Trends in Diagnoses of Non-Infectious Pneumonitis in Medicare Patients with Advanced Non-Small Cell Lung Cancer

Purpose: Non-infectious pneumonitis (NIP) is a common complication of treatment for non-small cell lung cancer (NSCLC). Often identified and labeled by its presumed precipitant (radiation vs. immune-related), NIP occurs in as many as 16% of patients with advanced NSCLC. We aimed to describe the trends of diagnoses of NIP in Medicare patients diagnosed with advanced NSCLC between 2013-2017.

Methodology: A retrospective cohort was assembled using the Surveillance, Epidemiology, and End Results-Medicare database, analyzing patients diagnosed with American Joint Committee on Cancer (AJCC) stages 3B-4 NSCLC between 2013-2017. We included patients who had at least 12 months of Medicare Parts A/B enrollment before diagnosis and were not enrolled in a health maintenance organization in the 12 months after diagnosis. Inpatient and outpatient claims data were searched between 2012-2018 using a previously validated algorithm for identifying cases of NIP. All patients with a diagnosis of NIP before their NSCLC diagnosis were excluded. Modified Poisson regression was performed to estimate the incidence rates (IR) of NIP for each year of diagnosis and to determine the IR ratios (IRR) for NIP evaluating year of diagnosis as a categorical and ordinal variable.

Results: There were 24,396 patients included in the analysis with 5,065 (20.8%), 5,047 (20.7%), 5006 (20.5%), 4,753 (19.5%), and 4,525 (18.6%) diagnosed in each year between 2013-2017, respectively. There were no differences in sex, AJCC stage, squamous vs. non-squamous histology, treatment with radiation therapy, and age at diagnosis among years of treatment (p>0.05). There was no significant trend of increased frequency of NIP diagnoses immediately after International Classification of Diseases, Tenth Revision, Clinical Modification coding was implemented (p=0.56). For patients diagnosed in 2013-2017, there were 3.7, 3.7, 4.6, 6.8, and 11.9 cases per 1000 person-years, corresponding to a 34% (95%CI: 27-42%) relative increase in the IR of pneumonitis per year. While there were no significant differences between the IR of NIP for patients diagnosed in 2014 (IRR=0.99, 95% confidence interval (95%CI): 0.79-1.24) and 2015 (IRR=1.24 (95%CI: 1.00-1.55) relative to 2013, there were significant increases in the IR of NIP in 2016 (IRR=1.85, 95%CI: 1.49-2.29) and 2017 (IRR=3.22, 95%CI: 2.56-4.05). There was a qualitatively higher increase in IR of NIP for those diagnosed with AJCC stage 4 NSCLC vs. AJCC stage 3B NSCLC (IRR=1.40, 95% CI: 1.30-1.50 vs. IRR=1.27, 95%CI: 1.16-1.40), although this interaction was not statistically significant (p=0.14).

Conclusions: Since 2013, patients diagnosed with NSCLC have an increasing risk of developing NIP. This may be attributable to the approval of immune checkpoint inhibitors (ICIs) during the study period, although that remains to be confirmed. Future research should directly evaluate how treatment patterns with ICIs and radiation therapy have impacted NIP diagnoses.

References: