PULMONARY HEMORRHAGE AS A CAUSE OF DEATH IN ALLOGENEIC BONE-MARROW RECIPIENTS WITH SEVERE ACUTE GRAFT-VERSUS-HOST DISEASE

Citation
Kj. Wojno et al., PULMONARY HEMORRHAGE AS A CAUSE OF DEATH IN ALLOGENEIC BONE-MARROW RECIPIENTS WITH SEVERE ACUTE GRAFT-VERSUS-HOST DISEASE, Transplantation, 57(1), 1994, pp. 88-92
Citations number
26
Categorie Soggetti
Immunology,Surgery
Journal title
ISSN journal
00411337
Volume
57
Issue
1
Year of publication
1994
Pages
88 - 92
Database
ISI
SICI code
0041-1337(1994)57:1<88:PHAACO>2.0.ZU;2-8
Abstract
Acute graft-versus-host disease (GVHD) has recently been associated wi th endothelial cell injury. The potential clinical significance was ex plored here in an autopsy review, Thirty-seven allogeneic bone marrow recipients were identified in the autopsy files at The Johns Hopkins H ospital with no evidence of systemic infection. Forty-one percent (15/ 37) of these patients were found to have extensive recent pulmonary he morrhage at autopsy which was thought to have led to terminal respirat ory failure and death. The 37 patients were divided into 2 groups: tho se with significant acute GVHD (stage 2 or greater) and those without GVHD (stage 0 or 1). Fifty-nine percent (10/17) of the patients with s ignificant acute GVHD died of acute respiratory failure due to recent pulmonary hemorrhage as opposed to 25% (5/20) of those without acute G VHD (P=0.032, Fisher's exact test). Terminal pulmonary hemorrhage was also associated with preparation for BMT, with 67% (12/18) of those pr epared with total body irradiation (TBI) having pulmonary hemorrhage a s opposed to 15% (3/19) of those prepared with chemotherapy using Busu lphan (P=0.002). There was no significant difference in posttransplant survival, engraftment, or final platelet count between the patients s tratified by GVHD or preparative protocol. The data support a strong a ssociation between significant acute GVHD and terminal hemorrhage, as well a possible association between TBI and pulmonary hemorrhage. Anal ysis of variance demonstrates that GVHD and TBI are independently asso ciated with increased pulmonary hemorrhage (P<0.01 for GVHD, P<0.001 f or TBI). We propose that GVHD contributes to terminal pulmonary hemorr hage by injuring the endothelium. However, this association could also be a secondary effect, i.e., toxicity from therapy for GVHD, or an ab normality in cytokines or growth factors. The pathogenic relationship between significant GVHD and terminal hemorrhage is discussed briefly.