Peripheral eosinophilia as a potential biomarker of response and immune-related adverse events in patients with gastrointestinal tumors treated with immune checkpoint inhibitors: A single-center experience.

Abstract

e15697 Background: Peripheral eosinophilia (PEBC), defined as an absolute eosinophil count (AEC) > 0.5 × 10⁹/L, has emerged as a biomarker in solid tumors, including gastrointestinal (GI) malignancies. While eosinophilia is well-documented in hematologic cancers, its role in solid tumors remains unclear. Evidence suggests increased eosinophil levels correlate with favorable responses to immune checkpoint inhibitors (ICIs) and may influence immune-related adverse events (irAEs), but its predictive value requires further investigation. Methods: This single-center observational study includes retrospective and prospective cohorts of 63 patients with GI tumors treated with ICIs from 2019 to 2024. All patients received at least one cycle of treatment, with PEBC monitored during the first six months. PEBC was assessed to determine its association with treatment response and irAEs incidence/severity. Clinical outcomes, including progression-free survival (PFS) and overall survival (OS), will be analyzed using R software (v4.4.2), with statistical significance set at p < 0.05. Results: Between March 2019 and November 2024, 63 patients with gastrointestinal tumors were included (median age: 58.8 years, 40 male (63.4%). The most common tumor sites were colon 26 (41.2%), esophagus 12 (19%), and stomach 12 (19%). Mismatch repair deficiency (dMMR) was observed in 33 (52.3%) with loss MLH1/PMS2 17(26.9%), and BRAF V600E mutations detected in 2 (3.1%). Eosinophilia ( > 500/μL) was present in 17 (26.9%), with a median peak eosinophil count of 520.7/μL (range: 0–2160/μL). At the time of ICIs treatment, no patient was receiving corticosteroids at doses equivalent to ≥10 mg prednisolone daily. ICIs were used as first-line therapy in 46 (73%), with 30 (47.6%) receiving monotherapy and 33 (52.4%) combination regimens. The overall response rate was 50 (79.8%), and 44 (69.8%) of patients remained alive at the time of analysis. irAEs occurred in 37 (58.7%), with 13 (20.6%) Grade 1, 13 (20.6%) Grade 2, 8 (12.6%) Grade 3, and 2 (3.2%) Grade 4. Higher eosinophil levels correlated with response (mean PEBC: 588.8 vs. 282.6; p = 0.0003) and survival (mean PEBC: 613.0 vs. 307.1; p = 0.0005). Patients with irAEs had lower progression risk (HR = 0.42, 95% CI: 0.25–0.73, p = 0.0018). Cox regression showed responders had improved OS (HR = 3.75, 95% CI: 1.81–7.76, p = 0.0004), while irAEs reduced mortality risk by 43.2% (HR = 0.57, 95% CI: 0.33–0.96, p = 0.035). Conclusions: This study, conducted in a public hospital in Argentina, suggests that peripheral eosinophilia is associated with improved treatment response and increased irAEs in GI cancer patients receiving ICIs. Elevated eosinophil counts correlated with better survival and response rates, while also being linked to higher irAEs. These findings support eosinophilia as a cost-effective biomarker for efficacy and toxicity prediction in immunotherapy. Further research is warranted to elucidate the underlying mechanisms and validate its predictive value in larger, multi-center studies.