Acute bleeding after bone marrow transplantation (BMT) was investigated in 1,402 patients receiving transplants at Johns Hopkins Hospital between January 1, 1986 and June 30, 1995. Bleeding categorization was based on daily scores of intensity used by the blood transfusion service. Moderate and severe episodes were analyzed for bleeding sites. Analysis of the cause of death and the interval of the bleeding episode to outcome endpoints was recorded. Survival estimates were computed for 1,353 BMT patients. The overall incidence was 34%. Minor bleeding was seen in 10.6%, moderate bleeding was seen in 11.3%, and severe bleeding was seen in 12% of all patients. Fourteen percent of patients had moderate or severe gastrointestinal hemorrhage, 6.4% had moderate or severe hemorrhagic cystitis, 2.8% had pulmonary hemorrhage, and 2% had intracranial hemorrhage. Sixty-one percent had 1 bleeding site and 34.4% had more than 1 site. Moderate and severe bleeding was more prevalent in allogeneic (31%) and unrelated patients (62.5%) compared with autologous patients (18.5%). Significant distribution of incidence was found among the different diagnoses, but not by disease status in acute myeloid leukemia, acute lymphoblastic leukemia, chronic myelogenous leukemia, Hodgkin's disease, and non-Hodgkin's lymphoma. Bleeding was associated with significantly reduced survival in allogeneic, autologous, and unrelated BMT and in each disease category except multiple myeloma. Survival was correlated with the bleeding intensity, bleeding site, and the number of sites. Although close temporal association was evident to mortality, bleeding was recorded as the cause of death in only the minority of cases compared with other toxicities after BMT (graft-versus-host disease, infections, and preparative regimen toxicity). Acute bleeding is a common complication after BMT that is profoundly associated with morbidity and mortality. Although bleeding was not a direct cause of death in the majority of cases, it has a potential prognostic implication as a predictor of poor outcome in clinical assessment of patients after BMT.

Our results support an association of bleeding with acute GVHD, suggesting that GVHD is a risk factor for bleeding after BMT. The occurrence of bleeding clearly identified poor outcome subgroup within GVHD, suggesting further evaluation for clinical application of bleeding in the assessment of GVHD severity.

Patients who bled were usually placed on a higher threshold before the onset of their major bleeding event and were not exposed to additional risk of bleeding from thrombocytopenia. Similarity in bleeding incidence between study periods appears to associate with adjustments to high-risk conditions and may not reflect consequences of the lower transfusion threshold.

Acute bleeding is a frequent complication that commonly associates with increased morbidity after bone marrow transplantation. Except for diffuse alveolar hemorrhage and cerebral hemorrhage, bleeding is infrequently recorded as a direct cause of death. Yet outcome analyses showed that bleeding from any reviewed site was associated with reduced survival. Reduced survival was correlated with bleeding intensity and the number of bleeding sites. These data point to the need to monitor all manifestations of bleeding, as bleeding may identify patients at risk for bone marrow transplantation toxicity. Until recently, prophylactic platelet transfusions were commonly given at a trigger of 20 x 10(9)/L. Whereas bleeding is more likely to occur when platelet counts drop to low levels, most bleeding episodes were recorded with platelet counts greater than 20 x 10(9)/L, suggesting causes other than profound thrombocytopenia in the pathogenesis of bleeding. Given that a trigger of 10 x 10(9)/L has become accepted for prophylactic platelet transfusions, care should be taken to ensure that parameters other than the incidence of bleeding have not been adversely affected.