Predictors of Imaging Surveillance for Surgically Treated, Early-Stage Lung Cancer

Abstract

Purpose: Current guidelines recommend routine imaging surveillance for non-small cell lung cancer (NSCLC) patients following treatment. Little is known about surveillance patterns for surgically resected, early-stage lung cancer patients in the community-at-large. We sought to characterize surveillance patterns in a national cohort. Methods: We conducted a retrospective study using Surveillance, Epidemiology, and End-Results (SEER)-Medicare database (1995-2010). Patients with stage I/II NSCLC treated with surgical resection were included. Our primary outcome was receipt of imaging between 4 and 8 months following surgery. Covariates included demographics and comorbidities. Results: Chest radiography (CXR) was the most frequent initial modality (60%) followed by chest computerized tomography (CT) (25%). Positron emission tomography (PET) was least frequent as initial imaging modality (3%). A total of 13% of patients received no imaging within the initial surveillance period. Adherence to National Comprehensive Cancer Network (NCCN) guidelines for imaging by overall prevalence was 47% for receipt of CT, however rates of CT imaging increased over time from 28% to 61% (p<0.01). Reduced rates of CT imaging were associated with stage I disease and surgery as the sole treatment modality. Conclusions: Imaging following definitive surgery for NSCLC predominantly utilizes CXR rather than CT. Most of this imaging is likely for surveillance and in that context, CXR has inferior detection rates for recurrence and detection of new cancers. Adherence to guideline recommended CT surveillance following surgery is poor, but the reason multifactorial. Efforts to improve adherence to imaging surveillance must be coupled with greater evidence demonstrating improved long-term outcomes.