Incidence of Infections and Malignancy Among Elderly Male Patients with IBD Exposed to Vedolizumab, Prednisone, and 5-ASA Medications: A Nationwide Retrospective Cohort Study

摘要

Introduction Vedolizumab (VDZ) is postulated to have a potentially safer side effect profile than other biologic medications owing to its gut-selective mechanism. However, extrapolating these safety data to older patients is challenging because of their underrepresentation in or exclusion from most clinical trials, higher rates of withdrawal, and higher rates of comorbidities. Our aim was to evaluate the absolute risk of infections and malignancies in an elderly group of patients with inflammatory bowel disease (IBD) exposed to VDZ vs. the absolute risks associated with 5-aminosalicyclic acid (5-ASA) medications and chronic steroid use. Methods We conducted a retrospective cohort study among the US national Veterans Affairs Healthcare System (VAHS). Our cohort comprised patients who were followed in the VAHS, had a diagnosis of IBD, and were aged 65 years or older. The patients were divided into three cohorts: primary exposure group (elderly patients on VDZ), assumed low-risk group (elderly patients on 5-ASA only), and assumed high-risk group (elderly patients on chronic prednisone). The low-risk and high-risk groups were matched to the VDZ group on race, gender, IBD type, age, and Charlson Comorbidity Index (CCI). Primary outcomes gathered and confirmed via chart review included mild infections, severe infections, malignancies, and non-melanoma skin cancers (NMSC). The results were based on a descriptive analysis. Results A total of 497 patients were included in our study with 213, 186, and 98 patients in the VDZ, 5-ASA, and steroid groups, respectively. The total patient-years (PYs) of follow up were 405, 656, and 303 in VDZ, 5-ASA, and steroid groups respectively. The incidence of mild infection was the lowest in the VDZ group with 93.1 outcomes per 1000 PYs as compared to the 5-ASA group (114.4 outcomes per 1000 PYs) and 155.1 outcomes per 1000 PYs in the steroid group. In regard to severe infections, the VDZ group had an incidence of 38.5 outcomes per 1000 PYs as compared to 30.6 outcomes per 1000 PYs in the 5-ASA group and 67.4 outcomes per 1000 PYs in the steroids group. Mild infections with the highest incidence in the VDZ group were upper respiratory infection (including pharyngitis and sinusitis) at 20.3 per 1000 PYs, Clostridium difficile (15.1 per 1000 PYs), and cellulitis (10.0 per 1000 PYs). The severe infection with the highest incidence was pneumonia for each group, with incidence rates of 10.0, 14.0, and 48.6 per 1000 PYs for the VDZ, 5-ASA, and steroid groups, respectively. Incidence of malignancies (excluding NMSC) was numerically similar in the VDZ and 5-ASA group (17.6 and 15.6 per 1000 PYs, respectively), while the steroid group showed a higher incidence of 42.6 per 1000 PYs. NMSC incidence was numerically similar in the VDZ and steroid groups (36.3 and 39.0 per 1000 PYs, respectively), with the 5-ASA group showing a much lower NMSC incidence (4.6 per 1000 PYs). Conclusion In a large nationwide cohort of elderly patients, we found the safety profile of VDZ among elderly patients with IBD with respect to the risk of infection and malignancy to be numerically similar to elderly patients with IBD taking 5-ASA, and favorable when compared to the elderly patients with IBD taking chronic steroids.