Sp. Kantrow et al., IDIOPATHIC-PNEUMONIA-SYNDROME - CHANGING SPECTRUM OF LUNG INJURY AFTER MARROW TRANSPLANTATION, Transplantation, 63(8), 1997, pp. 1079-1086
Background. The aim of our study was to describe the incidence, clinic
al course, and risk factors for the idiopathic pneumonia syndrome (IPS
), compared with those previously described for ''idiopathic pneumonia
,'' after bone marrow transplantation (BMT). Methods. Our study design
was a case-series review with determination of risk by comparison wit
h unaffected controls by log-rank or Fisher's exact (two-tailed) test
and logistic regression analyses. The study group comprised 1165 conse
cutive marrow recipients at a single center from 1988 to 1991. Results
. IPS was documented in 85 BMT recipients (7.3%) by bronchoalveolar la
vage (n=68), open lung biopsy (n=3), or autopsy (n=14). The calculated
actuarial incidence for IPS within 120 days after BRIT was 7.7%. Medi
an time to onset was 21 days (mean 34+/-30). Hospital mortality was 74
%, and 53 BMT recipients (62%) died with progressive respiratory failu
re. IPS resolved in 22 patients (26%); 18 patients (21%) survived to d
ischarge. Mechanical ventilation was required by 59 BMT recipients (69
%), within a median of 2 days of onset of infiltrates. Two of these 59
recipients (3%) survived to discharge. Pulmonary infection (predomina
ntly fungal) was noted in 7 of 25 (28%) BMT recipients who had an auto
psy. Potential risk factors for IPS were assessed in univariate and mu
ltivariate logistic regression analyses. Although the incidence was no
t significantly different between autologous (5.7%) and allogeneic mar
row recipients (7.6%), risks were identified only for the latter: mali
gnancy other than leukemia (odds ratio=6.5 compared with aplastic anem
ia), and grade 4 graft-versus-host disease (odds ratio=5.4 compared wi
th lower grades). No factors were associated with recovery. Conclusion
s. The incidence of idiopathic lung injury seems lower, the onset earl
ier, and the risk factors different from those previously reported. Th
e major risks seem to be regimen-related toxicity and multi-organ dysf
unction associated with alloreactive processes.