External validation of a composite bio-humoral index in anal cancer patients undergoing concurrent chemoradiation

A prognostic scoring system based on laboratory inflammation parameters, [Hemo-Eosinophils-Inflammation (HEI) index], including baseline hemoglobin level, the systemic inflammatory index and eosinophil count was recently proposed in patients with squamous cell carcinoma of the anus (ASCC). HEI was shown to discriminate disease-free (DFS) and overall (OS) survival in ASCC patients treated with concurrent chemoradiation (CRT). We tested the accuracy of the model on a multicentric cohort for external validation.

Patients treated with CRT were enrolled. The Kaplan-Meier curves for DFS and OS based on HEI risk group were calculated and the log-rank test was used. Cox proportional hazards models were used to assess the prognostic factors for DFS and OS. The exponential of the regression coefficients provided an estimate of the hazard ratio (HR). For model discrimination, we determined Harrell's C-index, Gönen & Heller K Index and the explained variation on the log relative hazard scale.

A total of 877 patients was available. Proportional hazards were adjusted for age, gender, tumor-stage, and chemotherapy. Two-year DFS was 77 %(95 %CI:72.0-82.4) and 88.3 %(95 %CI:84.8-92.0 %) in the HEI high- and low- risk groups. Two-year OS was 87.8 %(95 %CI:83.7-92.0) and 94.2 %(95 %CI:91.5-97). Multivariate Cox proportional hazards model showed a HR = 2.02(95 %CI:1.25-3.26; p = 0.004) for the HEI high-risk group with respect to OS and a HR = 1.53(95 %CI:1.04-2.24; p = 0.029) for DFS. Harrel C-indexes were 0.68 and 0.66 in the validation dataset, for OS and DFS. Gonen-Heller K indexes were 0.67 and 0.71, respectively.

The HEI index proved to be a prognosticator in ASCC patients treated with CRT. Model discrimination in the external validation cohort was acceptable.