IL1B rs1143634 genotype and differential cancer-specific survival by neoadjuvant strategy in locally advanced rectal cancer: an exploratory analysis of a prospective randomized noninferiority trial

Abstract Background Locally advanced rectal cancer (LARC) treatment selection between neoadjuvant intensified chemotherapy (NIC) and standard chemoradiotherapy (CRT) lacks validated predictive biomarkers. Interleukin-1β (IL1B) polymorphisms modulate IL-1β expression and may influence treatment-specific outcomes through tumor microenvironment remodeling. Methods In this prospective, single-center, randomized, open-label noninferiority trial (NCT05378919), 150 patients with stage II–III rectal cancer were randomized 1:1 to NIC (FOLFOX4, 8 cycles; n = 68) or CRT (50 Gy/25 fractions with concurrent fluoropyrimidine; n = 82). The primary endpoint was 5-year disease-free survival (DFS; noninferiority margin HR ≤ 1.29, NIC vs. CRT). IL1B rs1143634 and rs16944 were genotyped by TaqMan assay. Treatment–genotype interactions were assessed using Cox regression with multiplicative interaction terms. Results Noninferiority was not demonstrated: 5-year DFS was 49.3% (NIC) versus 64.1% (CRT) in the ITT population (HR 1.52, 95% CI 0.90–2.56; p = 0.111), with the upper CI bound exceeding the noninferiority margin of 1.29. Per-protocol DFS was significantly inferior for NIC (57.9% vs. 79.2%; HR 2.26, 95% CI 1.06–4.79; p = 0.029). Overall survival and cancer-specific survival (CSS) did not differ significantly between arms. A nominally significant treatment–genotype interaction was identified for rs1143634 in CSS (interaction p = 0.041 by Wald test; likelihood-ratio p = 0.028): T-allele carriage (CT + TT) was associated with inferior CSS exclusively in the NIC arm (HR 4.84, 95% CI 1.04–22.43; p = 0.044), with no prognostic effect in the CRT arm (HR 0.69; p = 0.494). No interaction was detected for DFS ( p = 0.294). rs16944 showed no interaction (CSS p = 0.214; DFS p = 0.851). The CSS interaction did not survive correction for multiple testing. Conclusions NIC (FOLFOX4) did not achieve DFS noninferiority versus CRT in unselected LARC patients, although the trial was underpowered for this endpoint. IL1B rs1143634 demonstrates a treatment-arm-specific prognostic effect on cancer-specific survival, with a nominally significant treatment–genotype interaction ( p = 0.041) identified within a prospective randomized design; this exploratory finding does not survive correction for multiple testing, is based on a small number of cancer-specific events, and does not establish rs1143634 as a predictive biomarker. Prospective validation in an adequately powered, biomarker-stratified trial is warranted. Trial registration ClinicalTrials.gov, NCT05378919, registered 18 May 2022 (retrospectively registered). Study start: September 2015.