Hepatocellular carcinoma (HCC) is the sixth most prevalent neoplasm and causes the third most cancer-related fatalities globally, accounting for an estimated 900,000 new cases and 830,000 deaths in 2020 (1,2). Recent advancements in systemic chemotherapy for advanced HCC, including the development of immune checkpoint inhibitors (ICIs) and molecular targeted agents, have markedly enhanced patient outcomes (3-7).

The IMbrave150 study demonstrated the efficacy of atezolizumab plus bevacizumab, monoclonal antibodies targeting programmed death ligand 1 (PD-L1) and vascular endothelial growth factor (VEGF), respectively. This combination therapy led to extended progression-free survival (PFS) and overall survival (OS) in comparison to sorafenib for patients with advanced HCC (3). Consequently, atezolizumab plus bevacizumab is emerging as the primary systemic chemotherapy for advanced HCC.

ICIs stimulate an intensified T-cell-mediated response to facilitate the eradication of tumor cells. However, this heightened immune activity can also provoke a broad spectrum of adverse effects termed immune-related adverse events (irAEs). These irAEs can impact diverse organs, including the skin, gastrointestinal tract, liver, respiratory system, thyroid, and pituitary glands (8). The importance of irAEs is substantial as they are prevalent and often cause severe complications that can profoundly affect the quality of life and prognosis of patients undergoing ICI therapy. Moreover, finding the optimal approach to manage irAEs without compromising the antitumor response and long-term survival remains a challenge (9). Interestingly, patients experiencing irAEs tend to have a more favorable cancer-related prognosis (10-12). Thus, evaluating the individual risk of toxicity in advance is pivotal to guide the early management of irAEs. This proactive approach can enable the continued administration of ICIs to patients who are vulnerable to immune-mediated complications and, paradoxically, allow them to benefit more from treatment. As the use of ICIs continues to expand in the field of oncology, there is a growing need for dependable and validated biomarkers capable of predicting irAEs.

Regrettably, our understanding of predictive factors for irAEs in patients with HCC is limited. Therefore, we initiated an investigation to identify predictive factors for irAEs in patients with HCC undergoing treatment with atezolizumab plus bevacizumab.

Patients. In this single-center, prospective study conducted at the Aso Iizuka Hospital from December 2020 to June 2023, we enrolled patients with HCC who were administered atezolizumab plus bevacizumab. Of the 84 patients treated, 15 with a follow-up duration of less than 12 weeks were excluded, leaving 69 patients for evaluation. This study was conducted in accordance with the guidelines of the Declaration of Helsinki and approved by the Ethics Committee of Aso Iizuka Hospital (approval No. 22008). The opt-out method was used to obtain consent for this study.

Definition of irAEs. Adverse events perceived to have a possible immunological root, necessitating enhanced surveillance and potential intervention with immune suppressants or endocrine therapies, were categorized as irAEs. They were rated based on the National Cancer Institute Common Terminology Criteria for Adverse Events (version 5.0).

Biomarker analysis. The neutrophil–to-lymphocyte ratio (NLR) was determined by the quotient of the total neutrophils to total lymphocytes in peripheral blood, whereas the platelet-to-lymphocyte ratio (PLR) was ascertained by dividing the total platelets by the total lymphocytes.

Treatment protocol. Following the IMbrave150 trial protocol (3), patients were intravenously administered atezolizumab (1,200 mg) and bevacizumab (7.5 mg/kg) from Chugai (Tokyo, Japan) every three weeks. Treatment persisted until disease aggravation (PD) or the emergence of insufferable adverse events.

Evaluation of efficacy. Every 6 to 12 weeks following the start of treatment, computed tomography or magnetic resonance imaging were used to assess the treatment response. The antitumor response was gauged using modified RECIST version 1.1 (13). Indications of disease control rate (DCR) included complete response (CR), partial response (PR), and stable disease (SD) that persisted for a minimum of four months. The summation of PR and CR defined the objective response rate (ORR). Every three weeks, patients underwent evaluations and received treatment until either disease escalation (PD) or adverse reactions became intolerable.

