Cancer Survival Analysis: A Pilot Study in Lebanon on Lung and Colorectal Cancers

Abstract

Background: Cancer survival rates (CSR) have never been estimated in Lebanon based on locally recorded data. The main obstacles to this otherwise routine computation have been the incomplete nature of cancer incidence data and the absence of usable cancer-specific mortality data. While a unique identifier is being tested to facilitate the linkage between an incident case and that case’s eventual death, the need for a preliminary CSR estimation has become crucial for Public Health decision- makers. Objective: To respond to this need, an approach is being tested in this pilot analysis, using a sub- sample of incident cancer cases recorded at the National Cancer Registry (NCR) in 2023 for colorectal and lung cancer. Methods: A survival analysis was conducted using a sub-sample of colorectal and lung cancer. This sub-sample includes patients who had requested free cancer medications through the Drug Dispensing Center (DDC) managed by MOPH, for whom identifying data had therefore been recorded as per DDC requirements. Among identifiers, telephone numbers allowed the team at NCR to contact the 2023 patients to inquire about their current situation with the previously diagnosed cancer, including a possible death event by the end of 2025. Cases that did not answer the call were dealt with in two ways: multiple imputation and grouping into 2 groups: status missing vs status known. Overall survival was calculated from the time of diagnosis to death. Survival probabilities were estimated using the KaplanMeier method. A Cox model was fitted to estimate the 3-year survival trend for the entire caseload, stratified by site of cancer, sex, age, and stage at diagnosis. All analyses were performed using STATA. Results: The CRC cohort consisted of 407 patients, of which 52% are males. The 3 year survival was 69.5%. Mean survival time decreased steadily but very slightly with age, from 21.3 months for those ≤60 years to 20.7 months for those > 60 years. Mean survival was similar between males (21.3 months) and females (20.7 months). The lung cancer cohort consisted of 703 patients, of which 61.6% were males. All patients were diagnosed in 2023, and the 3-year survival was 69%. The mean survival time did not vary between age groups and gender, but stage was an important predictor for survival when it came to lung cancer and colorectal cancer. Conclusions: This analysis creates the standard frame for future survival analysis conducted with NCR data. Results with this sub-sample are totally different from those of another preliminary analysis conducted in 2022 at AUBMC, or from general survival data worldwide. This difference can be attributed to the lack of validity in the data currently available to the NCR. In particular, the lack of follow-up reduced the probability of obtaining usable results. It is important to address issues of data validation prior to adopting a routine approach in survival analysis in Lebanon.