Response Rate Correlates with the Conversion Rate in Patients with Advanced Hepatocellular Carcinoma Treated with Systemic or Hepatic Artery Infusion Chemotherapy: A Systematic Review.

Introduction
Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related death worldwide [1]. Although potential curative therapies for early-stage HCC are surgical resection, liver transplantation, and local ablation, and standard therapy for intermediate-stage HCC is transarterial chemoembolization (TACE) [2–4], many patients with HCC are contraindicated for such therapies as they already have advanced or metastatic disease at diagnosis or recurrence after curative locoregional therapies. Systemic therapy, including combination immunotherapy and tyrosine kinase inhibitors (TKIs), has been recommended by several clinical practice guidelines as a first-line systemic therapy in patients with advanced HCC [2–4]; however, systemic therapy alone rarely cures these patients. As advanced HCC includes technically unresectable, which is usually decided based on the location, adjacent to major vessels, and remnant liver functional volume, and oncologically unresectable, which is defined as technically resectable but not suitable for upfront resection, and does not always run in parallel, the concept of borderline resectable and its definition for HCC based on vascular invasion and extrahepatic metastases were introduced by the Japan Liver Cancer Association and Japanese Society of Hepato-Biliary-Pancreatic Surgery [5].
Conversion therapy is the treatment strategy for converting tumors that are initially untreatable or borderline untreatable by curative-intent locoregional therapy to treatable tumors after they respond to systemic therapy. Several prior studies have indicated the feasibility and survival efficacy of conversion surgery for initially unresectable HCC [6, 7]. However, it is unclear which systemic treatment should be applied to patients who are potential candidates for conversion therapy. Sorafenib, the first TKI showing survival efficacy as a first-line treatment, has a low conversion rate [8, 9]. Lenvatinib, another multikinase inhibitor for advanced HCC used as a 1st-line treatment [10], has a relatively high conversion rate compared with that of sorafenib [11]. Although sorafenib and lenvatinib have similar survival efficacy and disease control rates (DCRs), the overall response rate (ORR) is significantly higher for lenvatinib than for sorafenib [10]. Therefore, the difference in the conversion rate between sorafenib and lenvatinib might be caused by the difference in the ORR between these treatments. Recently, combination immunotherapy, including atezolizumab plus bevacizumab (atezo + bev) and durvalumab plus tremelimumab (durva + treme), has shown superiority in terms of survival compared with sorafenib and is approved for advanced HCC as the 1st line of therapy [12, 13]. As these combination immunotherapies have shown higher DCRs and ORRs than does sorafenib, a higher conversion rate is also expected [14]. The present study aimed to assess the correlation of the conversion rate with the ORR by reviewing prior phase 2 and 3 studies for advanced HCC and to identify the appropriate therapy for patients with HCC who are potential candidates for conversion therapy.

Methods

Study Search and Selection Criteria
The current systematic review was carried out following PRISMA guidelines and the PRISMA 2020 checklist was used (online suppl. material; for all online suppl. material, see https://doi.org/10.1159/000550237), and MEDLINE on PubMed and Web of Science were used for the literature search. The search terms used were as follows: “hepatocellular carcinoma,” or “HCC,” “unresectable” or “advanced”; “prospective,” or “trial.” Studies published as full papers between January 2005 and April 2024 were retrieved. Initially, nonclinical studies, retrospective studies, case reports, reviews, editorials, protocol papers, studies for neoadjuvant therapy, studies reporting the use of TACE only, studies not published in English, animal studies, and studies not evaluating HCC treatment were excluded. Two reviewers (M.O. and H.N.) independently evaluated each article, and phase 2 or 3 studies assessing systemic therapies or systemic therapy and/or hepatic artery infusion chemotherapy (HAIC) combined with TACE and/or radiation therapy for advanced HCC with available data on ORR and DCR according to the Response Evaluation Criteria in Solid Tumors (RECIST) and the conversion rate were selected. Discrepancies were resolved by discussion and consensus. When multiple publications were identified from the same clinical study, only the most complete or recent publication was chosen. When multiple regimens were analyzed in one clinical study, each treatment arm was evaluated separately. The quality of the included observational studies was scored by 2 independent investigators screening the studies (K.O. and T.I.) using the Newcastle-Ottawa scale (NOS) with disagreements resolved after discussion (online suppl. Table S1) [15]. This study was not registered in any registry system.

