Background and Objectives

Therapeutic plasma exchange (TPE) is a blood purification treatment capable of removing large molecular weight substances from plasma. It is commonly used for the removal of circulating pathogenic immunoglobulins presumed to be the cause of many autoimmune diseases. TPE can be performed with a membrane-based system (mTPE) or a centrifugal-based system (cTPE). When plasma separation is performed with a membrane, filter clotting can lead to longer treatment time, higher cost and can negatively impact patient satisfaction. In this study, we examine the operational characteristics that might influence filter life.

Design, Setting, Participants, & Measures

We report on 24 patients, with a total of 135 mTPE treatments in a single tertiary care academic center using the NxStage machine. The study focuses on treatment specific parameters that may lead to procedure failure. The main parameters of interest were transmembrane pressure (TMP) and the filtration fraction as displayed on the machine (FFd) compared to the calculated filtration fraction (FFc). Primary outcome was to measure whether TMP, FFc, and FFd influenced filter survival. Secondary outcomes included factors that might have indirectly resulted in filter failure, including hematocrit (Hct), platelet count, heparin use, and intra-treatment calcium administration.

Results

In this study, we demonstrated that machine displayed filtration fractions (FFd) were lower than FFc and this difference was significantly larger in TPE sessions that experienced a clotting event (7.58 vs 6.22, P = .031). TPE sessions that clotted had a higher mean TMP (57.48 mmHg vs 44.43 mmHg, P = .001) and clotting events tended to have a lower mean blood flow rate (175.83 mL/min vs 189.55 mL/min, P = .002). In TPE sessions that received prefilter calcium administration, a higher mean dose of calcium gluconate was found in the sessions that experienced clotting (3.27 g vs 2.70 g, P = .013). Patients who experienced at least one clotting event were noted to be heavier than those patients without any clotting events (91.52 kg vs 72.15 kg, P = .040). Prefilter heparin administration was not associated with a lower incidence of filter clotting. We did not find a statistically significant difference in clotting events based upon type of intravenous access, pretreatment hematocrit, or pretreatment platelet counts.

Conclusion

Among patients undergoing mTPE, machine FFd on the NxStage system are consistently lower than FFc. Treatments where there was a greater difference between displayed and FFc had a greater likelihood of filter clotting. Treatments with higher TMP were associated with failed treatments. Prefilter calcium administration during treatment was associated with increased filter clotting. Lower blood flow rates and higher patient weight were also associated with increased filter clotting. Prefilter heparin administration did not reduce the incidence of filter clotting.