTR1900027102) as a random controlled study with SBRT group as the intervention group and non-SBRT group as the control group, and the protocol was published in Frontiers in Oncology (17). Then we realized that a noncomparative, randomized trial would be more appropriate because sorafenib rather than PD-1 antibody combined with antiangiogenic multikinase inhibitor was the standard treatment in clinical practice for advanced HCC at that time. Therefore, we revised our study protocol with ethical approval on the basis that the core elements of the study stayed unchanged.

The diagnosis of HCC was based on biopsy or clinical criteria of the European Association for the Study of Liver guidelines (18). PVTT was confirmed using the typical radiological patterns on dynamic contrast-enhanced CT, MRI, or angiography. PVTT was divided into five types according to Cheng's classification: I0, microscopic tumor thrombus; I, segmental or sectoral branches of the portal vein; II, right- or left-side branch of the portal vein; III, main trunk of the portal vein; and IV, superior mesenteric vein (19).

The key criteria were as follows: (i) patients with HCC with Cheng's type II/III/IV PVTT who did not receive prior systemic therapy; (ii) age ≥18 years; (iii) Eastern Cooperative Oncology Group (ECOG) performance status score of 0 or 1; (iv) Child-Pugh A or B; (v) at least one hepatic lesion measurable on CT or MRI as defined by RECIST 1.1, which has not been previously treated with locoregional therapies; (vi) prior locoregional therapies including surgery, radiotherapy, and TACE were permitted, but radioembolization was not allowed; (vii) locoregional therapy must be completed at least 4 weeks prior to the baseline scan and all toxic reactions > 1 (National Cancer Institute–Common Terminology Criteria for Adverse Events, NCI-CTCAE version 5.0) related to prior locoregional treatment must be resolved; and (viii) patients with hepatitis B virus (HBV) infection must be eligible to enroll if HBV DNA < 500 IU/mL or 2,500 copies/mL and must be administered at least 14 days of antiviral therapy.

All potentially eligible patients were randomly assigned to the SBRT or non-SBRT cohorts in a 2:1 ratio. Randomization was performed using computer-generated random number code. Stratification was not performed prior to randomization. The patients and investigators were not blinded to treatment allocation. Allocation concealment was performed by using sequentially numbered opaque sealed envelopes. Patients and investigators, except for those who assessed radiological and pathologic responses, were not blinded to the allocated study treatment.

SBRT began within 1 week of random assignment. Four-dimensional CT was used in the treatment planning to account for respiratory motion. The gross tumor volume (GTV), as determined by dynamic enhanced CT or MRI, encompassed the PVTT and contiguous primary hepatic lesion. In the case of large, multicentric, or diffuse primary disease where the entire tumor volume is large and adequate normal liver is not possible to spare from high radiation dose, only the PVTT was regarded as the GTV. The internal target volume (ITV) was generated to account for the extent and position of the tumor during the breathing cycle, and the planning target volume (PTV) had a uniform ITV expansion of 3 to 5 mm. PTV was adjusted manually to minimize overlapping the bowel when indicated. The initial prescription dose to PTV was 40 Gy in five fractions delivered by using 6 MV X-rays with a linear accelerator per week (Varian Medical Systems). When the PTV was adjacent to bowel or large enough to cause difficulties to meet the dose–volume constraints of the organs at risk, the dose was modified to 36 to 39 Gy in 5 to 6 fractions. The dose–volume constraints for organs at risk were defined as follows: liver constraint was ≥700 mL of uninvolved liver (liver minus GTV) receiving <15 Gy and/or the percentage of normal liver volume receiving more than 15 Gy (V₁₅)was less than one-third of the absolute normal liver volume (V_total); the maximum dose to 1 mL of stomach, small bowel, or duodenum was <25 Gy; for the spinal cord, the maximum dose to 1 mL was <15 Gy; and for the kidneys, V₁₅ was less than 1/3 V_total.

