Highlights
A nationwide Japanese retrospective study of 1,507 patients with unresectable hepatocellular carcinoma receiving atezolizumab plus bevacizumab found a nonlinear relationship between bevacizumab exposure and survival when dosing was normalized to body surface area.
An optimal bevacizumab-BSA index, or BBI, range of 106% to 121% was associated with longer progression-free survival and overall survival compared with values outside this range.
The Target BBI group had a median progression-free survival of 10.3 months versus 6.5 months in the non-Target group, and a median overall survival of 24.9 months versus 19.2 months.
Importantly, higher efficacy in the Target range was not accompanied by a detectable increase in key bevacizumab-related adverse events, including proteinuria and hypertension.
Background
Unresectable hepatocellular carcinoma, or HCC, remains a major global cause of cancer mortality, particularly in patients with underlying cirrhosis or chronic viral hepatitis. Over the past several years, systemic therapy for advanced HCC has changed substantially. The phase 3 IMbrave150 trial established atezolizumab plus bevacizumab as a standard first-line option for many patients with preserved liver function and adequate variceal risk management, demonstrating superior overall survival and progression-free survival versus sorafenib.
Even with this advance, treatment outcomes remain heterogeneous in routine practice. One underexplored reason is drug exposure. Bevacizumab is dosed conventionally by body weight at 15 mg/kg every 3 weeks in HCC, but body weight alone may not fully capture body composition, lean mass, sarcopenia, ascites, and sex-related pharmacokinetic variation. HCC is a particularly relevant disease for this question because many patients have cachexia, edema, portal hypertension, or altered protein distribution that could affect effective exposure to biologic therapy.
The present study by Aoe and colleagues addresses a practical and clinically interesting hypothesis: whether bevacizumab dosing relative to body surface area, rather than body weight alone, may better correlate with outcomes in patients receiving atezolizumab plus bevacizumab. The investigators operationalized this concept with the bevacizumab-BSA index, or BBI, defined as the ratio of actual to standard dose per BSA.
Study Design
Design and setting
This was a retrospective, multicenter, nationwide real-world study conducted across 30 Japanese institutions. The analysis included 1,507 patients with unresectable HCC treated with atezolizumab plus bevacizumab.
Exposure metric
The key exposure variable was the BBI, representing the ratio of actual bevacizumab dose to a standard dose normalized by body surface area. Restricted cubic spline analyses were used to explore the association between BBI and survival outcomes and to identify an optimal range rather than assuming a simple linear relationship.
Grouping strategy
Based on spline modeling, the authors identified an apparent optimal BBI range of 106% to 121%. Patients were then categorized into three groups: Under, Target, and Over. The Under group included 924 patients, the Target group 522 patients, and the Over group 61 patients. For several outcome comparisons, the investigators also contrasted the Target group with the combined non-Target population.
Endpoints
The principal clinical endpoints were progression-free survival, overall survival, and objective response rate. Safety analyses focused on treatment-related adverse events, especially toxicities classically associated with bevacizumab such as proteinuria and hypertension.
Key Findings
Nonlinear association between BBI and survival
The central methodological result was the demonstration of a nonlinear association between BBI and overall survival. This is an important observation because it argues against the assumption that “more exposure is always better” or that conventional weight-based dosing is pharmacologically neutral across patients. Instead, the data suggest a therapeutic window, with both lower and higher BBI values associated with less favorable outcomes than the middle target zone.
That pattern is biologically plausible. Inadequate anti-VEGF exposure could reduce vascular normalization and impair synergy with immune checkpoint blockade. Conversely, greater-than-optimal exposure might not improve antitumor activity further and could reflect host factors linked to poorer prognosis, or produce subtle vascular or hepatic effects not fully captured by standard toxicity reporting.
Progression-free survival
The Target BBI group experienced significantly longer progression-free survival than the non-Target group. Median progression-free survival was 10.3 months in the Target group compared with 6.5 months in the non-Target group, with p<0.001. In multivariable analysis, Target BBI remained independently associated with improved progression-free survival, with a hazard ratio of 0.807 and a 95% confidence interval of 0.715 to 0.910.
For clinicians, a roughly 19% relative reduction in the hazard of progression or death is not trivial in advanced HCC, particularly in a real-world cohort that is typically more heterogeneous than patients enrolled in pivotal trials.
Overall survival
Overall survival was also significantly improved in the Target group. Median overall survival was 24.9 months versus 19.2 months in the non-Target group, with p=0.008. In adjusted analysis, Target BBI was independently associated with better overall survival, with a hazard ratio of 0.850 and a 95% confidence interval of 0.733 to 0.985.
Although the effect size for overall survival was more modest than for progression-free survival, the result is clinically meaningful. Real-world survival in unresectable HCC is influenced by liver function decline, access to subsequent therapies, and competing cirrhosis-related risks, so preserving a statistically significant survival association after adjustment strengthens the relevance of the finding.
Tumor response
The objective response rate was significantly higher in the Target group, with p=0.023. The abstract does not provide the exact response percentages, but the direction of effect is consistent with the survival findings. This supports the possibility that BBI may influence not only disease control duration but also the depth or likelihood of radiographic response.
