Otolaryngology (ENT)

Chemokines in head and neck cancer: heterogeneous diagnostic and prognostic signals

Meta-analysis of 44 studies (7294 participants) found some chemokines associated with outcomes, but results were inconsistent across biomarkers and settings. MIP-3α was the most consistent prognostic marker, linking to poorer survival outcomes. Salivary IL-8 and chemerin showed promise for diagnosing oral squamous cell carcinoma, but supporting evidence was limited. Heterogeneity across tumor sites, sample types, and methods prevents immediate clinical adoption; larger standardized validation studies are needed. 1

Phase 2 monotherapy trials offer limited therapeutic value compared with phase 3

Pooled data from 130 phase 2 and 52 phase 3 trial arms showed objective response rates of 7% versus 24%, respectively. Median PFS and OS were shorter in phase 2 versus phase 3 (3.23 vs 5.43 months; 9.46 vs 14.44 months). Drug-related grade 3–4 adverse events were 30% in phase 2 versus 25% in phase 3. Interpret trials as primarily research opportunities; do not assume phase 2 monotherapy confers meaningful therapeutic benefit without clear supporting data. 2

Run-in phases or single-arm designs can shrink GA trial sample sizes

GA enlargement rates are stable within eyes and can be leveraged to reduce trial sample sizes using run-in or single-arm designs. In simulations, a 2‑year RCT required 156 participants versus 82, 40, and 26 with 3, 6, and 9‑month run-ins respectively. Single-arm designs with a run-in yielded even smaller sample size estimates than RCTs in comparable scenarios. These design efficiencies apply to GA trials measuring area change, but they rely on stable natural-history measurements and require careful bias control. 3