After enrollment, all patients received camrelizumab plus apatinib until disease progression, unacceptable toxicity, withdrawal of consent, investigator's decision, or study completion. The first dose of camrelizumab and apatinib was administered within 1 week after SBRT for the SBRT cohort and on day 1 of the first cycle for the non-SBRT cohort. Camrelizumab was administered intravenously at a standard dose of 200 mg every 3 weeks. Dose interruptions of camrelizumab for ≤12 weeks were permitted; however, dose reduction was not allowed. Apatinib was administered orally at a dose of 250 mg daily. Dose discontinuations, interruptions, and modifications in the dose frequency of apatinib (250 mg/day for 5 days on−2 days off, or 7 days on−7 days off) were allowed.

All patients were monitored regularly (Supplementary Fig. S1). During the treatment period, imaging assessment for efficacy was conducted using dynamic contrast-enhanced CT or MRI every 6 weeks until disease progression or treatment discontinuation (whichever occurred later). Survival data were collected every 12 weeks until death, loss to follow-up, or study completion. The RECIST guidelines version 1.1 (RECIST 1.1) were used to evaluate the treatment response of primary liver tumors and extrahepatic lesions. For PVTT, partial response (PR) was defined as any downstaging in the PVTT classification or conspicuous blood flow restoration in the portal vein; progressive disease (PD) as any upstaging in the PVTT classification; and stable disease (SD) as targets that satisfied neither PR nor PD. The overall effect was defined as PD if the primary liver tumor, extrahepatic lesions, or PVTT was regarded as PD and PR if the primary tumor, extrahepatic lesions, or PVTT was classified as PR, while the other was defined as SD (20). Adverse events were evaluated according to the Common Terminology Criteria for Adverse Events (version 4.0).

The primary endpoint was overall survival (OS; defined as the time from drug initiation to death resulting from any cause or last known date alive). The secondary endpoints were progression-free survival (PFS; time from drug initiation to the first documented PD or death from any cause), objective response rate [ORR; percentage of patients with a confirmed complete response (CR) or PR], disease control rate (DCR; percentage of patients who achieved CR, PR, or SD), duration of response (time from response to progression or death), TTP (time from drug initiation to the first documented PD), and safety.

On the basis of the outcomes of SHARP and ORIENTAL study (21, 22), the median OS of the historical control group (sorafenib treatment) in patients with HCC with PVTT was 4.5 months (H0). The estimate of median OS of the non-SBRT cohort was on the basis of several most recently published clinical trials regarding anti-PD-1/PD-L1 combined with antiangiogenic agents in advanced HCC. A retrospective study assessing the efficacy of camrelizumab combined with apatinib for HCC with PVTT reported that the OS periods of patients with type II and type III were 15.9 months (95% CI: 12.1–19.7), and 5.8 months (95% CI: 3.9–7.7), respectively (23). Also, the updated IMbrave150 data reported that the median OS period was 7.6 months (95% CI: 6.0–13.9) with atezolizumab + bevacizumab in patients with Vp4 (24). Because our non-SBRT patient cohort was composed of type II (45%), type III (30%), and type IV (25%) PVTT of patients with HCC, we hypothesized that camrelizumab–apatinib could increase the median OS to 8.5 months (H1). Seventy percent power by the one-sample log-rank test with 0.05 two-sided alpha was utilized for sample size calculation. On the basis of the hypothesis, 20 patients were required for the efficacy evaluation. For SBRT cohort, a total of 40 patients were required because of a 2:1 randomization ratio between the SBRT and non-SBRT cohort.

For the SBRT cohort, all patients who completed SBRT and received at least one dose of camrelizumab and apatinib were included in the full analysis set. For the non-SBRT cohort, we analyzed patients who received at least one dose of camrelizumab and apatinib. We summarized the tumor response results as frequencies and proportions. Kaplan–Meier analysis was used to estimate the median OS and PFS. The corresponding 95% CI of median OS and PFS was computed by the survfit() function in R software based on the method of Brookmeyer and Crowley (25). Safety data are presented as the frequency and proportion of patients who experienced each event.

Statistical comparison will not be conducted between the two cohorts. The sample size calculation was conducted using PASS 15 software (NCSS, LLC.); the median OS, PFS, TTP, and the corresponding 95% CIs were calculated in R software (version 4.3.1); and all other analyses were performed using SPSS software (version 21.0; SPSS, Inc., RRID:SCR_002865).

