The significance of pulmonary infiltrates developing in patients receiving granulocyte transfusions

Abstract

Summary. We reviewed a series of 109 patients given granulocyte transfusions between 1974 and 1978, to determine if pulmonary infiltrates developing during granulocyte transfusion therapy carried any specific prognostic significance. Eighteen patients developed new infiltrates while receiving granulocytes. Six of these patients died during the acute episode, an overall mortality rate no different from patients not developing new infiltrates. However, the subgroup with pulmonary infiltrates who died had some special features including: infiltrates that developed between 7 and 21 d after transfusions were begun; infiltrates that progressed or persisted during the period of transfusion; and infiltrates that were local rather than diffuse. Autopsies on five of the six patients who died revealed disseminated fungal infection in three, pulmonary haemorrhage in one, and alveolar hyaline membranes in association with Pseudomonas sepsis in one. Besides this high‐risk group specific aetiologies could rarely be assigned for other more diffuse or earlier/later developing infiltrates, despite usual diagnostic studies.We conclude that local infiltrates developing more than 1 week after transfusions are begun are life threatening; they are often caused by resistant infectious organisms and would seem to represent a failure of granulocyte therapy. Aggressive diagnostic and therapeutic interventions are warranted in patients developing this kind of infiltrate while receiving granulocytes. Conversely, pulmonary infiltrates occurring early or longer than 3 weeks after granulocyte tranfusions are begun are not clinically ominous and may be a manifestation of the beneficial effect of granulocyte therapy.