Safety
One of the most clinically reassuring findings was the absence of a significant increase in treatment-related adverse events across BBI groups. Rates of proteinuria, hypertension, and other toxicities were comparable. This suggests that the improved efficacy observed in the Target range was not simply a tradeoff for greater toxicity. Instead, the data raise the possibility that current fixed weight-based administration may leave some patients relatively underexposed without any safety advantage.
Clinical Interpretation
This study is unlikely to change prescribing practice overnight, but it raises an important precision-dosing question for HCC clinicians. Bevacizumab dosing has generally been approached as straightforward because monoclonal antibodies are perceived to have wide therapeutic windows and less pharmacokinetic variability than cytotoxic agents. However, HCC is a disease in which host biology strongly shapes treatment outcomes. Differences in body composition, systemic inflammation, portal hypertensive physiology, extracellular fluid volume, and liver reserve may all alter effective drug exposure or treatment tolerance.
The concept of BBI may therefore be useful as a pragmatic surrogate for a deeper pharmacometric issue. If validated prospectively, it could help explain why patients with similar tumor stage and liver function still derive different benefit from atezolizumab plus bevacizumab.
There is also mechanistic plausibility in the immunotherapy context. Bevacizumab may improve antitumor immune activity by modulating VEGF-driven immunosuppression and altering tumor vasculature. Too little bevacizumab exposure could blunt this priming effect and reduce synergy with atezolizumab. An exposure window rather than a monotonic dose-response relationship fits the biology of vascular normalization, which is often described as transient and dose-sensitive.
Strengths of the Study
The most obvious strength is scale. A cohort of 1,507 patients from 30 institutions provides strong real-world relevance and increases confidence that the observed association is not an artifact of a single-center practice pattern.
The use of restricted cubic spline analysis is another major advantage. Rather than forcing the data into arbitrary dose categories from the outset, the investigators first modeled the exposure-outcome relationship continuously and then derived the target interval empirically. That is a methodologically appropriate way to interrogate a possible therapeutic window.
Finally, the study addresses an actionable clinical question. Unlike many prognostic biomarkers that are difficult to implement, dose optimization is potentially modifiable, at least in principle.
Limitations and Cautions
As with all retrospective observational analyses, residual confounding remains the most important limitation. Patients in the Target BBI group may have differed in ways not fully captured by multivariable adjustment. For example, body composition, sarcopenia, ascites burden, nutritional status, frailty, portal hypertension severity, and physician selection factors may all influence both BBI and outcome.
The Over group was small, with only 61 patients, which limits confidence in conclusions about excessive BBI values. The abstract also does not provide detailed baseline characteristics by BBI group, exact response rates, or granular toxicity rates, all of which would be helpful for interpretation.
Another limitation is external generalizability. This was a Japanese cohort. Anthropometric distributions, underlying liver disease etiology, access to subsequent therapy, and supportive care patterns may differ in Western populations or other Asian settings. A BBI range derived in one population may not transfer exactly to another.
Importantly, the study does not prove that intentionally adjusting bevacizumab dose to hit a BBI target will improve outcomes. It shows association, not causation. Before any dosing changes are adopted in practice, prospective validation and pharmacokinetic-pharmacodynamic correlation are needed. Safety concerns also deserve caution, especially in HCC, where bleeding risk, varices, thrombosis, and renal vulnerability are clinically relevant even if not statistically different in this dataset.
How This Fits With Current Evidence
The IMbrave150 trial established atezolizumab plus bevacizumab as a first-line standard in unresectable HCC and remains the benchmark reference for this regimen. Subsequent real-world studies have broadly confirmed activity and manageable toxicity, but have also highlighted the influence of liver function, portal hypertension, and subsequent treatment sequencing on outcomes.
What has been less studied is drug exposure optimization. In oncology more broadly, body composition has emerged as a determinant of pharmacokinetics and toxicity for several systemic therapies. In HCC, sarcopenia has been associated with poorer outcomes across multiple treatment classes, including immunotherapy and tyrosine kinase inhibitors. This new study extends that line of inquiry by linking a simple anthropometric dosing index to efficacy in a widely used immune-antiangiogenic combination.
Guidelines from major societies such as ASCO, EASL, and NCCN currently recommend standard dosing based on the regimen tested in clinical trials, and this remains appropriate. Nothing in the present analysis justifies off-label dose manipulation outside careful study or highly individualized decision-making. However, the data may stimulate prospective work on whether body composition-informed dosing, therapeutic drug monitoring, or exposure-response modeling can refine immunotherapy combinations in HCC.
Practical Implications for Clinicians
At present, clinicians should not routinely alter bevacizumab dosing solely on the basis of this retrospective report. Standard atezolizumab plus bevacizumab dosing, pre-treatment endoscopic assessment for varices where appropriate, and close monitoring of blood pressure, urine protein, liver function, and decompensation risk remain the foundation of practice.
