Abstract
PURPOSE: Vulvar cancer (VC) is a rare malignancy with limited data in Southeast Asia. This 6-year retrospective cohort study aims to characterize temporal trends in VC cases in Indonesia, describe patients' sociodemographic, obstetric, and clinicopathological features, and identify predictors of advanced stage and distant metastasis (DM). PATIENTS AND METHODS: We analyzed 86 VC cases diagnosed between 2015 and 2020 at a tertiary hospital in Indonesia. Patients' sociodemographic and obstetric characteristics were described, and clinicopathological and treatment-related features were compared by stage and DM status. Temporal trends were assessed using Joinpoint regression, expressed as annual percentage change (APC). Subgroup analyses were conducted by age (<60 vs ≥60 years), residential status (rural vs urban), and DM presence (negative vs positive). Univariate and multivariate logistic regression identified predictors of advanced-stage VC and DM, with model performance evaluated using the area under the receiver operating characteristic curve (AUC). RESULTS: The number of VC cases increased significantly over six years (APC +21.43%, p=0.034), with the highest growth observed among older patients, rural residents, and cases without DM (+25.53% vs +19.61%; +22.25% vs +19.08%; +30.50% vs +9.95%, respectively). Multivariate analysis identified urban residence (OR 5.12), right-sided lesion (OR 7.93), bilateral lesions (OR 10.25), and tumor volume ≥70 cm³ (OR 10.88) as independent predictors of advanced-stage VC (AUC 0.81, p<0.001). Predictors of DM included normal/underweight nutritional status (OR 5.50), presence of pain (OR 4.88), right-sided lesion (OR 13.80), and non-keratinized tumor subtype (OR 5.72) (AUC 0.80, p<0.001). CONCLUSION: The experience from an Indonesian referral hospital shows that the number of VC cases treated is rising, particularly among older adults, rural populations, and patients without DM. Specific clinicopathological features, including tumor laterality, bilaterality, volume, nutritional status, pain, and histological subtype, predict adverse outcomes, providing a foundation for improved risk stratification, early detection, and targeted management in underrepresented populations.