Venous Thromboembolism among 1,570 Patients Undergoing Hematopoietic Stem Cell Transplantation: Incidence, Risk Factors, Treatment and Outcomes.

Introduction: Venous thromboembolism (VTE) is increasingly diagnosed among individuals with hematologic malignancies. It is unclear whether patients undergoing hematopoietic stem cell transplantation (HSCT) are at risk for VTE. We aimed to describe the incidence of VTE and risk factors among patients undergoing HSCT.

Methods: Using ICD-9 procedure codes in a discharge database, we identified all patients undergoing inpatient HSCT at Johns Hopkins between July 1, 1993, and June 30, 2005. We reviewed electronic medical records for each patient to identify VTE and bleeding. The associations between VTE and clinical characteristics were tested using parametric and non-parametric statistical tests and survival analysis methods.

Results: 1,570 patients had 3,425 person-years of follow-up. Median age was 46 years and 51% of patients were male. All patients had indwelling central venous catheters; pharmacological VTE prophylaxis was not used for any patient but ambulation was encouraged. Between admission and 180 days after HSCT, 75 VTE occurred in 70 patients (4.5%; 95% CI, 3.5–5.6%). VTE included 55 (73%) catheter-associated and 11 (15%) non-catheter-associated deep venous thromboses (DVT), and 9 (12%) pulmonary emboli (PE). Median platelet count at time of VTE was 77 K/mm³ (interquartile range [IQR] 42–159 K/mm³), and platelet count was < 50 K/mm³ at the time of VTE in 31% of cases. In multivariate analyses, VTE was associated with a history of prior VTE (OR 2.86; 95% CI, 1.25–6.55) and with graft-versus-host-disease (GVHD) (OR 2.36; 95% CI, 1.38–4.04). 86% of patients received anticoagulation for treatment. Clinically significant bleeding occurred in 238 patients after HSCT (15%), of whom 56 patients had fatal bleeding. Bleeding was associated with GVHD (OR 2.32; 95% CI, 1.71–3.16) and with initiation of anticoagulation for VTE diagnosed after HSCT (OR=3.00; 95% CI, 1.73–5.21). Bleeding was not associated with the continuation of anticoagulation initiated prior to admission (OR=1.13; 95% CI, 0.54–2.35).

Conclusions: VTE occurs relatively infrequently among patients undergoing HSCT and primarily in association with central venous catheter use. In contrast, clinically significant bleeding is a relatively common complication of HSCT, even in the absence of pharmacologic VTE prophylaxis. The unique hemostatic milieu associated with HSCT should be carefully considered when contemplating routine VTE prophylaxis in this patient population. The low risk of VTE associated with the high risk of bleeding makes pharmacologic VTE prophylaxis in this population both unnecessary and hazardous.