Population-based risks for cancer in patients with ALS

Objective: To estimate the risks for cancer (overall and site-specific) in an amyotrophic lateral sclerosis (ALS) cohort. Methods: In this observational longitudinal study, ALS and cancer cases were identified in a computerized Utah genealogy database (Utah Population Database) linked to a statewide cancer registry and death certificates. Hazard ratios (HRs) were estimated as the ratio of observed to expected number of cancers. Site-specific rates for cancer were estimated within the Utah Population Database; sex, birth year (5-year range), and birth state (Utah or not) cohorts were used to estimate the expected number of cancers among ALS cases. To account for an overall shortened lifespan, Cox regression was used to include years at risk in estimation of cancer risks for ALS cases. Results: An overall decreased hazard (hazard ratio [HR] 0.80, p = 0.014, 95% confidence interval [CI] 0.66-0.96) was found for cancer of any site in 1,081 deceased patients with ALS. A decreased hazard was found for lung cancer (HR 0.23, p = 0.002, CI 0.05-0.63). An increased hazard was found for salivary (HR 5.27, p = 0.041, 95% CI 1.09-15.40) and testicular (HR 3.82, p = 0.042, 95% CI 1.06-9.62) cancers. A nonsignificant hazard was observed for cutaneous malignant melanoma (HR 1.62, p = 0.12, 95% CI 0.88-2.71) for which increased risk has previously been reported. Conclusions: Using a unique population database, the overall risk of cancer of any site was found to be significantly reduced in cases with ALS, as was the risk of lung cancer. Significantly increased risk was observed for salivary and testicular cancers.