Definitions of viral and non-viral infection. Patients with either anti-HCV antibody (Ab) or HBs-antigen (Ag) presence were classified as having a viral infection. Conversely, those negative for both serum anti-HCV Ab and HBs-Ag were categorized as having non-viral infections.

Statistical analysis. For all data processing, JMP Pro version 11 (SAS Institute, Cary, NC, USA) was used. The data are presented as median values with interquartile ranges. The Kaplan-Meier method, log-rank test, receiver operating characteristic (ROC) analysis, and Cox hazard analysis were the statistical tests applied. To determine predictive factors for irAE manifestation, ROC and area under the curve (AUC) values were computed. Differences in groups were assessed using the chi-squared test. All p-values were derived from two-tailed tests, and p<0.05 was considered to indicate a statistically significant difference.

Patient characteristics. The characteristics of the 69 patients who received atezolizumab plus bevacizumab are shown in Table I. Twelve patients (17.4%) experienced grade >2 irAEs, which included five cases of adrenal insufficiency, two cases of interstitial pneumonia, two cases of tubulointerstitial nephritis, one case of enterocolitis, one case of stomatitis, and one case of immune thrombocytopenia. We classified the enrolled patients into those who experienced irAEs (irAE group) and those who did not (non-irAE group). The median age of the entire cohort was 73.0 (65.5-72.5) years old, and 82.6% of the patients were male. There were 12, 30, and 27 patients categorized as HBs-Ag, HCV Ab, and non-viral, respectively. In the irAE group, two patients had disease in Barcelona Clinic Liver Cancer (BCLC) stage A, four in BCLC stage B, and six in BCLC stage C. In the non-irAE group, one patient had disease in BCLC stage A, 26 in BCLC stage B, and 33 in BCLC stage C. Age, sex, body mass index, Child–Pugh grade, albumin-bilirubin score (ALBI) score, tumor size, number of intrahepatic lesions, microvascular invasion, extrahepatic spread, serum α-fetoprotein levels, protein induced by vitamin K absence or antagonist-II (known as PIVKA-II) levels, and NLR were similar between the two groups at the start of treatment. At three weeks after the start of treatment, patients in the irAE group had a lower NLR than those in the non-irAE group (p=0.019). There were more non-viral patients in the irAE group than in the non-irAE group (p=0.033).

Prognostic role of NLR for the development of irAEs. We determined the ROC values for the development of irAEs in the 69 patients based on multiple parameters, and set the cut-off values for the parameters (Table II). The cut-off value for NLR <2.04 at three weeks after the start of treatment was 2.04 (AUC=0.77).

Univariate logistic regression was performed to assess the risk factors for irAEs (Table III). Subsequent multivariate analysis revealed that NLR at three weeks (HR=0.23; p=0.037) and non-viral infection (HR=4.47; p=0.037) were independent factors that may contribute to irAE occurrence (Table III).

Efficacy of the association between irAEs and antitumor response. Among patients who received atezolizumab plus bevacizumab, the ORR (CR+PR) was 9/12 (75.0%) in the irAE group and 15/57 (28.1%) in the non-irAE group (p=0.005). The DCRs (CR+PR+SD) were 11/12 (91.6%) and 30/57 (52.6%) in the irAE and non-irAE groups, respectively (p=0.016) (Table IV). Therefore, there was a higher response rate among patients in the irAE group than in those in the non-irAE group following atezolizumab plus bevacizumab therapy.

Efficacy of the association between irAEs and PFS. The median PFS of all patients who received atezolizumab plus bevacizumab therapy was 6.8 months. Kaplan-Meier analysis revealed that the median PFS in the irAE group (12.1 months) was higher than that in the non-irAE group (6.0 months) (p=0.010) (Figure 1).

