Investigation Of Fresh Tissue and Effusion Samples From Immunocompromised Hematologic Patients Suspected For Invasive Fungal Infection With An Aspergillus-specific PCR Is a Promising Tool For Identifying Invasive Aspergillosis

Abstract

Background Although invasive fungal infection (IFI), especially invasive aspergillosis (IA) is a frequent and lethal complication in severely immunocompromised hematologic patients (pts) such as allogeneic stem cell recipients, diagnosis still remains difficult and conventional diagnostic tools lack sensitivity. Examining tissue or effusion samples might help to identify the causative pathogen of the underlying infection. However, culture-based methods for IFI detection only yield results in a minority of patients. The addition of a specific polymerase chain reaction (PCR) directly in these clinical specimens might therefore improve the detection of IA. Thus the performance of an established Aspergillus-specific nested PCR in fresh tissue and effusion samples of immunocompromised pts was evaluated. Methods By means of an established Aspergillus-specific PCR protocol, we investigated 98 fresh tissue and effusion specimens from 74 immunocompromised, predominantly hematologic patients. The diagnosis of IFI was suspected due to underlying immunosuppression and radiologic abnormalities. Of these patients 23/74 (31%) had proven (n=17) or probable (n=6) IFI using the 2008 EORTC/MSG criteria. The remaining 51 (69%) pts were classified as having either possible (n=27) or having no IFI (n=24) due to lack of host criteria or missing specific radiologic features. IA was identified as the underlying IFI in 18/23 proven/probable patients. Results PCR positivity in tissue and effusion samples was observed in 16/18 proven/probable IA and 6 possible IA patients, while patients with no IA according to EORTC/MSG did not show positive PCR signals. In 5 patients with proven IFI, non-Aspergillus moulds (2 Mucorales spp. and 3 Rhizopus spp.) were found while Aspergillus PCR remained negative. Thus, Aspergillus PCR analysis for diagnosing IA yielded sensitivity/specificity values of 0.89 (95%CI 0.67-0.97) / 1.0 (95%CI 0.86-1.0) respectively, as well as a positive likelihood ratio of >200 and a negative likelihood ratio of 0.1 leading to a diagnostic odds ratio of >200. Conclusions In our multicenter analysis of samples from predominantly hematologic pts with suspected IFI, we observed a good diagnostic performance of the PCR assay for detection of IA directly in fresh tissue and effusion samples. Thus, PCR testing of these samples represents a sensitive and complementary tool for identification of the underlying infection: A positive tissue/effusion PCR result makes the presence of IA highly likely whereas IA is unlikely in case of a negative result, however this could indicate non-Aspergillus IFI. For the clinician, the additional PCR testing of these specimens in pts with suspected IFI improves the diagnostic armamentarium and might help in identifying the causative pathogen. Disclosures: Lehrnbecher: Gilead: Honoraria; MSD: Honoraria; Pfizer: Honoraria; Astellas: Honoraria.