Introduction : There is little evidence to guide practice on the use of anticoagulation (AC) for patients with immune thrombocytopenia (ITP). In this study, we describe management of AC and clinical outcomes in patients with ITP who had a platelet count <50 x109/L while receiving AC.

Methods : Patients were identified from the McMaster ITP Registry, a prospective registry of adult patients with thrombocytopenia (<150 x109/L) referred to a tertiary hematology clinic in Canada. Patients who had a platelet count <50 x109/L and who were receiving AC at the same time were selected for this study. A detailed chart review was done to augment the clinical data from the registry. For every patient, we defined ' at-risk encounters ' as any occurrence of platelets <50 x109/L while simultaneously on AC. For each encounter, we described management decisions to either stop or continue AC; assessed bleeding events using a validated ITP bleeding tool, which graded the severity of bleeding from 0 (none) to 2 (severe); and recorded thrombotic events. This study was approved by the Hamilton Integrated Research Ethics Board.

Results : Of 615 patients from the registry, 44 received AC at some point. Of those, 13 patients with ITP had platelets <50 x109/L while simultaneously on AC (n=44 encounters). Indications for AC were atrial fibrillation (n=6) or venous thrombosis (n=7). Median age was 74 years and 53.8% were female. Median follow up was 9 months (IQR 22 months).

Of all encounters with AC management data available (n=35 encounters), 7 (20.0%) were associated with grade 2 bleeding, 6 (17.1%) were associated with grade 1 bleeding, and 20 (57.1%) were not associated with bleeding (2 were unknown). AC was stopped during 26 (74.3%) encounters; median platelet count was 15 x109/L. Of those encounters where AC was stopped, 20 (76.9%) resulted in administration of ITP therapies to raise the platelet count; 7 (26.9%) were associated with a subsequent thrombotic event; and none were associated with subsequent grade 2 bleeds. There were 2 deaths among patients who stopped AC: 1) A 74 year-old female who had been receiving AC for atrial fibrillation (CHADS2 = 2) presented with a platelet count of 1 x109/L and a grade 2 bleed. Ten days after stopping AC and receiving ITP treatments, repeat platelet count was 55 x109/L and she developed a massive ischemic stroke resulting in death. 2) A 49 year-old male who had been receiving AC for recurrent deep vein thromboses had 3 encounters with platelets <50 x109/L and a grade 2 bleed. Two years after stopping AC and receiving multiple ITP treatments (repeat platelet count = 167 x109/L), he developed sepsis and ultimately died of presumed pulmonary embolism. AC was continued during 9 (25.7%) encounters (median platelet count, 38 x109/L). Of those, 5 (55.6%) resulted in the use of ITP therapies. There were no subsequent grade 2 bleeds. For encounters with platelets <10 x109/L (n=12), 11 stopped AC and there were 2 thrombotic events; AC was continued for 1 patient without any subsequent grade 2 bleeds.

Conclusions : For patients with ITP who had platelets <50 x109/L and who were on AC, grade 2 bleeding events occurred during 20% of encounters. When AC was held, 27% of encounters resulted in thrombotic events; when AC was continued, no subsequent grade 2 bleeds were reported. These results suggest that for some patients with ITP, stopping AC can result in more harmful outcomes than continuing AC. Further studies are needed to define safe platelet count thresholds for AC in patients with ITP.


Arnold: Novartis: Consultancy, Research Funding; UCB: Consultancy; Dova: Consultancy; Amgen: Consultancy, Research Funding; Bristol Myers Squibb: Research Funding; Rigel: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution