Prognostic impact of latency interval on survival in second primary malignancies after head and neck cancer: A population-based SEER study.

e18098 Background: Survivors of head and neck cancer (HNC) are at increased risk of developing second primary malignancies (SPMs), which substantially contribute to long-term mortality. However, the prognostic significance of the timing of non-synchronous SPM development remains poorly defined. We evaluated the association between latency interval and overall and cancer-specific mortality among HNC survivors with non-synchronous SPMs. Methods: Using the Surveillance, Epidemiology, and End Results (SEER) database, we identified adults diagnosed with primary squamous cell HNC who subsequently developed a non-synchronous SPM occurring ≥2 years after HNC diagnosis, a threshold chosen to minimize misclassification of synchronous disease and early recurrence. Latency intervals were categorized as 2–4 years, 5–9 years, and ≥10 years. Modified Poisson regression with robust variance was used to estimate unadjusted and adjusted relative risks (aRRs) for overall mortality and cancer-specific mortality, adjusting for age at HNC diagnosis, race/ethnicity, tumor grade, SPM histology and site, and receipt of surgery, chemotherapy, and radiation. Results: This analysis spans contemporary SEER treatment eras. Among 19,686 HNC survivors with non-synchronous SPMs, shorter latency intervals were independently associated with worse outcomes. Compared with latency ≥10 years, a latency of 2–4 years was associated with significantly higher overall mortality (aRR 1.30, 95% CI 1.26–1.33) and cancer-specific mortality (aRR 1.29, 95% CI 1.24–1.34), and a latency of 5–9 years also conferring increased risk (overall mortality aRR 1.14; cancer-specific mortality aRR 1.12). Older age, Black race (aRR 1.12), higher-grade HNC, and SPMs of the pancreas (aRR 1.66), esophagus (aRR 1.53), lung (aRR 1.39), and stomach (aRR 1.27) were significantly associated with increased overall mortality and cancer-specific mortality. Breast, prostate, melanoma, and kidney SPMs were associated with lower mortality risk. Conclusions: In this large population-based analysis, earlier development of non-synchronous SPMs following HNC was associated with a graded increase in both overall and cancer-specific mortality, independent of demographic, tumor, and treatment factors. Mortality risk is further modified by the anatomic site of the second primary malignancy, with pancreatic, esophageal, and lung cancers conferring the highest lethality, while prostate, melanoma, and breast cancers are associated with more favorable outcomes. These findings provide population-level validation of latency and SPM site as complementary markers of survivorship risk and support more individualized, risk-adapted post-treatment surveillance strategies.