The association between lymphopenia and serious infection risk in rheumatoid arthritis

Objectives: To investigate the relationship between occurrence of serious infection (SI) and lymphocyte counts in patients with RA using data from a single centre. Methods: We used routinely captured data from a single tertiary rheumatology centre to explore the relationship between lymphopenia and SI risk. Adult RA patients were included over a 5-year follow-up period. Admissions due to confirmed SI were considered. SI rate with 95% confidence intervals was calculated. The association between SI with baseline lymphocyte counts, time-averaged lymphocyte counts throughout all follow-up, and a nadir lymphocyte count was assessed using Cox proportional hazards regression. The relationship between lymphopenia over time and SI was analysed using a mixed-effect model of lymphocyte counts prior to SI. Results: This analysis included 1095 patients with 205 SIs during 2016 person-years of follow-up. The SI rate was 4.61/100 patient-years (95% CI: 3.76, 5.65). Compared with patients with nadir lymphocyte counts >1.5 x 10(9) cells/l, nadir lymphopenia <1 x 10(9) cells/l was significantly associated with higher SI risk (HR 3.28; 95% CI: 1.59, 6.76), increasing to HR 8.08 (95% CI: 3.74, 17.44) in patients with lymphopenia <0.5 x 10(9) cells/l. Lymphocyte counts were observed to be reduced in the 30-day period prior to SI. Conclusion: Lymphocyte counts below <1.0 x 10(9) cells/l were associated with higher SI risk in RA patients; the strongest association between lymphopenia and SI was observed when lymphocyte counts were below <0.5 x 10(9) cells/l. Lymphopenia may be used as a measure to stratify patients at risk of SI.