The data generated in this study are available upon request from the corresponding author. Readers should contact Jun Xue to initiate the individual subject data request process. Individual subject data described in this article will be shared after deidentification. To gain access, researchers need to provide a methodologically sound proposal and sign a data access agreement.

From January 1, 2020 to February 28, 2022, a total of 64 patients were screened for eligibility. Sixty patients were enrolled and randomly assigned to receive camrelizumab (200 mg every 3 weeks) and apatinib (250 mg every day) with or without SBRT (95% PTV 36–40 Gy/6–8 Gy). All 60 patients were included in the modified intention-to-treat analysis (Fig. 1). The baseline characteristics of the patients were summarized in Table 1 (Supplementary Table S1). The median age was 53 years and 85.0% of the patients were men. All patients had an ECOG performance status of 0 or 1, and most had Child-Pugh class A liver function [47 (78.3%)]. Chronic HBV infection was the leading cause of liver disease [49 (81.7%)]. Most patients had Cheng's type II/III PVTT [49 (81.7%)], multiple intrahepatic lesions [43 (75.0%)], and extrahepatic metastasis [32 (71.7%)]. Representativeness of study participants is shown in Supplementary Table S2.

By the data cut-off date for the final analysis (February 28, 2022), 6 patients were still under treatment. SBRT was designed to target only the PVTT in 37 patients and both the PVTT and primary intrahepatic tumors in 3 patients due to the dose constraints. The most common reason for treatment discontinuation was disease progression, which occurred in 24 (60.0%) and 12 (60.0%) patients in the SBRT and non-SBRT cohorts, respectively. Median follow-up was 8.4 months [interquartile range (IQR), 6.4–12.9] for SBRT and 7.9 months (IQR, 4.2–10.3) for non-SBRT cohort; median duration of treatment was 4.4 months (IQR, 2.5–7.7) and 2.5 months (IQR, 1.6–5.2), respectively. Median duration of response was 5.3 months [range, 2.2–not available (NA)] for SBRT and 4.6 months (range, 3.2–NA) for non-SBRT cohort. Median TTP was 4.6 months (95% CI, 3.3–7.0) and 3.6 months (95% CI, 2.5–NA) for the SBRT and non-SBRT cohorts, respectively.

A total of 15 (37.5%) deaths had occurred in the SBRT cohort and 14 (70%) in the non-SBRT cohort. Median OS and PFS were 12.7 months (95% CI, 10.2–NA; Fig. 2A) and 4.6 months (95% CI, 3.3–7.0) in the SBRT cohort (Fig. 3A), and 8.6 months (95% CI, 5.6–NA; Fig. 2B) and 2.5 months (95% CI, 2.0–7.6) in the non-SBRT cohort (Fig. 3B). The lower bounds of 95% CI for median OS in both cohorts exceeded the point estimate of 4.5 months from the historical control (sorafenib treatment) in patients with HCC with PVTT. In this case, both groups did better than historical control. Kaplan–Meier estimates of 6-month OS rates and PFS rates were 89.3% and 46.3% for the SBRT cohorts, and 62.9% and 20.9% for the SBRT and non-SBRT cohorts, respectively (Table 2).

Regarding tumor response, we evaluated PVTT and primary liver tumor, and showed the best responses in Table 2. For PVTT assessment, CR was observed only in one patient, and PR was observed in 11 of 40 patients (27.5%) in the SBRT cohort and in 1 of 20 patients (5.0%) in the non-SBRT cohort. For primary liver tumors, 13 (32.5%) patients in the SBRT cohort and 3 (15.0%) patients in the non-SBRT cohort had a PR according to RECIST 1.1 (Fig. 4A and B). For overall effect, the ORR and DCR were 47.5% and 72.5% in the SBRT cohort, and 20.0% and 40.0% in the non-SBRT cohort, respectively. We also conducted the subset analysis for Child-Pugh grade and extrahepatic involvement in Supplementary Table S3.

