Characteristics of cabozantinib treatment in advanced hepatocellular carcinoma

Cabozantinib is a molecular targeted agent (MTA) used for treatment of advanced hepatocellular carcinoma (HCC). Although its superiority over placebo has been proven, its effectiveness and risk factors in real‐world practice are needed to be elucidated.


| INTRODUC TI ON
Hepatocellular carcinoma (HCC) is one of the leading causes of cancer-related death worldwide. 13][4] Currently, the combination therapy of atezolizumab and bevacizumab is considered as first-line therapy for advanced HCC 5 ; however, no clear evidence has been presented for the appropriate choice of second-line therapy drugs and beyond.
Because of the lack of effective companion diagnostics, which is the ideal way to select appropriate treatments, it is necessary to examine the outcomes of the therapies in practice precisely and determine the indications for appropriate treatment selection.
Cabozantinib is an inhibitor of tyrosine kinases, including vascular endothelial growth factor (VEGF) receptors 1, 2 and 3, MET, and AXL, and is a candidate for second-line therapy for HCC. 6Since HCC is a hypervascular tumour, inhibition of VEGF is effective in HCC treatment.Furthermore, inhibition of MET and AXL helps overcome resistance to agents that target the VEGF receptor pathway. 7The

| Diagnosis
HCC was diagnosed based on the findings of imaging modalities, such as computed tomography (CT), magnetic resonance imaging (MRI) or angiography.The diagnostic criteria for HCC were based on the practice guidance by the American Association for the study of liver diseases. 8

| Treatment
The patients received either a 60-mg tablet of cabozantinib or one to two 20-mg tablets taken orally once daily.The starting dose of cabozantinib was determined by each physician based on the patient's age and status.The adverse events (AE) were managed by treatment interruption or dose reductions according to the instructions provided by the supplier.The severity of AE was assessed according to the National Cancer Institute Common Terminology Criteria for AE, version 4.0.Tumour response was assessed using CT or MRI every 2-3 months according to the Response Evaluation Criteria in Solid Tumours, version 1.1. 9

| Statistical analysis
The baseline characteristics are presented as medians and ranges.
The continuous variables were compared using the Wilcoxon ranksum test, and categorical values were analysed using the chi-square test.The prognosis was analysed using Cox hazard analysis, Kaplan-Meier method and log-rank test.All significance tests were twosided; p-values <0.05 were considered statistically significant.All analyses were performed using JMP software program (version 16.0; SAS Institute Japan Ltd.).

| Characteristics of the patients
The median age of the patients was 72. 5

| Treatment response
The best treatment are shown in Table 2.The objective response rate and disease control rate were 3.7% and 40.7%, respectively.The median overall survival (OS) and progression-free survival (PFS) were 6.9 and 4.4 months, respectively (Figure S1).The patients with partial response (PR)/stable disease (SD) showed better survival rates than those with progressive disease (PD) (Figure 1).
The median OS of the patients with PR/SD, PD and not evaluable (NE) disease were 12.7, 5.8 and 4.6 months, respectively (p = 0.010).

| Safety
Unpredictable AE was not observed in this study (Table 3).AE of any grade was observed in 47 (87.0%) patients, and appetite loss (48.1%) and fatigue (40.7%) were the two major AE associated with cabozantinib.Grade 3/4 AE were observed in 20 patients (37.0%), and 30 patients (55.6%) discontinued the treatment due to AE. Hepatic decompensation (ascites or encephalopathy) was observed in 8 patients; however, there was no significant difference in OS between these patients and the rest of the group.
The incidence of AE (Grade 3/4) did not differ between the patients with and without a reduced starting dose, with a reported incidence of 82.9% and 94.7% (37.1% and 36.8%),respectively.

