The impact of curative conversion therapy aimed at a cancer‐free state in patients with hepatocellular carcinoma treated with atezolizumab plus bevacizumab

Abstract Background and Aims We aimed to validate the predictive factors for tumor response and the prognostic impact of conversion therapy aimed at cancer‐ and drug‐free states in patients with unresectable hepatocellular carcinoma (u‐HCC) undergoing atezolizumab plus bevacizumab (Atez/Bev) therapy. Methods This retrospective study enrolled 156 patients who were Child‐Pugh class A with u‐HCC treated using Atez/Beva. The profile of objective response was investigated using decision‐tree analysis. Progression‐free, recurrence‐free, and overall survival were assessed. Results The progression‐free and overall survival were 6.1 and 18.0 months, respectively. Objective response and disease control rates were 32.0% and 84.0%, respectively. Decision‐tree analysis revealed that neutrophil‐to‐lymphocyte ratio (NLR) <3, modified albumin‐bilirubin grade (m‐ALBI) 1 or 2a, and age < 75 were sequential splitting variables for the objective response, respectively. In the multivariate analysis, NLR <3 and m‐ALBI grade 1 or 2a were identified as predictive factors for objective response. We successfully achieved eligibility for conversion therapy in 17 patients after Atez/Bev therapy significant response. Following conversion therapy, the curative therapy group, including surgical resection or radiofrequency ablation (RFA), had significantly higher recurrence‐free survival than did the transcatheter arterial chemoembolization (TACE) and Atez/Bev discontinuation (surgical resection or RFA; not reached vs. TACE; 5.3 months, p = 0.008, Atez/Bev discontinuation; 3.9 months, p = 0.048, respectively) groups. Conclusions NLR <3 and m‐ALBI grade 1 or 2a were predictive factors for conversion therapy, leading to cancer‐ and drug‐free states in patients with u‐HCC undergoing Atez/Bev therapy. Moreover, surgery or RFA may be suitable for conversion therapy for cancer‐free status.


| INTRODUCTION
Hepatocellular carcinoma (HCC), the major malignant primary liver cancer, is the third leading cause of cancerrelated deaths worldwide. 1,2 Many patients with unresectable HCC (u-HCC) receive systemic therapy because HCC is often diagnosed at an advanced stage. 3,4 Thus, several molecular targeted agents for treating u-HCC have been developed, such as sorafenib, 5 lenvatinib, 6 ramucirumab, 7 and cabozantinib. 8 Moreover, immune checkpoint inhibitors have recently revolutionized the treatment strategy for u-HCC.
Atezolizumab plus bevacizumab (Atez/Bev), a combination of the atezolizumab (anti-programmed death ligand-1 inhibitor) and bevacizumab (vascular endothelial growth factor inhibitor), was approved for the firstline systemic therapy for patients with u-HCC based on the IMbrave150 trial's finding. 9 Atez/Bev treatment improved progression-free survival (PFS) and overall survival (OS) compared to sorafenib, and it is expected to shrink tumors more than the previously used systemic therapy. [9][10][11] Currently, the treatment strategy for u-HCC is sequential therapy with systemic therapy. [12][13][14] However, cancerfree conversion therapy, such as surgery, radiofrequency ablation (RFA), or curative transarterial chemoembolization, can be performed if adequate tumor reduction is achieved with Atez/Bev therapy. 15 Several reports have discovered that OS was prolonged in patients who received liver resection or RFA as conversion therapy for colorectal cancer liver metastasis after tumor reduction with chemotherapy. [16][17][18][19] Thus, establishing predictors of tumor shrinkage is important. Previous studies reported that neutrophil-to-lymphocyte ratio (NLR), albumin-bilirubin (ALBI) grade, 20,21 neo-Glasgow prognostic scores, 22 and prognostic nutritional index 23 were the predictive factors associated with therapeutic effect or OS in Atez/Bev treatment for HCC. However, the predictive factors for the response to Atez/Bev treatment have not yet been established. Moreover, few studies have reported the impact of conversion therapy aimed at cancer-and drug-free status after Atez/Bev treatment, and the associated outcomes are unknown.
Therefore, in this study, we aimed to investigate the predictive factors for tumor response in u-HCC patients treated with Atez/Bev. Additionally, we aimed to evaluate the impact of conversion therapy aimed at cancer-and drug-free states after an effective response to Atez/Bev treatment.

