Frequency of surveillance computed tomography in non-Hodgkin lymphoma and the risk of secondary primary malignancies: A nationwide population-based study.

With increasing usage of computed tomography (CT) for lymphoma patients receiving curative-intent treatment, development of secondary primary malignancy (SPM) related to radiation from CT scans becomes an emerging issue in these long-term survivors. We conducted a nationwide population-based study analyzing non-Hodgkin lymphoma (NHL) patients receiving curative-intent treatment between January 1997 and December 2010. Patients were divided into two populations by the medium number of CT performed. The cumulative incidence of SPM in these two groups was compared using the Kaplan-Meier method. Propensity score matching was applied to eliminate potential confounders. Group stratification and multivariate analyses calculated by Cox proportional hazard models using competing risk analyses adjusted for mortality were performed to identify independent predictors for SPM. Patients receiving >8 CT scans had a significantly greater risk for developing SPM (hazard ratio [HR] 2.25, 95% confidence interval [CI] 1.61-3.13; p < 0.001) than those with ≤8 scans and this difference remained significant even after correction with propensity score matching. Among the 180 SPM identified, those receiving more CT scans had significantly higher SPM incidence in cancers of the breast (HR 11.22), stomach (HR 5.22) and liver and biliary tract (HR 2.18) in comparison to those with less exposure. The risk of SPM was estimated to increase 3% per one more CT scan performed. Our study demonstrated that after curative-intent treatment, patients with NHL receiving more frequent surveillance CT scans would have an increased risk of SPM.