Background: Serum galactomannan (GM) antigen is routinely used to aid the diagnosis of invasive aspergillosis, but the sensitivity of the test is reduced in recipients of allogeneic HSCT (allo-HSCT) and with antifungal prophylaxis.
Methods: We examined the sensitivity and specificity of a pulmonary vascular occlusion sign visualized by CTPA in allo-HSCT patients with a diagnosis of possible, probable or proven invasive mold disease (IMD) according to EORTC/MSG definitions. Patients who received allo-HSCT (n=20) for AML/MDS (n=8), ALL (n=7), lymphoma (n=1), or chronic myelodysplastic syndromes (n=4) underwent CTPA if they had a lesion on initial CT scans suggestive of IMD. IMD diagnosis was subsequently confirmed on the basis of culture, histopathology and/or antigen results.
Results: CTPA was positive in 16/20 of patients, 13 of whom eventually met EORTC/MSG definitions for proven or probable IMD (12 Aspergillus, 1 Mucormycosis + Aspergillus). The remaining 3/20 of CTPA positive patients had possible IMD. Among the 4 CTPA negative patients, 3 had bacterial pneumonia, and 1 patient had probable IMD based on GM antigen and CT. In CTPA positive patients, GM antigen was positive in serum of 12/16, and in the bronchial alveolar lavage (BAL) fluid of 4/4 patients. Considering only proven or probable IMD, a positive CTPA finding had a sensitivity (92%), specificity (50%), positive predictive value (81%) and negative predictive value (75%) for IMD.
Conclusions: CTPA may be an alternative to serum or BAL GM antigen testing for establishing the diagnosis of probable or proven IMD in allo-HSCT patients.
Full conference title:
- EBMT 39th (2013)