Abstract / Summary
Sarcopenia, characterized by loss of skeletal muscle mass and function, has emerged as a potential prognostic biomarker in oncology. Its role in patients with head and neck squamous cell carcinoma (HNSCC) treated with immune checkpoint inhibitors (ICIs) remains inadequately defined. A systematic review and meta-analysis was conducted in accordance with PRISMA 2020. PubMed, Scopus, and the Cochrane Library were searched from inception to March 2026. Observational studies reporting baseline computed tomography (CT)-defined sarcopenia, based on a dichotomized skeletal muscle index (SMI) or skeletal muscle area (SMA), in adults with HNSCC receiving ICIs were included. Studies using non-radiological assessment, continuous-only muscle indices, exposure definitions incorporating on-treatment change, or populations restricted to progression after ICI therapy were excluded, and cohorts were screened for patient overlap. Hazard ratios (HRs) for overall survival (OS) and progression-free survival (PFS) were pooled using a random-effects model with the Hartung-Knapp-Sidik-Jonkman (HKSJ) method incorporating the truncated variance correction. Risk of bias was assessed with the QUIPS tool and certainty of evidence with GRADE as adapted for prognostic factor reviews. Three retrospective cohort studies (354 patients) met the eligibility criteria. Under the pre-specified primary estimator the pooled hazard ratio for OS was 2.05 (95% CI 1.00-4.20), an interval that marginally includes the null (common-effect HR 2.05, 95% CI 1.48-2.84; I2 = 0%); for PFS it was 1.85 (95% CI 1.05-3.27; I2 = 0%), excluding the null only narrowly. The wide HKSJ intervals reflect the small number of studies. The pooled OS estimate is closely concordant with an independent published synthesis of the same question (HR 2.05). Certainty of evidence was rated low for both outcomes. Baseline CT-defined sarcopenia was associated with an approximately two-fold increase in the hazard of death, and of progression or death, in patients with HNSCC treated with ICIs. The primary OS analysis did not reach conventional statistical significance: under the pre-specified HKSJ estimator its 95% confidence interval marginally included the null, and the PFS interval excluded the null only narrowly. The association therefore rests on the size and consistency of the point estimates, the absence of heterogeneity, the common-effect interval and concordance with an independent published synthesis, rather than on the primary interval alone. The evidence base comprises three retrospective cohorts and certainty is low. Sarcopenia should be regarded as a prognostic rather than a predictive marker in this setting and must not be used to justify withholding immunotherapy. Prospective studies with standardized CT-based definitions are required.