3781 - A Patient Vulnerability Score Predicts Radiation Treatment Modification and Survival in Gynecologic External Beam Radiation Therapy

Purpose/Objective(s): Radiation treatment completion is a critical determinant of outcomes in gynecologic malignancies. Identifying patients at risk for treatment modification prior to initiating therapy remains an unmet need. We developed a Patient Vulnerability Score (PVS) and evaluated its association with EBRT treatment modification and mortality in women with gynecologic cancers.

Materials/Methods: We conducted a retrospective cohort study of 186 patients who received EBRT for gynecologic malignancies at our institution from 2018 to 2025. A PVS (theoretical range 5–19; observed 5–16) was constructed from five clinical domains: age at diagnosis (0–4 points), ECOG performance status (2–5 points), primary tumor site (1–3), systemic therapy type (1–5), and surgical status (1–2). Treatment-related variables were excluded to avoid confounding. The primary endpoints were treatment modification (delivery of <100% of planned fractions) and mortality. Patients were stratified into low (PVS 5–8, n=55), moderate (PVS 9–13, n=102), and high (PVS =14, n=29) risk groups. Associations were assessed using the Cochran-Armitage test for trend, chi-square tests, independent samples t-tests, odds ratios with 95% confidence intervals.

Results: Primary sites included uterine corpus (55.3%), cervix (32.3%), vulva (10.8%), and vagina (1.6%). Treatment modification rates increased across PVS strata: low 14.5%, moderate 24.5%, high 37.9% (Cochran-Armitage trend Z=2.40, p=0.016). High-risk patients had significantly greater odds of modification versus low-risk (OR 3.59, 95% CI 1.24–10.37, p=0.018). Non-completion rates (<90% of planned fractions) also increased: 9.1%, 13.7%, 24.1% (high vs. low OR 3.18, 95% CI 0.91–11.12, p=0.070). Mortality increased across strata: 9.1%, 11.8%, 27.6% (Cochran-Armitage trend Z=2.12, p=0.034). High-risk patients demonstrated nearly 4-fold increased odds of death versus low-risk (OR 3.81, 95% CI 1.12–13.01, p=0.033). Deceased patients had significantly higher mean PVS than survivors (11.4 vs 10.0, p=0.019).

Conclusion: A five-domain PVS was significantly associated with treatment modification and mortality in women receiving EBRT for gynecologic cancers. High-risk patients demonstrated 3.6-fold and 3.8-fold increased odds of treatment modification and death, respectively. The significant linear trend across risk strata and higher continuous PVS among deceased patients supports a graded relationship between patient vulnerability and adverse outcomes. This pre-treatment tool may enable identification of vulnerable patients who would benefit from intensified supportive care. Prospective validation is warranted.