Kaye JA, Margulis AV, Plana E, Calingaert B, Perez-Gutthann S, Arana A. Cancer rates over time after initiation of overactive bladder drugs. Poster presented at the 31st ICPE International Conference on Pharmacoepidemiology & Therapeutic Risk Management; August 2015. Boston, MA.

Background: Genitourinary cancers may cause symptoms similar to those of overactive bladder (OAB). We investigated incidence rates (IRs) of genitourinary cancers and of other common cancer types among initiators of antimuscarinic OAB drugs, taking into account time since first prescription.

Objectives: Estimate IR of the 10 most common cancers stratified by time after initiation of treatment.

Methods: Using the Clinical Practice Research Datalink (CPRD), we assembled a cohort of new users of oxybutynin, tolterodine, solifenacin, fesoterodine, trospium, or darifenacin in 2004-2012 aged ≥18 years and not known to have cancer before cohort entry. Follow-up (FU) ended with cancer diagnosis, death, disenrollment, or end of study period. CA diagnoses were identified in primary care, cancer registry, and/or hospitalization records. Crude IR per 1,000 person-years and 95% confidence interval (CI) were estimated by year since cohort entry (YSCE) for each study cancer (bladder, breast, colorectal, lung, melanoma, non-Hodgkin lymphoma (NHL), pancreas, prostate, renal, and uterine).

Results: Of 119,913 new users of study drugs (mean age at cohort entry 62 years; 70% women), 4,117 with incident study cancers were identified during 399,375 person-years of FU (534 bladder, 886 breast, 545 colorectal, 495 lung, 182 melanoma, 144 NHL, 138 pancreas, 932 prostate, 125 renal, 136 uterine). Bladder cancer IR (95% CI) was greater in earlier periods: 2.5 (2.2–2.8) less than 1 YSCE, 1.2 (0.9–1.4) 1 to less than 2 YSCE, and ≤ 1 in most later years. Prostate cancer IR was 14.2 (12.9–15.6) less than 1 YSCE, 6.8 (5.8–7.9) 1 to less than 2 YSCE, then decreased more gradually. IRs were higher less than 6 months after OAB drug start: bladder 3.5 (3.0–4.0); prostate 19.5 (17.5–21.8). In contrast, risk of other cancers did not show this effect of time since cohort entry.

Conclusions: Patterns of change in IR over time since cohort entry must be considered in etiologic studies of cancer risk related to use of OAB drugs. Protopathic bias and/or detection bias are plausible explanations for higher IRs of bladder and prostate CAs during the first few years after starting OAB drug treatment than in subsequent periods.

Share on: