Background: Breast cancer remains the leading cause of cancer-related mortality among women worldwide, and outcomes are disproportionately poor in low- and middle-income countries, largely as a result of delays along the diagnostic–therapeutic pathway. However, in Bolivia, quantitative data on care intervals and stage at diagnosis remain scarce.
Methods: We conducted a retrospective, single-centre, cross-sectional study of women aged ≥18 years with histologically confirmed invasive breast cancer treated at a public referral centre in Cochabamba, Bolivia (January 2019–December 2021). Clinical, pathological and molecular data were extracted. Following Aarhus Statement guidelines, we calculated the partial diagnostic interval (first diagnostic imaging to histopathological confirmation) and the early pretreatment interval (diagnosis to first medical oncology consultation). Advanced stage was defined as III–IV. Group comparisons used Kruskal–Wallis tests (significance: p < 0.05). Multivariate logistic regression with multiple imputations (five imputations) was performed to identify independent predictors of advanced-stage disease (significance: p < 0.05).
Results: Among 242 patients, the mean age was 52.1 years. Overweight/obesity prevalence was 74.8%. Invasive ductal carcinoma predominated (89.7%), and Luminal B-like was the most frequent molecular subtype (40.1%). Notably, 58.7% presented with advanced-stage disease (stage III: 46.3%; stage IV: 12.4%). The mean partial diagnostic interval was 79.0 days (SD 119.0), and the mean early pretreatment interval was 65.7 days (SD 111.7). Interval durations did not differ significantly by molecular subtype or tumour stage in univariate comparisons (p > 0.05). In the multivariate analysis, molecular subtype was the only independent predictor of advanced-stage presentation. Compared to triple-negative breast cancer, patients with Luminal A had significantly lower odds of presenting with advanced-stage disease (OR 0.09, 95% CI 0.02–0.32, p < 0.001), and those with Luminal B also showed reduced odds (OR 0.30, 95% CI 0.10–0.96, p = 0.043). Crucially, neither the partial diagnostic interval nor the early pretreatment interval was significantly associated with advanced-stage presentation after multivariate adjustment.
Conclusion: In this Bolivian reference institution, advanced-stage breast cancer (III–IV) was prevalent alongside partial system delays. However, multivariate analysis demonstrated that measured diagnostic and pretreatment intervals were not independently associated with stage at diagnosis. Despite methodological constraints regarding partial intervals, these findings underscore the critical role of tumour biology over system delays in stage presentation. There is an urgent need for structured early-detection programs and prospective studies mapping the complete diagnostic–therapeutic pathway in Bolivia.