In patients with advanced HCC, systemic therapy with a blend of ICI and VEGF inhibitors, notably atezolizumab and bevacizumab, is now endorsed (3). Therefore, it is imperative to evaluate irAE risks prior to treatment initiation. Recent research has focused on identifying predictive biomarkers for irAEs (14). The deployment of blood cell metrics for early irAE detection is gaining traction among medical professionals due to its accessibility, cost-efficiency, and straightforward interpretation. Current data, although not uniform, indicate that certain baseline blood cell counts and their increment during monitoring sessions correlate with increased irAE risks (15-34). Remarkably, no prior studies have delved into predictive determinants of irAEs within the HCC demographic. In our research, an NLR<2.04 at three weeks after the start of treatment manifested substantial predictive capability for patients with HCC undergoing atezolizumab and bevacizumab treatment.

The NLR is indicative of the cancer-induced inflammatory reaction and has correlations with patient outcomes and ICI treatment responsiveness across diverse cancer types (35-40). Persano et al. reported the prognositc index including NLR in HCC patients being treated with atezolizumab and bevacizumab (41). Eso et al. analyzed 40 patients with HCC receiving atezolizumab and bevacizumab and discovered significant NLR discrepancies between different response groups, with the most favorable responses showcasing the lowest NLR values (42). A concurrent multi-institutional study in Japan corroborated these findings, where survival outcomes varied markedly based on NLR levels (43). It has been reported that the NLR did not significantly worsen within 6 weeks after atezolizumab and bevacizumab initiation in HCC patients (44). In our study, Multivariate analysis demonstrated that NLR at three weeks was an independent predictor for OS (data not shown). In summary, NLR at three weeks was considered a vital predictor of irAEs and prognosis.

Recently, several studies have reported that irAEs were associated with the efficacy of ICI therapy in patients with HCC (45). Fukushima et al. reported that low-grade irAEs exhibited a strong association with better patient outcomes when treated with atezolizumab and bevacizumab (46). We observed that patients with irAEs had a better ORR and PFS compared to patients without irAEs. Although not statistically significant, OS tended to be longer in the irAE group compared to that in the non-irAE group (22.0 months vs. 16.2 months, p=0.10). This result can be attributed to the unmatched liver function at the start of chemotherapy due to the small number of cases.

Interestingly, our study revealed that liver etiology (non-viral) was another predictive marker associated with irAEs. Hatanaka et al. reported no significant differences in efficacy, and all treatment-related adverse events, including irAEs, were found between patients with non-viral infection and those with viral infections (47). There have been no previous reports on the association between irAEs and patients with non-viral liver etiology. Nucleoside analogs have been associated with significant improvement in liver inflammation and fibrosis in chronic HBV patients (48). Furthermore, direct-acting antivirals have been shown to reduce fibrosis in chronic HCV patients (49). In our study, most HBV and HCV patients were treated with these drugs, potentially leading to suppressed inflammation. Non-viral patients may have non-viral factors, such as alcohol consumption or obesity, which contribute to pre-existing inflammatory activity and susceptibility to an acute inflammatory response, further heightened by ICIs. More extensive studies are necessary to analyze the relationship between irAEs and liver etiology.

Limitations of this study include the small number of patients with HCC due to its single-center nature. Our results suggest that a lower NLR might enable a more efficient anti-tumor immune response, but also indicate a higher risk of immune-related toxicity.

An NLR <2.04 at three weeks after the start of treatment may be a biomarker for predicting irAEs in patients with HCC treated with atezolizumab plus bevacizumab therapy. Further studies are needed to predict and validate biomarkers for irAEs.

All Authors declare no competing interests in relation to this study.

A.K., M.Y., A.M., and K.M. designed the study. A.K., Y.K, S.N., K.T., and M.Y. assisted with the data analyses. A.K. wrote the initial draft of the manuscript. M.Y. contributed to the analysis and interpretation of the data. M.Y., A.M., and K.M. assisted in the preparation and critical review of the manuscript. All Authors approved the final version of the manuscript and agreed to be accountable for all aspects of the work.

The Authors are grateful to Y. Ishibashi for his assistance with manuscript preparation. The Authors thank Amanda Holland, Ph.D., from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.