ESMO Congress 2024 | OncologyPRO

Abstract 1158P

Background

There is no generally accepted model to predict overall survival (OS) of patients treated with peptide receptor radionuclide therapy (PRRT) for small intestinal neuroendocrine tumors (siNET).

Methods

We searched literature for non-experimental clinical and biochemical predictors of OS after treatment with PRRT. Among 13 potential markers related to patient baseline status, function of organs affected by PRRT toxicity, inflammation, tumor burden and tumor proliferation, seven candidate variables were selected based on previous publication results, and analyzed in a retrospective cohort of 326 patients treated with PRRT between 2005 and 2019 in a tertiary referral center. An accelerated failure-time model was used and a nomogram was constructed. The model was internally validated with bootstrapping, and its discrimination and calibration evaluated with Harrell’s c-index and calibration plots.

Results

During a median follow-up of 40 months, there were 230 deaths. All selected variables were significantly associated with OS in unadjusted analysis. A prognostic nomogram was created using age (0-68) and performance status (PS, 0-44), tumor markers chromogranin A (CgA, log, 0-100) and 5-hydroxyindoloacetaic acid (5HIAA, 0-57), albumin (0-15), alkaline phosphatase (ALP, log, 0-20) and hemoglobin (Hb, 0-32). The full model had good discriminatory ability (optimism-corrected c-index 0.74) and calibration. A backward selection algorithm was used to approximate the full model. The final model contained age, PS, CgA, 5HIAA and Hb.

Conclusions

We suggest a novel nomogram, based on a small number of readily available clinical and biochemical parameters, to predict OS after treatment with PRRT for siNET. Our nomogram may assist clinicians in more accurate clinical decision-making and patient counseling.