| Patients
Between 2020 and October 31, 2022, this multicenter retrospective cohort study enrolled 177 consecutive u-HCC patients treated with Atez/Bev at seven Japanese institutions. The following eligibility criteria were used: (i) age > 18 years; (ii) Eastern Cooperative Oncology Group performance status (PS) <2; and (iii) clinical data were available and it could follow-up until study cessation (December 31, 2022) or death. Exclusion criteria were as follows: (i) Child-Pugh class B or C; (ii) history of autoimmune disease; and (iii) active esophageal varices. Thus, 21 patients were excluded. Finally, this study enrolled 156 patients ( Figure S1).
This research was conducted according to the Helsinki Declaration and approved by the Ethics Committee of the Kurume University School of Medicine (approval code: 20183). An opt-out approach was used to obtain informed consent from patients.

| Confirmation of HCC
HCC confirmation was based on hypervascularization in the arterial phase and washout in the portal venous phase or delayed phase on CT or MRI examination. Additionally, in MRI, focal areas with a suspicious hypointense signal in the hepatobiliary phase were used to detect HCC. For the cutoff of tumor marker level, AFP ≥7 was considered positive in this study. HCC was diagnosed using a combination of serum markers and imaging following recommendations of The Japan Society of Hepatology. 24

| Treatment protocol
The patient was administrated Atez/Bev according to the recommended dosage (1200 mg of Atez and 15 mg/kg of Bev intravenously every 3 weeks). The clinical guidelines for Atez/Bev created by the manufacturer were used K E Y W O R D S atezolizumab plus bevacizumab, conversion therapy, hepatocellular carcinoma for the discontinuation of each component agent if a treatment-related adverse event occurred. Atezolizumab therapy was interrupted when any unacceptable grade 2 or over grade 3 immune-related AEs occurred such as liver injury, skin disorder, and drug-induced pneumonia. 9,25 These patients received bevacizumab monotherapy. Bevacizumab was interrupted when any unacceptable grade 2 or over grade 3 bevacizumabrelated AEs occurred such as proteinuria and bleeding. 26 These patients received atezolizumab monotherapy. Patients received the treatment until the development of unacceptable adverse events (AEs) or tumor progression. Atez/Bev could be continued even beyond tumor progression if clinical benefit is still observed; among patients who refused other systemic treatments, Atez/ Bev was continued if disease progression was slow and adverse events of Atez/Bev treatment were acceptable.

| Assessment of therapeutic response and safety
The Response Evaluation Criteria In Solid Tumors version 1.1 (RECIST v1.1). 27 was used to evaluate therapeutic response using CT/MRI 6 weeks after the initiation of Atez/Bev. The Common Terminology Criteria for Adverse Events version 5.0 was adopted to assess AEs every 3 weeks until study cessation or death. 28

| Definition of eligibility for conversion therapy after Atez/Bev and additional conversion therapy
We defined eligibility for conversion therapy as follows: (1) no viable lesion on CT or MRI (such as necrotic cases) scans and range within normal values for tumor markers; AFP was the independent risk factor for early recurrence after surgical resection for HCC, 29,30 and (2) condition in which additional treatment was expected to achieve cancer-and drug-free statuses. We recommended all patients who fulfilled the eligibility criteria for conversion therapy to achieve cancer-and drug-free state, and the optimal treatment for conversion therapy for each patient was determined by discussions among hepatologists, radiologists, and surgeons following the Japanese practice guidelines for HCC. 24

| The decision-tree algorithm
A decision-tree algorithm was constructed to reveal the profiles' relationship with conversion therapy and objective response (OR) following the instructions provided with the R software package as described previously. 14,31 The profiles were identified by data-mining methods.