All the 60 patients were included in the safety analysis (Table 3). The safety profiles were similar between the SBRT and non-SBRT cohorts. All patients experienced at least one treatment-related adverse event (TRAE). The most common TRAEs of any grade were hypertension [33 (55.0%)], hand-foot syndrome [31 (51.7%)], and leukopenia [30 (50.0%)]. Patients in the SBRT cohort experienced more instances of hyperbilirubinemia than did patients in the non-SBRT cohort [22 (55.0%) vs. 7 (35%)]. Grade ≥ 3 TRAEs were reported in 13 (21.7%) patients. Of the 60 patients, 11 (18.3%) discontinued treatment because of serious TRAEs, 35 (58.3%) patients had TRAEs leading to any treatment interruption or dose modification and 1 (1.7%) patient died because of TRAEs (myocarditis).

PVTT is one of the most important prognostic indicators in patients with HCC. However, the optimal therapeutic regimens for patients with HCC with PVTT have not been identified (26). Recently, there has been a growing interest in combination therapies, including combination locoregional and systemic therapies, to provide superior treatment responses (27, 28). Therefore, there is an urgent need to conduct randomized clinical trials of combination therapies, specifically for these patients. In this study, we observed the clinical benefits of combination treatment with camrelizumab–apatinib with or without SBRT in patients with systemic treatment-naïve HCC and PVTT. We found that both treatment approaches showed better results than historical sorafenib therapy.

Radiotherapy is an integral component in the treatment of various cancers, including HCC (29). The increasing interest in combining radiotherapy, immune checkpoint blockade (ICB), and antiangiogenesis agents stems from the rationale of potential synergistic antitumor effects (27, 30). Radiotherapy can convert what is considered the "cold" tumor microenvironments (TME) to an immune-reactive "hot" TME, which will improve the effectiveness of ICB (31). In turn, ICB could block the upregulation of immune checkpoint molecules induced by radiotherapy and restore effective antitumor immune responses. Meanwhile, antiangiogenesis can significantly normalize the tumor vasculature, which boosts cytotoxic immune cell infiltration and improves the efficacy of radiotherapy (30). Despite promising preclinical outcomes, clinical data on the efficacy of combination therapies for HCC are relatively scarce. A retrospective cohort study showed that SBRT combined with a PD-1 inhibitor had an ORR of 71.0% and a median PFS of 19.6 months in patients with intermediate-stage HCC refractory to TACE (32). A phase I study reported that axitinib (selective VEGFR 1–3 inhibitor) in combination with radiotherapy was well tolerated, with a 66.7% ORR in 9 patients with advanced HCC (33). To our knowledge, this study is the first prospective randomized noncomparative clinical trial to assess the role of radiotherapy followed by ICB and antiangiogenic agents in HCC with PVTT.

In the current study, camrelizumab plus apatinib with or without SBRT yielded clinical benefits in patients with systemic treatment-naïve HCC and PVTT. Median OS and PFS were 8.6 months (95% CI, 5.6–NA) and 2.5 months (95% CI, 2.0–7.6) for patients treated with camrelizumab–apatinib. Nevertheless, the survival outcomes reported in the RESCUE trial (median OS: not reached, median PFS: 5.7 months) were better than those reported in our study (11). This comparison should be made with caution because the proportion of patients with PVTT in the RESCUE trial was only 28.6%. According to the latest findings of a phase III trial released at the European Society for Medical Oncology 2022 Congress, treatment with camrelizumab plus apatinib as first-line therapy resulted in 25.4% ORR among patients with unresectable HCC, which was similar to the 20.0% ORR for non-SBRT cohort in our study (34). For SBRT cohort, median PFS was 4.6 months (95% CI, 1.6–7.6 months), and 6-month and 12-month OS rates were 89.3% and 51.3%, respectively. ORR and DCR were 47.5% and 72.5%, respectively. Only 3 (7.5%) patients had both the PVTT and primary intrahepatic tumors treated in SBRT cohort, which may cause a lower ORR than we expected. The latest prospective trial (TRIPLET study) assessed the efficacy and safety of hepatic artery infusion chemotherapy (HAIC) combined with apatinib and camrelizumab for BCLC stage C HCC. The median PFS time was 9.37 months (95% CI, 7.00–11.73), and the 6-month, 12-month, and 18-month OS rates were 93.1%, 85.8%, and 65.8%, respectively. ORR and DCR were 70.96% and 87.10%, respectively. The population in the TRIPLET study was characterized by 100% Child-Pugh class A liver function, 12.90% extrahepatic metastasis, and PVTT types I-II/III-IV (25.81%/45.16%). In the current study, 13 (21.6%) patients had Child-Pugh class B liver function, 32 (53.3%) had extrahepatic metastasis, and 29 (48.3%) and 11 (18.3%) had types III and IV PVTT, respectively. RTOG 1112 reported that OS significantly improved with SBRT followed by sorafenib versus sorafenib alone in patients with advanced HCC (15.8 vs. 12.3 months). Although these results seem better than that found in the current study, our patients had poorer liver function and more advanced tumor stages (35, 36).