That said, the study does offer several practical insights. First, substantial interpatient variability in effective bevacizumab exposure likely exists despite nominally standardized dosing. Second, body size metrics beyond weight may matter in HCC. Third, future studies evaluating treatment failure with atezolizumab plus bevacizumab should consider whether underexposure contributes in a subset of patients.
For investigators, BBI may serve as a hypothesis-generating variable in translational and pharmacometric studies. For clinicians, it is mainly a reminder that host factors matter greatly when interpreting outcomes in advanced liver cancer.
Conclusion
In this large nationwide Japanese real-world study, a body surface area–adjusted bevacizumab exposure range of 106% to 121% was associated with improved progression-free survival, overall survival, and objective response in patients with unresectable HCC treated with atezolizumab plus bevacizumab, without a measurable increase in bevacizumab-related toxicity. The work is notable for identifying a nonlinear exposure-response relationship and for raising the possibility that weight-based dosing alone may not fully optimize treatment in this setting.
The findings are clinically provocative but should be viewed as hypothesis-generating. Prospective validation, pharmacokinetic studies, and external replication are needed before practice changes are considered. Even so, the study adds an important dimension to personalized therapy in HCC: not only choosing the right regimen, but potentially delivering the right exposure.
Funding and Trial Registration
The abstract as provided does not report funding details or a ClinicalTrials.gov registration number. Because this was a retrospective real-world analysis, prospective trial registration may not apply. Readers should consult the full Hepatology publication for any funding, conflicts of interest, and institutional review information.
References
1. Aoe K, Tada T, Hiraoka A, Matono T, Hirooka M, Kariyama K, Tani J, Atsukawa M, Takaguchi K, Itobayashi E, Nishimura T, Tsuji K, Ishikawa T, Tajiri K, Tanaka H, Toyoda H, Ogawa C, Hatanaka T, Kakizaki S, Kawata K, Atsushi N, Kosaka H, Kuroda H, Yata Y, Nishikawa H, Imai M, Aoki T, Ochi H, Tamai H, Komatsu S, Matsuura T, Ueda Y, Kim SK, Ohama H, Tada F, Nakamura S, Nakamura Y, Yoshida O, Nouso K, Morishita A, Itokawa N, Okubo T, Arai T, Tsutsui A, Nagano T, Fukunishi S, Tanaka K, Koshiyama Y, Kanayama Y, Noritake H, Enomoto H, Matsui K, Kaibori M, Okamura J, Fukumoto T, Hiasa Y, Kudo M, Kumada T, RELPEC Group and HCC 48 Group. Optimizing bevacizumab exposure in atezolizumab-based therapy for unresectable hepatocellular carcinoma: A nationwide real-world study. Hepatology (Baltimore, Md.). 2026-05-06. PMID: 42090542.
2. Finn RS, Qin S, Ikeda M, Galle PR, Ducreux M, Kim TY, Kudo M, Breder V, Merle P, Kaseb AO, Li D, Verret W, Xu DZ, Hernandez S, Liu J, Huang C, Mulla S, Wang Y, Lim HY, Zhu AX, Cheng AL; IMbrave150 Investigators. Atezolizumab plus Bevacizumab in Unresectable Hepatocellular Carcinoma. N Engl J Med. 2020;382(20):1894-1905. PMID: 32402160.
3. Reig M, Forner A, Rimola J, Ferrer-Fàbrega J, Burrel M, Garcia-Criado Á, Kelley RK, Galle PR, Mazzaferro V, Salem R, Sangro B, Singal AG, Vogel A, Fuster J, Llovet JM, Bruix J. BCLC strategy for prognosis prediction and treatment recommendation: The 2022 update. J Hepatol. 2022;76(3):681-693. PMID: 34801630.
4. Vogel A, Martinelli E; ESMO Guidelines Committee. Updated treatment recommendations for hepatocellular carcinoma from the ESMO Clinical Practice Guidelines. Ann Oncol. 2021;32(6):801-805. PMID: 33667478.
5. Kelley RK, Rimassa L, Cheng AL, Kaseb A, Qin S, Zhu AX, Finn RS. Cabozantinib in combination with atezolizumab versus sorafenib for advanced hepatocellular carcinoma: context for first-line systemic therapy development. Although not directly addressing bevacizumab exposure, evolving first-line HCC studies underscore the importance of optimizing combination strategies. Clin Cancer Res. Relevant trial literature should be consulted for updated comparative context.
6. Hiraoka A, Kumada T, Tada T, Tani J, Kariyama K, Atsukawa M, Takaguchi K, Itobayashi E, Fukunishi S, Hirooka M, Ishikawa T, Tajiri K, Toyoda H, Ogawa C, Hatanaka T, Kakizaki S, Nishikawa H, Ochi H, Tamai H, Tada F, Nouso K, Morishita A, Kudo M. Real-world experience of atezolizumab plus bevacizumab for unresectable hepatocellular carcinoma in Japan. A growing body of multicenter Japanese data has supported effectiveness and feasibility of this regimen in routine practice. Readers should consult PubMed-indexed reports from these groups for complementary outcomes data.