Converging Cancer Profiles in Bangladesh: A Comparative Statistical Analysis of Blood, Lung, Breast, and Cervical Cancer Datasets

Abstract

Background: Bangladesh faces a growing cancer burden that remains inadequately characterized because of limited population-based data and fragmented cancer surveillance systems. This study integrates four publicly available datasets from Bangladeshi clinical and academic settings, comprising 19,516 records across blood, lung, breast, and cervical cancers. This integrated approach provides a unique opportunity to identify cross-cancer patterns within a low- and middle-income country (LMIC) setting, where comprehensive clinical data remain limited.

Materials and Methods: Descriptive statistics, chi-square tests of independence, one-way analysis of variance (ANOVA), and independent-samples t-tests were used to summarize patient characteristics, examine associations between clinical variables, and compare outcomes across four independent datasets: (1) a multivariate blood cancer clinical dataset (n = 2,295); (2) a lung cancer survival dataset (n = 1,782) collected from four tertiary hospitals in Dhaka; (3) a breast cancer molecular and clinical dataset (n = 15,054); and (4) a cross-sectional survey of cervical cancer awareness and HPV vaccine uptake among 385 female undergraduate students.

Results: Significant associations were identified between cancer subtype and treatment outcome in the blood cancer cohort (χ² = 2307.76, p < 0.001), disease stage and survival in the lung cancer cohort (χ² = 375.26, p < 0.001), histologic grade and tumour size in the breast cancer cohort (F(2,15045) = 140.66, p < 0.001), and cervical cancer awareness and HPV vaccine knowledge (χ² = 32.88, p < 0.001). A substantial knowledge-to-behaviour gap was observed in cervical cancer prevention: although 75.1% of unvaccinated participants expressed willingness to receive the HPV vaccine, only 8.3% had been vaccinated. Smoking status was not significantly associated with survival among patients with lung cancer (p = 0.762), and no individual genetic marker (BCR-ABL, FLT3, or TP53) was significantly associated with treatment outcomes in the blood cancer cohort.

Conclusions: Disease stage emerged as the most consistent prognostic determinant across the cancer cohorts, underscoring the importance of early detection and timely diagnosis. The substantial gap between HPV vaccination willingness and actual vaccine uptake, together with the absence of a significant smoking-related survival effect in the lung cancer cohort, highlights important structural barriers to cancer prevention and healthcare delivery in Bangladesh. These findings support strengthening early detection programmes, expanding HPV vaccination strategies, and improving equitable access to cancer diagnosis and treatment.