Characteristics and survival of patients with lung cancer at Uganda Cancer Institute

Abstract

Background: Lung cancer incidence is increasing and is the 7th leading cause of cancer-related deaths in Uganda. However, the characteristics, survival, and prognostic factors among lung cancer patients in Uganda are not well described, resulting in the use of extrapolated data from developed countries with different demographic and logistical challenges. We describe the baseline characteristics, survival, and prognostic factors among lung cancer patients at Uganda Cancer Institute (UCI). Methodology: This was a retrospective chart review of 207 lung cancer patients at UCI between January 2008 and August 2018. Data on demographic, clinical, laboratory, and treatment characteristics, and vital status were abstracted and analyzed. We determined survival as the time from histological diagnosis to death. The Kaplan Meier survival function was used to calculate the yearly event probabilities and median survival time and the 5-year overall survival rate. The factors associated with survival were analyzed using bivariate (log-rank) and then multivariate analysis through the Cox-proportional hazards regression model. Results: Majority were mostly female (56.5%) with a median age of 60 years (range 20-94 years) and never-smokers (78.7%). Difficulty in breathing (83.6%), cough (80.7%) and history of pleural effusion (66.2%) were the commonest symptoms. Presumptive anti-tuberculosis treatment was given to 23.2% for a median duration of 12 weeks (range 2-32 weeks) before a lung cancer diagnosis was made. Majority had NSCLC (96.6%) of which 74.5% was adenocarcinoma and 19% squamous cell carcinoma. All had advanced (stage 3 or 4) lung cancer with 96.1% having stage 4 disease. Chemotherapy (44.9%) and biological therapy (34.8%) were the commonest treatments used. Median survival was 4.4 months while 5-year survival was 1.7%. Duration since first symptom (HR 0.993), ECOG performance status, ALP (1.003), Total bilirubin (HR 1.068), Squamous cell carcinoma (HR 2.132) and biological therapy (HR .502) were associated with survival on multivariate cox regression analysis. Conclusion: Lung cancer in Uganda occurred predominantly among middle aged female never smokers with adenocarcinoma histology, stage 4 disease, and poor survival. Survival was low and treatment with biological agents was associated with longer survival. This is in contrast with patients in the western world. Recommendations: We recommend the development of an algorithm for lung cancer screening and early diagnosis in our population, training of health workers on lung cancer screening and further research on lung cancer in Uganda.