| Statistical analysis
All provided data were described as medians (ranges). Between-group comparisons were carried out using the Mann-Whitney U test, the Kruskal-Wallis test, and a nonparametric analysis of variance. If the oneway analysis of variance was significant, differences between individual groups were analyzed using the Fisher least significant difference test. Evaluation of PFS, recurrence-free survival (RFS), and OS was carried out using the Kaplan-Meier method, and statistical differences were evaluated with the log-rank test or Bonferroni method. A decision-tree analysis was carried out to investigate the factors of conversion therapy and objective response (OR), as previously described. 14 Statistical significance difference was defined as p < 0.05. Statistical analyses were performed using JMP software (JMP Pro version 15, SAS Institute Inc.).

| Decision tree associated with possible conversion therapy
In the present study, the conversion therapy rate in all participants was 10.9% ( Figure 1). A decision-tree analysis was performed to determine the profiles related to conversion therapy. RECIST was detected as the first splitting factor for the conversion therapy rate. The conversion therapy rate was 0.0% in patients with SD or PD treated with Atez/Bev; however, the conversion therapy rate was 32% in patients with PR treated with Atez/Bev.

| Overall survival according to treatment response by RECIST
The MST was not reached, 15.7 months, and 8.4 months among patients who achieved PR, SD, and PD, respectively ( Figure 2). The PR group had a significantly higher MST than the SD group (not reached vs. 15.7 months, p = 0.001). In addition, the SD group had a significantly higher MST than the PD group (MST: 15.7 months vs. 8.4 months, p < 0.001).

| Decision-tree for OR
In the present study, the OR rate in all participants was 32.0% ( Figure 3). A decision-tree analysis was conducted to determine the profiles related to OR. NLR was detected as the first splitting variable for OR. The OR rate was 40.1% in patients with NLR <3, while it was only 16.3% in patients with NLR ≥3. In patients with NLR <3, the 2nd and 3rd splitting factors were m-ALBI grade and age. The OR rate was 58.9% in patients with NLR <3, m-ALBI grade 1 or 2a, and age < 75.

| Multivariate analyses of variables associated with OR and OS
NLR and m-ALBI grade were identified as independent variables for OR (Table 2), while NLR, AFP, and m-ALBI grade were identified as independent variables for OS (Table 3).

| Patient characteristics with NLR <3 and ≥3
The characteristics of patients in the NLR <3 and ≥3 groups are summarized in Table S2. There were no significant differences in age, sex, m-ALBI grade, extrahepatic spread, macrovascular invasion, AFP, or treatment lines between the two groups. However, tumor size and tumor number were significantly more frequent in the NLR >3 group than those in the NLR <3 group.

| The eligibility for conversion therapy after Atez/Bev therapy
The proportion of patients who were eligible for conversion therapy was 10.9% (17/156). Five, three, and three patients underwent surgery, RFA, and transarterial chemoembolization, respectively, as a treatment for conversion therapy. The prevalence of curative therapy as surgery and RFA was 47.1% (Table 4). All six patients for whom Atez/Bev treatment was discontinued belonged to the no viable lesion on imaging category and tumor markers within normal ranges; however, it was impossible to predict pathological CR among these patients, and so we recommended these patients for additional treatment aimed at a cancer-free state because of the risk of developing adverse events due to continued Atez/Bev treatment. However, these patients requested to discontinue Atez/Bev treatment,  Figure S3).