Definition of Outcome and Statistical Analysis
The current analysis included technically unresectable and oncologically unresectable or borderline resectable cases due to the high liver tumor burden, vascular invasion and extrahepatic spread [5]. The disease stage was defined according to the Barcelona Clinic Liver Cancer (BCLC) staging system [2]. ORR and conversion rate were used as the common measure of association across studies. The definition of conversion therapy in the current analysis was the treatment strategy of systemic therapy and/or HAIC combined with or without TACE followed by curative-intent surgical resection, transplantation, or local ablation therapy. TACE was not considered conversion therapy after systemic therapy or HAIC. Since the ORR, DCR, and conversion rate data were not normally distributed on a test of normality, Spearman’s correlation coefficient was calculated to measure the strength of the relationship between the conversion rate and the ORR or DCR based on RECIST. If the studies included ORR and DCR assessment according to modified RECIST (mRECIST), these data were also included into the analysis [16]. Correlation coefficients between the ORR or DCR and median progression-free survival (PFS) and median overall survival (OS) were also calculated via Spearman’s test. JMP Pro 17 (SAS Institute Inc., Cary, NC, USA) was used for the statistical analysis. All tests were two-tailed, and p < 0.05 was considered significant.

Results

Studies Included in the Analysis
A total of 42 treatment groups in 32 phase 2 or 3 studies were selected and included in the analysis (Table 1) [17–48]. When the conversion rate was shown as 0, the studies were included in the analysis (Fig. 1). As the median PFS and OS were described in 37 patient groups and 36 patient groups, respectively, these groups were included in the analysis of the correlation between ORR/DCR and PFS/OS.

Correlation Analysis between the ORR and Conversion Rate
A total of 3,516 patients were analyzed in the included studies, and the ORR (based on RECIST) and conversion rates were 24.7% and 8.3%, respectively. First, the correlations between the ORR based on RECIST and the conversion rate was evaluated with the overall included patient groups. The ORR (ρ = 0.569, p < 0.0001) was moderately to strongly and significantly correlated with the conversion rate (Fig. 2a). Second, to assess the efficacy of presently available or potentially usable regimens for advanced HCC, we evaluated the correlation only in treatment groups with 1st-line immune checkpoint inhibitors (ICIs), TKIs and HAIC (27 treatment groups). Strong and significant correlation of the ORR (ρ = 0.647, p = 0.0003) with the conversion rate was also observed (Fig. 2b). Third, subgroup analyses of treatment groups with TKIs only as a 1st-line treatment was separately performed. As no study of ICIs only, only 4 treatment groups of ICIs + TKIs, 3 treatment groups of HAIC only, and 4 treatment groups of HAIC + TKIs were included in the current analysis, subgroup analyses with these treatment groups were not performed. There was no significant correlation of the ORR (ρ = 0.657, p = 0.16) based on RECIST with the conversion rate in the treatment group of TKIs only (online suppl. Fig. 1a). Fourth, the correlations between the ORR based on mRECIST and the conversion rate was evaluated with the overall included patient groups. The ORR (ρ = 0.579, p = 0.0047) was moderately to strongly and significantly correlated with the conversion rate (online suppl. Fig. 2a). As for the correlation analyses between DCR and conversion rate, the results are shown in the online supplementary Figure 3.

Correlation Analysis between ORR and PFS/OS
The correlation of the ORR (Fig. 3a) based on RECIST with the median PFS was strong and significant (ρ = 0.772, p < 0.0001). The ORR (Fig. 3b) based on RECIST was also strongly and significantly correlated with the median OS (ρ = 0.754, p < 0.0001) in the overall patient groups. In addition, in only the 27 patient groups treated with ICIs, TKIs, and HAIC, the median PFS (ρ = 0.843, p < 0.0001, Fig. 3e) and OS (ρ = 0.831, p < 0.0001, Fig. 3f) were strongly and significantly correlated with the ORR based on RECIST. Subgroup analysis in the treatment group of TKIs only revealed no significant correlations between ORR based on RECIST and median PFS (ρ = 0.543, p = 0.27, online suppl. Fig. 1b), and ORR based on RECIST and median OS (ρ = 0.500, p = 0.39, online suppl. Fig. 1c). The correlations of the ORR based on mRECIST with the median PFS (ρ = 0.671, p = 0.0012, online suppl. Fig. 2b) and the median OS (ρ = 0.786, p = 0.0002, online suppl. Fig. 2c) were strong and significant in the overall patient groups. The results of correlation analyses between DCR and median PFS/OS are shown in the online supplementary Figure 4.

Correlation Analysis between the Conversion Rate and PFS/OS
The conversion rate was significantly and moderately correlated with PFS (ρ = 0.644, p < 0.0001; Fig. 3c) and OS (ρ = 0.599, p = 0.0004; Fig. 3d) in the overall patient groups. Significant and moderate correlations of the conversion rate with PFS (ρ = 0.547, p = 0.038; Fig. 3g) and OS (ρ = 0.606, p = 0.0002; Fig. 3h) were also observed in the patient groups treated with ICIs, TKIs, and HAIC. Subgroup analysis in the treatment group of TKIs only revealed a strong and significant correlation of the conversion rate and median OS (ρ = 0.900, p = 0.037), while there was no significant correlation between the conversion rate and PFS (ρ = 0.714, p = 0.11) (online suppl. Fig. 1d, e).

