Recent data have suggested a higher incidence of clinically significant prostate cancer in men with inflammatory bowel disease (IBD), with Burns et al demonstrating a 4-times increased risk of high-grade prostate cancer in patients with IBD.¹ Previous epidemiologic studies have noted an association between chronic inflammation and prostate cancer as well as increased risk of gastrointestinal malignancy and extraintestinal malignancies, including lymphoma and skin cancer, in patients with IBD.^2,4 The risk of urological tumors in patients with chronic inflammatory conditions, including but not limited to IBD, is not well elucidated.^5,6

Our study grew out of the initial Burns et al paper and sought to investigate the possible link between prostate cancer and IBD. As one of our avenues of exploration, we looked into inflammation as a common cause, and one of the most common classes of anti-inflammatory medications used worldwide are the tumor necrosis inhibitors (TNF-I). There are multiple on-label indications for these medications including IBD, psoriasis, rheumatoid arthritis, and many others. The literature has not established any link between cancer development and TNF-I exposure aside from certain nonmelanoma skin cancers and lymphomas. However, very few studies have follow-up data beyond 1 year of drug treatment, which is a nonconclusive follow-up time frame to evaluate the risk of solid malignancy.⁷ The studies that have longer follow-up are also plagued by other issues, including reliance on a spontaneous reporting system for adverse events or use of multiple medications concomitantly.^8,9 Furthermore, there are no studies that specifically evaluate the risk of urological cancer after TNF-α-I exposure in the published literature.¹⁰ As such, we sought to examine the risk of urological malignancies in patients on long-term TNF-α-I immunosuppression through a multicenter, single–health system, retrospective cohort.

We queried for adult patients who presented to any clinic within the Northwestern Medicine network from July 1996 through January 2020. Patients exposed to TNF-α-I were identified using the generic medication names for any of the 5 TNF-α-I (adalimumab, infliximab, etanercept, certolizumab, and golimumab) along with the chronic inflammatory condition for which they were prescribed. Prostate cancer in this population was identified using ICD-9 and ICD-10 codes followed up by manual chart review. Because there was a significant time period in which TNF-I exposed patients were in the system before initiation of the TNF-I, we employed a time-dependent analysis across exposure groups of each inflammatory condition. We used control groups as the patients with chronic inflammatory conditions without TNF-I exposure. After a long internal discussion and search of the literature, we decided to only include a malignancy if it was diagnosed at least 6 months after initial TNF-I exposure. Although there is no literature to estimate biologic feasibility, this was the consensus of a panel of oncologic experts on our team.

These data, which we presented at AUA 2023 in Chicago, found a total of 15,190 men with chronic inflammatory conditions for which TNF-I is a Food and Drug Administration–approved treatment (Figure 1). There were 4,209 men exposed to a TNF-I and 10,981 men who remained unexposed to TNF-I. Median post-TNF-I exposure follow-up was 46 weeks (IQR 21-78 weeks) with 53 patients (1.3%) subsequently developing prostate cancer. Median follow-up time without TNF-I exposure was 87 weeks (IQR 38-144 weeks) with 490 patients (4.5%) subsequently developing prostate cancer (Table 1). After our time-dependent analysis, TNF-I exposure was associated with a decreased risk of prostate cancer (HR 0.58, 95% CI 0.42-0.80, P = .001; Figure 2). There was no difference in PSA at time of diagnosis, Grade Group on biopsy specimen, number of positive biopsy cores, or rates of adverse pathology on prostatectomy specimen (Table 2).

Our manuscript is the first publication of the finding that TNF-I exposure may be protective against prostate cancer. Our data include more than 15,000 men with a median follow-up of over 1 year and define a sample size similar to previous interventional and observational meta-analyses and comparable to previous large registry-based cohort studies.^7-9 Overall, our findings are very exciting, and we plan to further investigate these findings with increased analyses on our current data set as well as attempting to set up multi-institutional prospective studies to aggregate more robust data on this possible connection. Whether or not this process is driven by inflammation or other processes in the tumor microenvironment, we plan to continue to explore this new clinical phenomenon.

Table 1. Baseline Characteristics of Prostate Cancer Patients

Table 2. Prostate Cancer Characteristics