| Risk factors for survival
Cox proportional analysis revealed that high albumin-bilirubin score 10 and neutrophil-lymphocyte ratio (NLR) 11,12 were significant risk factors for survival in the univariate analysis (Table 4).We set the cut-off at 4 based on findings from our previous study. 11The survival curves were significantly better in patients with high NLR (>4) than in those with low NLR (Figure 2).4][15][16] This analysis revealed that the occurrence of Grade 3/4 AE was a negative risk factor for survival (hazard ratio for death, 0.36; 95% confidence interval [CI], 0.16-0.83;p = 0.016), whereas high NLR (>4) was a positive risk factor (hazard ratio for death, 2.35; 95% CI, 1.41-4.82;p = 0.020).
The same analysis was conducted for PFS (Table 5).Occurrence of Grade 3/4 AE was the only negative risk factor for PFS in univariate analysis (hazard ratio for disease progression or death, 0.43; 95% CI, 0.20-0.90;p = 0.025).Furthermore, multivariate analysis was conducted with the same variables used in the OS analysis and the occurrence of Grade 3/4 AE was the only negative risk factor (hazard ratio for disease progression or death, 0.33; 95% CI, 0.15-0.73;p = 0.006).
Neither preceding therapy with atezolizumab/bevacizumab nor reduced starting dose correlated with patient survival.

| DISCUSS ION
This study demonstrated the efficacy of cabozantinib in the treatment of HCC in real-world practice in Japan.The OS and  The OS and PFS reported in the CELESTIAL study (10.2 and 5.2 months, respectively) were longer than those in this study.8][19][20][21][22] The OS and PFS in those studies were 6.8-7.7 and 2.1-3.7 months, respectively.There are three major differences between the CELESTIAL study and real-world practice.First, the proportion of Child-Pugh Class A cases in the CELESTIAL study was 100%, whereas that in the published reports and this study was 72%-84% and 79.6%, respectively.Second, only one MTA (sorafenib) was used in over 70% of the patients before starting cabozantinib in the CELESTIAL study.Many treatment guidelines for HCC recommend combination therapy with atezolizumab and bevacizumab, and several other MTAs can be used after the first-line treatment. 4,23,24Therefore, more than half of the patients (59.3%) were treated with atezolizumab/bevacizumab, not sorafenib, and 88.9% of the patients were treated with two or more MTAs before starting cabozantinib in this study.8][19][20][21][22] Although prior immune therapies might work, multiple prior treatments may worsen the liver function and induce drug resistance.Therefore, it is better to use cabozantinib as early as we can to benefit better from this drug.
Third, it is common to start treatment with reduced dose of MTAs and subsequently increase the dosage in Japan.Cabozantinib follows the same practice.Each doctor made the decision regarding the reduced starting dose and frailer patients may have been included in the reduced starting dose group.However, we did not observe any significant effect of the reduced starting dose on survival or PFS (Tables 4 and 5).
Appetite loss (48.1%) and fatigue (40.7%) were two major AE in this study, and their percentages were similar to those reported in the CELESTIAL study (48.2% and 45.4%, respectively).However, the incidence of elevated aspartate aminotransferase levels, diarrhoea and hypertension was lower in this study than in the CELESTIAL study (13.0% vs. 22.5%, 13.0% vs. 53.7%and 5.6% vs. 29.3%,respectively).
The possible reason for these discrepancies could be the treatment duration.4][15][16] The same association was observed clearly for cabozantinib in this study.
NLR, an indicator of systemic inflammation, is reportedly associated with poor outcomes in many malignancies, including HCC. 11 We previously reported that a high NLR is associated with poor outcomes in patients with HCC treated with lenvatinib, 12 and the same association was observed with cabozantinib.
This study has several limitations.First, it was a retrospective study with no scheduled radiological assessment, resulting in variations in PFS, and the sample size of patients was small.Due to the small sample size, we could not analyse the relationship between each AE and effectiveness of cabozantinib.Second, it is unclear whether similar results can be obtained for other regions.The data in this study were obtained from multiple centres; hence, the results are representative of Japan.However, we cannot determine the outcomes outside of Japan.Third, we could not eliminate the possibility of immortal bias concerning the impact of the occurrence of 3/4AE.Time-dependent Cox regression or landmark analysis at different timepoints are ideal methods to eliminate the immortal time bias.However, the median duration of cabozantinib treatment was too short (1.18 months) to conduct such analysis.
Nevertheless, it is clear that cabozantinib can be used safely in real-world practice, although its usage is different from that in the CELESTIAL study.Moreover, we found that the NLR and occurrence of Grade 3/4 AE are closely related to patient survival.Further largescale studies are warranted to confirm these results.