Discussion
The current study aimed to investigate the correlation of the ORR with the conversion rate from 42 treatment groups in 32 already published phase 2 and 3 clinical studies using systemic chemotherapy, HAIC, ICI or TKIs for advanced HCC. The results revealed that the ORR based on RECIST are moderately to strongly correlated with the conversion rate. In addition, the ORR based on RECIST are strongly correlated with PFS and OS. Additionally, ORR assessed by mRECIST showed moderate-to-strong correlation with conversion rate and PFS/OS. These results, particularly the correlations between ORR and conversion rate, would be natural and predictable. However, no previous study has demonstrated these correlations through scientific analysis. Therefore, we believe the findings of the current analysis provide concrete evidence supporting what was previously hypothetical.
Our group reported a correlation between the ORR and conversion rate in the field of colorectal liver metastases [49]; however, it is uncertain whether high ORRs are also associated with high conversion rates in patients with HCC. The current analysis of patients with HCC revealed a moderate-to-strong correlation between the ORR according to both RECIST and mRECIST and conversion rate, similar to the findings of a prior study of patients with colorectal liver metastases.
One of the main focuses of conversion therapy for advanced HCC is determining which treatment regimen is most appropriate for induction therapy. Because of the heterogeneity of the included patients between the groups, the efficacy of conversion or survival cannot be directly compared between the groups included in the current analysis. As shown in Figure 2, the ICI + TKI group had a relatively high ORR, leading to a high conversion rate. However, it should be noted that these patient groups initially received induction therapy with the intention of undergoing conversion surgery, and selection bias may exist in these studies. In addition, some patient groups with other treatment regimens also had high conversion rates or ORRs. Although the subgroup analysis with patient groups with TKI only showed no significant correlations of ORR and conversion rate, this result might be due to the small number of samples (n = 6). Therefore, a specific treatment regimen cannot be recommended based on the current analysis.
Importantly, not only the correlation between the ORR and the conversion rate but also strong correlations between the ORR and PFS/OS were observed. Kudo et al. [50] also reported that the response rate is a predictor of OS in patients with advanced HCC on the basis of an analysis of 65 prior randomized clinical trials (RCTs). Although several reports have shown that patients who underwent conversion therapy had more favorable oncological outcomes than patients who did not [7, 11], the limited number of patients included in these 65 RCTs had completed conversion therapy, and it is uncertain whether the results of Kudo’s analysis can be applied to patients who are potential candidates for conversion therapy. However, the findings of the current analysis suggest that patients with advanced HCC who are potential candidates for conversion therapy should be treated with systemic therapy or HAIC, which has a high ORR or DCR and leads to prolonged PFS or OS even if these patients cannot complete conversion therapy. While the conversion rate was significantly and moderately correlated with RFS and OS in the current analysis, it is still unclear whether conversion therapy itself is directly associated with prolonged PFS and OS according to the current analysis. Further studies evaluating the efficacy of conversion therapy for advanced HCC may be needed to answer this question.
Some treatment markers other than the ORR and DCR, such as depth of response, which is defined as the maximum percentage change in tumor size compared with baseline, and duration of response (DoR), which is defined as the time from onset of response to progression or death for any reason, have been introduced as surrogates of survival in several cancers as well as HCC [51]. As these markers may be associated with PFS and OS in patients with HCC, the correlations between these markers and the conversion rate are another area of interest. However, no studies included in the current analysis described the actual value of the DoR. In addition, because only 7 patient groups had DoR data, no significant correlation was observed between the DoR and conversion rate in the current study (data not shown). Therefore, research in this regard is needed in the future.
The limitation of the current study is heterogeneity in patient inclusion criteria in each clinical study included in this analysis. Given that the definition of unresectable or advanced HCC differs between studies, direct comparisons between treatment groups are inappropriate. However, because this study comprehensively analyzed the correlation coefficient of each patient group, differences in patient selection between the groups might have had little effect on the results. Another limitation includes publication bias. Although the patient group whose conversion rate was reported as 0 was included in the current analysis, we did not include studies that did not report the conversion rate because we could not decide if the conversion rate was actually 0 or was merely unreported. Given that studies reporting the conversion rate may more or less consider conversion therapy, applying the results of the current analysis to patients who are potentially candidates for conversion therapy may be acceptable.

Conclusion
Regimens for systemic treatment or HAIC with a high ORR should be selected for patients with advanced HCC who are candidates for conversion therapy. Even if the conversion treatment is not successful, a high ORR is still related to prolonged PFS and OS.

Statement of Ethics
A statement of ethics is not applicable because this study is based exclusively on published literature.

Conflict of Interest Statement
The authors have no conflicts of interest to declare.

Funding Sources
This study was partially supported by JSPS KAKENHI, Grant No. 21K16486.

Author Contributions
M.O., H.N., and E.H. made substantial contributions to the conception, design, acquisition, analysis, and interpretation of the data. K.O., T.I., and H.H. made substantial contributions to the acquisition and interpretation of the data. M.O., H.N., K.O., and E.H. drafted the article and revised it critically for important intellectual content. T.I. and H.H. critically revised the article for important intellectual content. All authors approved the final version of the manuscript to be published.