Abstract
BACKGROUND: Breast cancer is the most common cancer among women in India, and a substantial proportion present with metastatic disease, which worsens prognosis. Evidence on socio-demographic, clinical, and health system determinants of metastasis remains limited. We aimed to identify factors associated with metastatic breast cancer using hospital-based cancer registry data. METHODS: This multicentre observational study included 76,356 women older than or equal to 15 years diagnosed with primary breast cancer between 2009 and 2020. The primary outcome was metastasis at diagnosis. Multivariable Poisson regression estimated incidence risk ratios (IRRs) for metastasis risk with 95% confidence intervals, and random forest models assessed variable importance. FINDINGS: Among 76,356 women, metastasis occurred in 12.96%, with bone as the most frequent metastatic site (2487 [25.1%]). Metastasis risk was higher for diagnoses occurred during 2015-2020 compared with 2009-2014 (IRR 1.15, 95% CI 1.04-1.28). Lower incidence of metastasis was observed in private (IRR 0.50, CI 0.38-0.65) and NGO hospitals (IRR 0.32, CI 0.20-0.51) compared with government hospitals. Larger tumour size (5 to <10 cm: IRR 2.92, CI 1.76-4.83; 10-20 cm: IRR 3.72, CI 2.21-6.27) and grade 2 tumours (IRR 1.62, CI 1.01-2.59) were associated with increased risk. Random forest analysis identified supraclavicular node involvement, tumour size, skin involvement, lympho-vascular invasion, and receptor status as key predictors. INTERPRETATION: Metastatic breast cancer in India is strongly associated with indicators of tumour burden and disease progression. Strengthening early detection, improving diagnostic completeness, and ensuring equitable access to specialised cancer care may help reduce metastatic disease burden. FUNDING: Intramural grant of the ICMR-NCDIR, India.