ACK N O WLE D G E M ENTS
This study was conducted on behalf of the Real-life Practice Experts for HCC (RELPEC) Study Group and HCC 48 Group (hepatocellular carcinoma experts from 48 clinics in Japan).

FU N D I N G I N FO R M ATI O N
This research received no external funding.

CO N FLI C T O F I NTER E S T S TATEM ENT
None of the authors have a conflict of interest to disclose.

2 |
PATIENTS AND ME THODS 2.1 | Patients This study enrolled 54 consecutive patients with advanced HCC, who were treated with cabozantinib at 19 different hospitals between December 2020 and July 2022 (Kansai Medical University, Kagawa University, Takasaki General Medical Center, Teine Keijinkai Hospital, Ehime Prefectural Central Hospital, Saiseikai Niigata Hospital, Osaka Medical and Pharmaceutical University, Hamamatsu University School of Medicine, University of Toyama, Asahi General Hospital, Ehime University, Kagawa Prefectural Central Hospital, Takamatsu Red Cross Hospital, Saiseikai Maebashi Hospital, Matsuyama Red Cross Hospital, Nippon Medical School Chibahokusoh Hospital, Okayama City Hospital, Ogaki Municipal Hospital and Himeji Red Cross Hospital).The final observation date was 22 December 2022.All patients provided permission to review their clinical records for this study.The study protocol conformed to the tenets of the Declaration of Helsinki and was approved by the Institutional Ethics Committee of Ehime Prefectural Central Hospital (IRB # 30-66) (UMIN000043219) and of each participating institution.
The median survival time of the patients with high NLR (>4) and low NLR (≤4) were 2.1 and 9.1 months, respectively.We performed a multivariate analysis of the factors showing p-values <0.10 and the presence of Grade 3/4 AE, which reportedly correlated with the survival associated with TA B L E 2 Treatment effect.

TA B L E 4 F I G U R E 2
Risk factors for survival.Effect of neutrophil-lymphocyte ratio on the overall survival (OS) of patients with hepatocellular carcinoma treated with cabozantinib.The OS of the patients with high NLR (>4; thin line) is significantly shorter than that of the patients with low NLR (≤4; thick line) (p = 0.022).The median survival time of the patients with high NLR (>4) and low NLR (≤4) was 2.1 months and 9.1 months, respectively.NLR, neutrophil-lymphocyte ratio.PFS were 6.9 months and 4.4 months, respectively.Patients with PR/ SD showed better survival than those with PD.Most of the patients experienced AE (87.0%), and Grade 3/4 AE were observed in 37% of the patients.In addition, we demonstrated that neither preceding therapy with atezolizumab/bevacizumab nor reduced starting dose correlated with patient survival.Notably, this is the first study to reveal that the presence of Grade 3/4 AE is significantly correlated with better OS and PFS and that high NLR (>4) is a significant factor for poor OS.
Patient characteristics.
mALBI grade (1/2a/2b/3) 14/11/28/1 Child-Pugh stage A 43 (79.6%) The most important finding of this study was that both Grade 3/4 AE and NLR correlated with the OS and PFS.Several studies TA B L E 5 Risk factors of progressionfree survival.waslonger than that in our study (3.8 vs. 1.2 months); hence, the chance of these AE occurring was high.Overall, the profiles of AE in real-world practice are not different from those in the CELESTIAL study, indicating that cabozantinib can be used safely in real clinic settings.