In our study, the safety profile showed no new or unexpected toxicities. Reactive cutaneous capillary endothelial proliferation (RCCEP), hypothyroidism, and myocarditis are likely associated with camrelizumab, whereas hand-foot syndrome, hypertension, and proteinuria are likely to be associated with apatinib (11, 37, 38).The incidence of alanine aminotransferase increase, aspartate aminotransferase increase, hyperbilirubinemia, decreased appetite, nausea, and thrombocytopenia was slightly higher in the SBRT cohort, which might be associated with SBRT. It was worth noting that a decline in liver function, including elevated transaminases and hyperbilirubinemia, was transiently observed within 3 months of radiation. This risk might be mitigated by careful attention to normal liver constraints (39). The most common ≥ grade 3 TRAEs are hypertension and hand-foot syndrome, which can be controlled by medicine or dose modification and are usually not life threatening. No patients who received SBRT developed radiation-induced liver disease. To date, no cases of viral hepatitis flares have been reported.

This study has several limitations. First, it was performed during the COVID-19 pandemic, which affected the regular treatment and visits. Some patients experienced delays in medication administration, which may accelerate tumor progression. Radiographic assessment delays resulting in an extended observation period may prolong the PFS. Therefore, the effect of COVID-19 on the PFS remains unclear. Second, blinding was not possible because of the nature of the treatments. Bias in the open-label trial may have occurred in this study. Third, the generalizability of our results may be limited because most patients [49 (81.7%)] in our study were infected with HBV. Finally, although this was a noncomparative trial, baseline extrahepatic disease and Child-Pugh score appeared to be the factors with potential imbalances between the two cohorts, which was probably due to chance. These two factors could potentially have contributed to the observation of shorter OS in non-SBRT group. Randomization stratified by these two factors should be a consideration in our future trial design.

In conclusion, this randomized clinical trial demonstrated that treatment with camrelizumab–apatinib with or without SBRT had promising antitumor activity with acceptable tolerability in patients with HCC with PVTT not previously treated with systemic therapy. This study also indicated that a combination of locoregional and systemic therapies may be a suitable treatment option for this patient population.

No disclosures were reported.

Y. Hu: Data curation, formal analysis, validation, investigation, writing–original draft, writing–review and editing. M. Zhou: Data curation, investigation, writing–review and editing. J. Tang: Resources, validation, investigation, writing–review and editing. S. Li: Resources, supervision, investigation, project administration. H. Liu: Investigation, visualization, project administration. J. Hu: Data curation, investigation. H. Ma: Investigation. J. Liu: Investigation. T. Qin: Data curation, software, formal analysis. X. Yu: Investigation. Y. Chen: Investigation. J. Peng: Investigation. Y. Zou: Investigation. T. Zhang: Conceptualization, resources, supervision, funding acquisition, validation, investigation, visualization, project administration, writing–review and editing. J. Xue: Data curation, software, formal analysis, writing–original draft, writing–review and editing.

This work was supported by Jiangsu Hengrui Medicine Co., Ltd. This study was supported by grants from the National Natural Science Foundation of China (U21A20376, to J. Xue; 81902934, to Y. Hu; and 81901714, to M. Zhou). We thank all patients, their families, and all participating clinical teams for making this study possible.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.