Efficacy of treatment based on TKIs in combination with PD-1 inhibitors for unresectable recurrent hepatocellular carcinoma

This study retrospectively collected 213 consecutive patients who were found to have HCC recurrence after initial hepatectomy of primary HCC from January 2017 to November 2022 at the Chinese People’s Liberation Army (PLA) General Hospital. Recurrence was diagnosed via pathological examination findings or with the non-invasive criteria used by the American Association for the Study of Liver Diseases [16]. Among these patients, 79 were diagnosed with unresectable recurrent HCC. We excluded 35 patients because they had a history of systemic therapy or histopathological findings not consistent with HCC (Fig. 1). The inclusion criteria for this study included (1) the recurrence of primary HCC after radical surgery; (2) upper abdominal magnetic resonance imaging (MRI) findings suggest at least one of the following features: (a) more than three recurrent lesions in the liver; (b) tumor invading macrovascular such as the portal vein, hepatic vein or inferior vena cava; and (c) extrahepatic metastases; (3) expected survival ≥ 12 weeks; (4) pathological examination confirms the diagnosis of HCC; (5) Eastern Cooperative Oncology Group performance status (ECOG PS) score ≤ 1; and (6) Child-Pugh score < 10. Exclusion criteria include (1) systemic therapy (including new molecular targeted drugs or immunotherapy drugs) before HCC recurrence and (2) history of hepatic encephalopathy and oesophageal or gastric variceal bleeding. The determination of the HCC tumor stage using the Barcelona Clinical Liver Cancer (BCLC) system and assessment of portal vein tumor thrombosis (PVTT) according to Japan’s liver cancer study group [17, 18]. PVTT can be graded as Vp1 (portal invasion at the third or more peripheral portal branch), Vp2 (portal invasion at the second portal branch), Vp3 (portal venous invasion at the first portal branch), and Vp4 (portal invasion at the main portal trunk) [17]. The study was approved by the ethics committee of the Chinese PLA General Hospital (Approval No. S2018-111-01).

All patients received TKIs in combination with anti-PD-1 antibodies. TACE or RFA treatment depends on the size or the number of tumors and the anatomical location of the patient’s intrahepatic lesions. Lenvatinib (bodyweight ≥ 60 kg, 12 mg/day; < 60 kg, 8 mg/day) [14] or apatinib (750 mg/day) [19] were administered orally as TKIs. Anti-PD-1 antibodies were intravenously administered as follows: nivolumab (240 mg every 2 weeks) or pembrolizumab (200 mg every 3 weeks), or sintilimab [20] (200 mg every 3 weeks). TACE was performed by super-selective catheterization of the tumor-feeding hepatic arteries with the 2.7 F microcatheter (Progreat, Terumo Corporation, Japan). An emulsion of 6–20 mL of lipiodol (Lipiodol Ultrafluide; Guerbet, Paris, France), 30–50 mg of epirubicin (Farmorubicin; Pharmacia, Tokyo, Japan), 100–150 mg of oxaliplatin, 500–750 mg of 5-fluorouracil, 200–300 mg of calcium folinate, and 10–14 mg of mitomycin C (Kyowa Hakko Kogyo, Tokyo, Japan) were injected into the tumor supply vessels through the microcatheter. TKIs and Anti-PD-1 antibodies were initialized simultaneously, followed by TACE within 2 weeks. TKIs and anti-PD-1 antibodies were administered continuously until unacceptable toxicity or disease progression.

All patients are followed up every 2 months after the start of treatment. Blood tests include liver function, thyroid function, and tumor biomarkers. MRI was performed every 8–10 weeks, and Positron emission tomography-computed tomography (PET-CT) was performed as appropriate. Treatment outcomes were assessed for PFS, OS, ORR, disease control rate (DCR), and adverse events calculating PFS from the first day of combination therapy to tumor progression or death and OS from the first day of combination therapy to patient death. ORR is defined as the percentage of patients achieving complete response (CR) and partial response (PR). DCR is defined as the percentage of patients achieving CR, PR, and stable disease (SD). The last follow-up was performed on November 2022. Efficacy assessment is mainly based on modified response evaluation criteria in solid tumors (mRECIST) [21], and treatment-related adverse events were defined using the National Cancer Institute–Common Terminology Criteria for Adverse Events (NCI-CTCAE) version 5.0.

We used SPSS 25.0 (SPSS, IL, USA) and R version 4.1.0 (Institute for Statistics and Mathematics, Vienna, Austria) to analyze and visualize our results. Quantitative data that conforms to a normal distribution were expressed as the means ± standard deviation. Non-normally distributed quantitative data were expressed as median values (range). Categorical data were expressed as a number (percentage). The PFS and OS were illustrated using Kaplan–Meier curves.

A total of 44 patients were enrolled in this study (Table 1). The median duration of follow-up was 22 months (range, 5–54 months). Forty of the 44 patients were male (90.9%) with a mean age of 57 ± 8.9 years, and a mean body mass index (BMI) of 25.0 ± 3.3. 37 patients previously had hepatitis B infection (90.0%), 34 of whom had liver cirrhosis. Five patients (11.4%) had previous hepatitis C infections. Twenty-four patients (54.5%) recurred within 1 year after radical hepatectomy, while 20 patients (45.5%) recurred more than 1 year after radical hepatectomy. There were 23 patients with more than three intrahepatic lesions, seven with PVTT, 23 patients with extrahepatic metastases, seven patients with both more than three intrahepatic lesions and extrahepatic metastases, one patient with both more than three intrahepatic lesions and PVTT, and one patient with both extrahepatic metastases and PVTT in the study. Bone metastases, lung metastases, lymph node metastases, abdominal metastases, and diaphragmatic metastases were found in five, seven, three, five, and three patients, respectively. The liver function of most patients (90.9%) was normal (Child-Pugh A class). Furthermore, the mean levels of alanine aminotransferase, aspartate aminotransferase, albumin, and total bilirubin were 32.1 ± 23.6 U/L, 36.6 ± 23.1 U/L, and 41.0 ± 5.8 g/L, and 15.0 ± 6.9 mg/dL, respectively.

