Unintentional platelet removal by plasmapheresis.

Therapeutic plasmapheresis may remove platelets as well as plasma. Unintentional platelet loss, if not recognized, may lead to inappropriate patient assessment and treatment. A patient with thrombotic thrombocytopenic purpura-hemolytic uremic syndrome (TTP-HUS) is reported in whom persistent thrombocytopenia was interpreted as continuing active disease; thrombocytopenia resolved only after plasma exchange treatments were stopped. This observation prompted a systematic study of platelet loss with plasmapheresis. Data are reported on platelet loss during 432 apheresis procedures in 71 patients with six disease categories using three different instruments. Comparing the first procedure recorded for each patient, there was a significant difference among instrument types (P<0.001); platelet loss was greater with the Fresenius AS 104 (17.5%, N = 21) than with the COBE Spectra (1.6%, N = 26) or the Haemonetics LN9000 (2.6%, N = 24). With all procedures, platelet loss ranged from 0 to 71%. Among disease categories, platelet loss was greater in patients with dysproteinemias who were treated for hyperviscosity symptoms. Absolute platelet loss with the first recorded apheresis procedure, in the 34 patients who had a normal platelet count before the procedure, was also greater with the AS 104 (2.23 x 10(11) platelets) than with the Spectra (0.29 x 10(11) platelets) or the LN9000 (0.37 x 10(11) platelets). In 39 patients in whom data were collected on consecutive days, platelet removal by plasmapheresis correlated with a decreased patient platelet count (r = 0.40, P = 0.011). In these 39 patients, the platelet counts were significantly decreased at 24 hours (P = 0.002).