Pneumocystis pneumonia incidence in metastatic solid tumor and lymphoma patients from an administrative health care claims database.

Abstract

e12500 Background: Pneumoncystis jiroveci pneumonia (PJP) causes severe, often life-threatening illness in immunocompromised patients, including patients with cancer. PJP incidence estimates have been reported for several different cancer types based on case series from single institutions. In order to obtain incidence estimates from a larger pool of patients, we studied PJP claims among patients with metastatic solid tumors (MSTs) or lymphomas in a large administrative healthcare claims database. Methods: Adults diagnosed with MSTs or lymphomas between January 1999-September 2011 were identified in the MarketScan commercial, Medicare supplemental and Medicaid databases based on ICD-9 diagnosis codes. First PJP after cancer diagnosis was defined by PJP ICD-9 code (136.3) associated with an inpatient hospital stay of at least 1 day. Patients with a PJP claim before their cancer diagnosis were excluded. PJP incidence proportions and rates (with 95% CIs) were calculated by age group, sex, HIV status, and tumor type. Results: There were 158 PJP cases (0.04%) among 392,369 HIV-negative MST patients, an incidence rate of 3.0/10,000 pt-yrs (95% CI 2.6-3.5). Incidence rates by metastatic cancer type were 6.8/10,000 pt-yrs for lung, 5.4/10,000 pt-yrs for pancreas, 4.1/10,000 pt-yrs for prostate, 3.0/10,000 pt-yrs for breast, and 0.9/10,000 pt-yrs each for colorectal and ovary. There were 337 PJP cases (0.21%) among 162,799 HIV-negative lymphoma patients, an incidence rate of 9.4/10,000 pt-yrs (95% CI 8.5-10.5). Incidence was higher among males compared to females in both MST and lymphoma patients. PJP incidence rates in HIV-positive patients were approximately 14 times higher than in HIV-negative patients. Conclusions: PJP is a rare occurrence among patients with MSTs or lymphomas. In this analysis of a large administrative healthcare claims database, PJP incidence was higher among lymphoma patients compared to MST patients. Incidence estimates were lower than reports from institutional case series, but patterns of occurrence were consistent in terms of higher incidence in patients with lymphomas or HIV. Differences in incidence by tumor type were also observed among MST patients.