All patients received TKIs in combination with PD-1 inhibitors therapy, including 15 patients who received TACE, and three patients who received TACE in combination with RFA treatment (Table 2). During the follow-up period, 21 of 44 patients died of tumor progression. The assessment of treatment efficacy data is shown in Table 3. Six patients had CR (13.6%), and 12 patients (27.3%) had PR, two of whom underwent liver transplantation and secondary liver resection, respectively. SD was achieved in 10 patients (22.7%). PD was observed in 16 patients (36.4%), of which 9 patients died by the last follow-up. The ORR was 40.9%, whereas DCR was 63.6% according to the mRECIST. Median overall survival was 27.0 months (95% CI 21.2, 32.8) with a 1-year OS rate of 83.6% (95% CI 77.9%, 89.3%). Median PFS was 15.0 months (95.0% CI 12.1, 17.9), with a 1-year PFS rate of 77.0% (95% CI: 70.6%, 83.4%) (Fig. 2). The media follow-up time was 17 months after treatment.

The most common treatment-related adverse events (TRAEs) were elevated alanine aminotransferase (ALT) or aspartate aminotransferase (AST) (16 of 44, 36.4%), followed by diarrhea (13 of 44, 29.5%), post-embolization syndrome (PES) (13 of 18, 72.2%), skin reactions (12 of 44, 27.3%), and hypertension (10 of 44, 22.7%) (Table 4). Grade 3 TRAEs in four patients (9.1%) with the main symptoms and presentations being skin reaction (2.3%), diarrhea (6.8%), hypertension (4.5%), fatigue (2.3%), increased ALT or AST (2.3%), and hyperbilirubinemia (2.3%). There were no treatment-related deaths.

A 47-year-old male patient (described as patient A) underwent right hepatic tumor resection on September 1, 2019, for primary liver cancer. The postoperative pathological examination was suggestive of a moderately and poorly differentiated hepatocellular carcinoma. The tumor was 12×7.4×7 cm in size, with intravascular tumor thrombus (> 5) around the tumor. No tumor was found in the liver margin, and the pathological stage was BCLC-A. The patient received an MRI re-examination on December 9, 2019, suggesting a new lesion in the liver with PVTT (Vp3) (Fig. 3A–C). After a comprehensive evaluation, the patient received apatinib in combination with sintilimab therapy for financial reasons firstly. After 9 weeks of therapy, the MRI re-examination suggested no significant change in tumor and PVTT (Fig. 3D–F). For a better outcome, the patient subsequently received lenvatinib in combination with sintilimab therapy. After 7 months of therapy, the MRI re-examination suggested a significant reduction in tumor diameter, and no significant changes in PVTT (Fig. 3G–I). The efficacy could be assessed as PR according to mRECIST criteria, and no significant adverse effects were observed during the treatment. After a preoperative evaluation (Fig. 4A–C), the patient underwent a right hemihepatectomy, cholecystectomy, and portal vein right branch tumor embolectomy on September 22, 2020 (Fig. 4D–F). The postoperative pathological examination revealed a localized poorly differentiated HCC (the proportion of residual tumor cells was about 20%). The total size of the tumor area was approximately 5×3×3 cm. A small amount of regressed cancerous tissue was seen in the portal vein (the proportion of residual tumor cells was about 20%). No tumor was found in the liver margin, and the pathological stage was BCLC-C. The patient recovered well and continued to receive lenvatinib in combination with sintilimab therapy 1 month after surgery. The MRI re-examination on April 11, 2021, showed no recurrence in the coelom (Fig. 4G–I). The patient has achieved survival of 34 months after combined treatment and recurrence-free survival of 25 months after repeat hepatectomy, as of November 2022.

Similarly, another 40-year-old male patient (described as patient B) underwent radical hepatectomy on April 28, 2019. The postoperative pathological examination was suggestive of a moderately and poorly differentiated hepatocellular carcinoma. The tumor was 2×1.8×1.2 cm in size, and no tumor was found in the liver margin. This patient on August 26, 2019, had a re-examination MRI suggesting intrahepatic recurrent lesions with hilar lymph node metastasis (Fig. 5A–C). After 12 weeks of treatment with lenvatinib combined with pembrolizumab, MRI re-examination showed significant regression of the hilar lymph nodes and a significant weakening of the arterial enhancement phase (Fig. 5D–F). The treatment efficacy can be assessed as CR according to the mRECIST criteria. The patient subsequently underwent a liver transplant on February 27, 2020, and postoperative pathology revealed no surviving carcinoma tissue in the lymph nodes. Six months after surgery, the abdominal computed tomography (CT) scan showed normal portal veins, hepatic veins, and bile ducts, with no recurrent lesions (Fig. 5G–I). The patient’s survival had reached 37 months after combined treatment, and the recurrence-free survival after liver transplantation had reached 32 months, as of November 2022.