Not All Child‐Pugh B Is the Same: Refining Immunotherapy Eligibility in Hepatocellular Carcinoma

The therapeutic landscape of unresectable hepatocellular carcinoma (HCC) has been transformed by immune checkpoint inhibitor-based combination regimens. Following the IMbrave150 trial, atezolizumab plus bevacizumab (Atez/Bev) has become the preferred first-line systemic therapy for patients with preserved liver function 1. However, this paradigm has been built almost exclusively on evidence derived from patients with Child-Pugh class A cirrhosis, leaving a substantial proportion of real-world patients, that is, those with Child-Pugh class B (CP-B), in a persistent evidence gap. In daily practice, clinicians are frequently confronted with the dilemma of whether to offer active systemic therapy to patients with CP-B liver function or to default to best supportive care. CP-B is traditionally viewed as a relative contraindication to immunotherapy, largely due to concerns regarding safety (e.g., bleeding risk) and limited survival benefit 2, 3. Yet CP-B represents a heterogeneous population, and treating it as a uniform category may oversimplify a far more nuanced clinical reality. In this context, the multicenter retrospective study by Sasaki and colleagues 4 provides timely and clinically relevant insights. By analysing nearly 800 patients treated with Atez/Bev across 10 Japanese institutions, including a sizable cohort of CP-B patients, the authors address a question that randomised trials have largely avoided: ‘which patients with CP-B liver function may still benefit from first-line immunotherapy-based combination therapy?’ Consistent with prior real-world reports, the study confirms that patients with CP-B experience shorter progression-free and overall survival compared with those with CP-A 5, 6. Of note, the safety profile of Atez/Bev was broadly comparable between the two groups, with similar rates of treatment discontinuation and immune-related adverse events. While bleeding-related events were more frequent in CP-B patients, this excess risk was attenuated when analysis was restricted to first-line therapy, reinforcing the importance of patient selection and treatment timing. The most compelling contribution of this study lies not in reaffirming that CP-B outcomes are inferior, but in demonstrating that CP-B outcomes are not uniform. Notably, the Child-Pugh score itself failed to meaningfully stratify prognosis within CP-B. In contrast, the modified albumin–bilirubin (mALBI) grade clearly separated patients with substantially different outcomes. Specifically, patients with CP-B and mALBI grade ≤ 2b achieved progression-free and overall survival that compare favourably with previously reported outcomes of systemic therapy in this population, whereas those with mALBI grade 3 had dismal outcomes, which is similar to best supportive care. These findings suggest that hepatic functional reserve, rather than Child-Pugh class alone, should guide treatment decisions. The mALBI grade appears to offer a more granular and clinically actionable assessment of liver function in this setting. Rather than viewing CP-B as a binary exclusion criterion, this study supports a more refined approach in which selected CP-B patients, for example, those with mALBI grade ≤ 2b, may be reasonable candidates for Atez/Bev. Conversely, patients with mALBI grade 3 appear unlikely to derive meaningful benefit and may be better served by supportive care or enrollment in clinical trials. The findings also raise important considerations regarding regimen selection. Bevacizumab-containing therapy carries an inherent risk of bleeding, particularly in patients with portal hypertension 1, 2. Although the present study suggests acceptable safety in carefully selected CP-B patients, it also highlights the need for vigilant risk assessment, especially in later lines of therapy. Whether VEGF-free immunotherapy regimens might offer a safer alternative for patients with marginal hepatic reserve remains an open question. The current study by Sasaki et al. 4 underscores a simple but important message: CP-B is not a monolith. By shifting the focus from categorical classification to functional stratification, this study offers a pragmatic framework for extending immunotherapy-based treatment to carefully selected patients who might otherwise be denied active therapy. In doing so, it moves the field one step closer to truly individualised treatment for advanced HCC. Given the inherent challenges associated with conducting prospective trials in patients with CP-B liver function, high-quality real-world data such as this can provide critical guidance for an underserved patient population. The author declares no conflicts of interest. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.