The utility of thrombophilia testing in patients with newly diagnosed portal vein thrombosis

Thrombophilia testing is frequently performed in both seemingly provoked and unprovoked portal vein thrombosis (PVT), yet the clinical implications of these expensive laboratory tests are unknown. We investigated the frequency of clinical management changes in patients with newly diagnosed PVT. This is a retrospective analysis of adult patients with a newly diagnosed PVT at a single institution. The primary outcome is change in clinical management, defined as documented change in choice, dose, or duration of anticoagulation, future thromboprophylaxis, or counseling of asymptomatic family members. Five-hundred and forty-four patients with PVT were identified, 438 (80.5%) of whom had an identifiable pretesting provoking factor, most commonly cirrhosis (39.2%). Two-hundred ninety-one patients (53.5%) had at least one hypercoagulable laboratory test performed. The most frequently positive test was PAI-1 polymorphism, followed by elevated homocysteine and MTHFR mutational analysis. However, the only test that was frequently positive and consistently altered management was JAK2 mutational analysis (15.3%). Factor V Leiden was commonly positive but rarely changed clinical decision-making (1.5%), as was flow cytometric testing for paroxysmal nocturnal hemoglobinuria (0.8%), and antiphospholipid antibodies (0.7%). Patients with cirrhosis rarely had thrombophilia testing results that were clinically significant. A rough cost estimate was dramatically reduced from $231 000 to $76 000 if only clinically meaningful tests were employed in the hypercoagulable work-up. These results highlight the need for focused thrombophilia testing in patients with PVT.