High Incidence of Thromboembolic Complications and PF4 Positivity in Patients with Ventricular Assist Devices.

Abstract

Patients with ventricular assist devices (VAD) are at high risk for thromboembolic complications (TEC) probably due to platelet activation and excessive thrombin generation once the device is implanted. This warrants prophylactic use of full dose anticoagulation, usually with unfractionated heparin (UFH), which increases the risk for development of heparin-induced thrombocytopenia (HIT). We retrospectively evaluated the incidence of TEC and PF4 positivity in 50 sequential patients who underwent VAD placement for severe ventricular dysfunction between November 2002 and April 2005. Median age at VAD placement was 52 years (range 18–72), 68% were men, and all patients were on full dose UFH post-VAD insertion. TEC were observed in 22/50 patients (44%), with 8 patients experiencing >1 event. CNS embolic phenomena occurred in 11/22 (50% of TEC), with watershed/non-embolic CNS infarcts being identified in 2 additional patients. Splenic infarcts were identified in 6/22 (27%). Two patients had lower extremity deep vein thromboses (DVT), and two patients had pulmonary emboli. One patient died from complications of multiple infarcts involving his spleen and gut that were attributed to HIT. All patients became thrombocytopenic (<150K) at some point during VAD/UFH use. 44 patients (88%) dropped the platelet count below 100K with an average nadir of 56K. Antibodies to the PF4/heparin complex were checked using a standard enzyme immunoassay (EIA) in 38 (76%) patients, and found to be positive in 17 (44.7%). PF4 antibodies were detected using the EIA in 50% of patients with TEC and in 25% of patients with no documented TEC. A functional antibody assay was checked in a total of 17 patients, the serotonin release assay (SRA) in 13 and heparin-induced agglutination (HIA) in 4 cases. The SRA was only positive in 2/13 cases, and concordant with the PF4 result both times. The HIA was positive once, and concordant with the PF4 result. These results highlight the high incidence of TEC in patients with a VAD, despite anticoagulation therapy. There is a particularly high rate of embolic stroke in this patient population, and better ways of preventing this complication need to be explored. Thrombocytopenia is also extremely common post-VAD placement, prompting evaluation of HIT in over ¾ of the patients, often necessitating use of direct thrombin inhibitors. The frequent detection of PF4 antibodies in this patient group and lack of correlation with functional assays illustrates the need for more specific laboratory means to quickly confirm those patients with HIT.