BACKGROUND: Whether therapeutic plasma exchange (TPE) results in clinically relevant bleeding or thrombotic events is not well characterized.

STUDY DESIGN AND METHODS: We conducted a single-institution retrospective cohort study to estimate the incidence of bleeding and thrombotic outcomes of patients treated with TPE from 2017 to 2021. Patients with ≥3 treatment sessions within a 14-day period were included. Two independent physician reviewers manually reviewed patient charts. We calculated the 30-day cumulative incidence of bleeding and thrombotic outcomes with 95% confidence intervals (CI). We used a regression model to determine the association between clinical and laboratory features and study outcomes.

RESULTS: The study comprised 145 patients who underwent 176 treatment courses, representing a cumulative total of 890 individual TPE procedures. The cumulative incidence of clinically relevant bleeding (including major bleeding and clinically relevant nonmajor bleeding) was 8.0% (95% CI 4.4%-13.0%), and thrombosis was 4.0% (95% CI 1.6%-8.0%). There was one fatal bleeding event. Older age and chronic kidney disease were associated with odds ratios of 1.08 (95% CI 1.03-1.13, p = .002) and 4.68 (95% CI 1.33-14.90, p = .01), respectively, for bleeding. Patients who received any quantity of fresh frozen plasma as replacement fluid had a bleeding rate of 16.4% compared to 6.6% in patients who did not receive plasma (p = .05).

DISCUSSION: TPE-associated bleeding complications occur at a higher rate than previously understood. Further prospective investigations are merited to characterize bleeding risk factors and optimal preventive strategies.