Prognostic association of a SEER-derived HIPEC-related treatment proxy with cancer-specific survival in surgically treated appendiceal epithelial malignancies: a SEER-based study

Appendiceal epithelial malignancies are rare and heterogeneous. Confirmed hyperthermic intraperitoneal chemotherapy (HIPEC) is not directly recorded in the Surveillance, Epidemiology, and End Results (SEER) database. We constructed a SEER-derived HIPEC-related treatment proxy and evaluated its association with cancer-specific survival and its incremental prognostic information in surgically treated patients. We included 5,520 adults diagnosed in 2007–2017 from SEER and divided them into training ( n = 3,863) and internal validation ( n = 1,657) cohorts. HIPEC-proxy positivity required recorded intraoperative systemic therapy and chemotherapy. Cause-specific Cox analyses were used to evaluate prognostic associations, and a model based on conventional clinicopathologic variables was compared with a model additionally including the proxy. Performance at 1, 3, and 5 years was characterized using discrimination, calibration, Brier score, and IDI. Both models were externally evaluated without refitting in 209 institutional patients. Internal validation AUCs at 1, 3, and 5 years were 0.815, 0.810, and 0.782 for the HIPEC-proxy model versus 0.812, 0.808, and 0.780 for the base model, indicating useful prognostic discrimination but minimal incremental gain from the proxy; paired differences were not statistically significant. In the external cohort, corresponding AUCs were 0.775, 0.797, and 0.783 versus 0.753, 0.759, and 0.751. The 3- and 5-year paired AUC differences were modest but statistically supported, whereas the 1-year estimate remained imprecise. External calibration was generally acceptable, with some overprediction at longer horizons. Sensitivity analyses addressing proportional-hazards violations yielded similar proxy estimates. In this retrospective surgically treated cohort, the SEER-derived HIPEC-related proxy was associated with cancer-specific prognosis and provided minimal incremental discrimination internally and modest improvement at 3 and 5 years externally. The proxy should not be interpreted as confirmed HIPEC exposure, evidence of treatment benefit, or a basis for individual postoperative prediction or treatment selection. These findings support further evaluation of the proxy as prognostic treatment-pattern information in datasets with more detailed treatment characterization.