356. Bronchoalveolar Lavage Fluid Cytology by GMS Stain for the Diagnosis of Invasive Pulmonary Aspergillosis in Patients With Hematologic Malignancies: Analysis of 67 Episodes

BACKGROUND: The yield of direct fungal visualization by GMS (Gomori–methenamine–silver) stain in bronchoalveolar lavage (BAL) cytology has rarely been studied in the diagnosis of invasive pulmonary aspergillosis (IPA) in patients with hematological malignancies (HM). METHODS: We analysed a series of patients with proven or probable culture-documented IPA (EORTC/MSG criteria) in HM patients (1999–2015). All patients had BAL cultures that were positive for Aspergillus spp. and had concurrently obtained BAL cytology GMS available for analysis. RESULTS: We identified 67 such patients. BAL cytology based on GMS showed hyalohyphomycetes consistent with Aspergillus in 28/67 (41.8%) patients, whereas only in 2/67 (3.6%) direct smear Calcofluor White stain was positive. Based on BAL GMS cytology, co-infections were identified in six patients: two Pneumocystis and five viral infections with cytopathic changes (one had both). The yield of cytology was not different in patients with IPA caused by non-fumigatus Aspergillus, although patients with IPA and >1 Aspergillus in BAL culture had more often positive cytology GMS (100% vs. 0%, p 0.027). Cytology was also more often positive when obtained from a lesion-targeted BAL as compared with non-targeted bronchial washings (60.7% vs. 7.1%, p 0.038). Patients with IPA and cavitary lesions (32.1% vs. 5.1%, P = 0.006), history of SCT (64.3% vs. 33%, P = 0.015) or prior exposure to itraconazole (75% vs. 41%, P = 0.007) had positive cytology GMS results more often than did patients without these characteristics. In the multivariate analysis, only cavitary lesions were significantly associated with positive BAL GMS cytology. CONCLUSION: GMS stain in cytology of BAL in patients with HM and culture-documented IPA had a sensitivity of 41.8% and was more often positive in patients with cavitary lesions. Although there were no differences in the proportion of GMS-positive cytology rates among differing Aspergillus spp. causing IPA, mixed Aspergillus spp. IPA was associated with an increase in positive cytology. BAL cytology was diagnostic for co-infections in more than 10% of patients. BAL cytology should be part of the diagnostic wok up in HM patients with suspected IPA. DISCLOSURES: All authors: No reported disclosures.