| Patient characteristics at the time of receiving conversion therapy with surgery or RFA and TACE
The characteristics of patients at the time of receipt of conversion therapy are listed in Table S3. There were no significant differences in age, ALBI score, tumor size, tumor number, or AFP between the two groups. Figure 4 shows RFS after curative therapy, TACE, and Atez/Bev discontinuation. The curative therapy group had a significantly higher RFS than that in the other therapy groups.  20 Moreover, several studies have reported that NLR can predict outcomes in patients with u-HCC treated with Atez/Bev, and these results are similar to those of previous studies that revealed that low NLR was associated with therapeutic response compared to high NLR. 21,32 NLR high is neutrophilia, which is related to high tumorassociated macrophage infiltration and inflammatory cytokine production. 33 In addition, relative lymphopenia impairs the host immune response against tumor cells, making it a poor prognostic factor for cancer patients. 34 NLR is expected to serve as a baseline biomarker for patients receiving Atez/Bev because it is a simple biomarker for predicting inflammation in the body. 35 Conversion therapy aims to reduce tumor volume and provide a chance at curative surgery for unresectable cancer patients. 36 This concept is an important treatment strategy for several solid tumors, including colorectal cancer, gastric cancer, and pancreatic cancer. Various therapies are candidates for conversion therapy 37,38 ; however, this concept has never been developed because there is no evidence of treatment with a high therapeutic response for HCC. However, conversion therapy could be made possible for HCC through advances in systemic therapies. Several studies have reported that patients with u-HCC could undergo hepatectomy after systemic therapy with combined tyrosine kinase inhibitor/anti-PD-1 antibodies. 39,40 Moreover, a previous study reported a significantly higher therapeutic effect among patients receiving Atez/ Bev as the first-line therapy than that in patients receiving Atez/Bev later. 41 Although this study included approximately 40% of patients who were treated with Atez/Bev as the second-line or later treatment strategies, the proportion of patients who were eligible for conversion therapy was 10.9%. In other words, as more cases were treated with Atez/Bev as the first line, the proportion of patients eligible for conversion therapy may increase. In this study, we also revealed that NLR <3 was associated with conversion therapy, leading to cancer-and drug-free states. However, the OR rate was 16.3% in patients with NLR ≥3 by the decision tree. This response rate was higher than other systemic therapies. Therefore, we should not avoid patients from Atezo/Bev therapy group even without favorable factors. However, not all patients who achieved an OR are candidates for conversion therapy. Among the patients who had achieved OR, there were many cases wherein conversion therapy for a cancer-free state could not be performed because of tumor and host factors, such as the number of tumors, tumor diameter, extrahepatic spread, vascular invasion, and deterioration of liver function. Moreover, the evidence of the clinical benefit of switching conversion therapy for patients receiving Atez/ Bev is still not elucidated sufficiently. Thus, further study should focus on evaluating conversion therapy in u-HCC patients treated with Atez/Bev.

| AEs in the Atez/Bev treatment
In this study, our findings demonstrated that the curative therapy group had a significantly higher RFS than the non-curative therapy group after Atez/Bev significant response. To our knowledge, this is the first study to show F I G U R E 4 Recurrence-free survival time after conversion therapy with curative or other therapies. The red, blue, and green line indicates the curative therapy, Atez/Bev discontinuation, and TACE group, respectively. Atez/Bev, atezolizumab plus bevacizumab; TACE, transarterial chemoembolization. the importance of curative conversion therapy after Atez/ Bev significant response. Recently, Xiao et al. reported that RFS was 75% at 12 months in HCC patients who underwent hepatectomy after treatment with tyrosine kinase inhibitor and anti-PD-1 Antibody combination therapy. 36 In addition, several reports have demonstrated that disease-free survival and OS were prolonged in patients who underwent liver resection or RFA as conversion therapy when tumors shrank after chemotherapy for colorectal cancer liver metastasis. [16][17][18][19] A previous study reported that the retreating liver metastasis leaves unresected, and more than half of patients with colorectal cancer will experience recurrence. 42 These results suggest that the tumor should be completely removed whenever curative conversion therapy can be performed because the tiny cluster of viable tumor cells left could become a source of relapse, and predicting pathological CR before resection is impossible. Therefore, patients with tumor responses evaluated via images might benefit from curative therapy of residual lesions to achieve long-term tumor-free survival.
The present study had several limitations. First, it was a retrospective study. Second, the study included approximately 40% of patients who were treated with Atez/Bev as second-line or later treatment. Third, HCC was not evaluated using The Liver Imaging Reporting and Data System. 43 Fourth, a sufficient observation period for patients with curative conversion therapy after Atez/Bev significant response was not assessed. Thus, future evidence from prospective real-world studies is needed to prove which conversion therapy is effective for patients with advanced HCC.
In conclusion, NLR <3 and m-ALBI grade 1 or 2a were significant predictive factors for therapeutic response and conversion therapy, leading to cancer-and drug-free patients with u-HCC treated with Atez/Bev. Moreover, curative conversion therapy may be important in achieving long-term RFS after Atez/Bev significant response.