Overall survival, poverty differentials, and mediating pathways among women with breast cancer: South African Breast Cancer and HIV Outcomes cohort

Maureen Joffe(University of the Witwatersrand), Wenlong Carl Chen(National Health Laboratory Service), Ashleigh Craig(South African Medical Research Council), Daniel S. O’Neil(Yale Cancer Center), Alfred I. Neugut(Columbia University), Judith S. Jacobson(Columbia University), Paul Ruff(South African Medical Research Council), Rofhiwa Mathiba(University of the Witwatersrand), Nivashini Murugan(University of the Witwatersrand), Herbert Cubasch(University of the Witwatersrand), Charmaine Blanchard(University of the Witwatersrand), Sarah Nietz(University of the Witwatersrand), Jennifer Edge(University of the Witwatersrand), Ines Buccimazza(Queen Nandi Regional Hospital), Sharon Čačala(Stellenbosch University), Valerie A McCormack(Centre international de recherche sur le cancer), Yoanna S Pumpalova(Columbia University), Shane A. Norris(South African Medical Research Council)
The Oncologist
December 11, 2025
Cited by 2Open Access
Full Text

Abstract

BACKGROUND: Breast cancer survival rates in sub-Saharan Africa are low. In a prospective, multi-center cohort study, we estimated 5-year overall survival rates, overall survival determinants, and mediating effects between socioeconomic status on overall survival among South African women diagnosed with invasive BC. PATIENTS AND METHODS: Patients from 4 public hospitals were enrolled between July 1, 2015 and January 31, 2019. Survival determinants were assessed using Cox proportional hazard models adjusted for age, background mortality, and treatments. Socioeconomic pathway effects on overall survival were determined through generalized structural equation models. RESULTS: Of 2838 participants, 58% had advanced-stage (III/IV) disease. Five-year crude overall survival was 44.3% (95% CI 42.5-46.2). Significant mortality risks were late stage at diagnosis (hazard ratio [HR] = 2.31 [95% CI 1.99-2.69] [stage III]; 4.79 [95% CI 3.96-5.80] [stage IV]), HIV-positive status (HR = 1.45 [95% CI 1.25-1.67]), unemployment HR = 1.25 [95% CI 1.09-1.44], and low education HR 1.19 [95% CI 1.04-1.37]). Age and treatment-adjusted socioeconomic status effects on overall survival were mediated through HIV status (81.7% of the effect) and stage at diagnosis (81.7%), both P < .001. Poor breast cancer knowledge had an indirect effect on overall survival, accounting for 77.6% of the total effect (P = .001), fully mediated by late-stage presentation. Socioeconomic status had no significant direct path to mortality after accounting for these mediators. CONCLUSION: Interventions should prioritize early breast cancer detection. For patients with low socioeconomic status, particularly those with comorbid HIV, we must mitigate multifaceted barriers to healthcare access, including limited awareness and knowledge of breast cancer.


Related Papers

No related papers found

Powered by citation graph analysis