Person:

Koo, Sophia

Abstract Background: Timely diagnosis of IM remains a major challenge in clinical mycology. Because of the lack of specific diagnostic methods for IM and the frequently fulminant nature of this infection, IM-associated mortality remains high. Methods: We examined breath volatile metabolite profiles in a neutropenic murine model of IM, using the 3 Mucorales species that most commonly cause human IM - Rhizopus arrhizus var. arrhizus, R. arrhizus var. delemar, and R. microsporus - and for comparison, Aspergillus fumigatus. We infected female balb/c mice (N = 4 per group) treated with cyclophosphamide and cortisone followed by intranasal administration of 106 conidia of each species. 3 days post-infection, we collected breath samples from each mouse via tracheostomy using a flexiVent murine ventilator, examining breath volatile metabolites using thermal desorption gas chromatography/tandem mass spectrometry (GC-MS/MS). We also sampled breath prospectively from five patients eventually diagnosed with proven IM caused by R. microsporus, analyzing breath volatile metabolites using thermal desorption GC-MS/MS. Results: Each Mucorales species produced a consistent profile of breath sesquiterpene secondary metabolite VOCs in our murine models, which distinguished these species from each other and from murine invasive aspergillosis (Figure A). These fungi shifted their secondary metabolism significantly in vivo, compared with their previously characterized in vitro metabolism. We found overlapping VOC sesquiterpene metabolites between breath samples from the murine model of R. microsporus infection and 5 of 5 patients with R. microsporus IM, with additional sesquiterpene secondary metabolites detected in patient breath, compared with the murine IM model (Figure B). In one patient with serial breath samples, these sesquiterpenes declined in abundance and disappeared with antifungal therapy, in parallel with clinical improvement (Figure C). Conclusion: The three Mucorales species that cause most human IM have distinct breath sesquiterpene profiles that can be used to identify these infections in vivo noninvasively. These profiles distinguish these infections from each other and from aspergillosis, and may be useful in monitoring clinical response to treatment. Disclosures F. M. Marty, Astellas Pharma US: Consultant and Grant Investigator, Consulting fee and Grant recipient; Chimerix: Consultant and Grant Investigator, Consulting fee and Grant recipient; Fate Therapeutics: Scientific Advisor, Consulting fee; Gilead Sciences: Consultant and Grant Investigator, Consulting fee and Grant recipient; LFB: Consultant, Consulting fee; Merck: Consultant, Grant Investigator and Scientific Advisor, Consulting fee and Grant recipient; Roche Molecular Systems: Consultant, Consulting fee; Shire: Consultant and Grant Investigator, Consulting fee and Grant recipient; D. P. Kontoyiannis, Pfizer: Research Contractor, Research support and Speaker honorarium; Astellas: Research Contractor, Research support and Speaker honorarium; Merck: Honorarium, Speaker honorarium; Cidara: Honorarium, Speaker honorarium; Amplyx: Honorarium, Speaker honorarium; F2G: Honorarium, Speaker honorarium

Abstract Background: Invasive aspergillosis (IA) is the most common invasive mold infection (IMI) and early diagnosis is critical to improving clinical outcomes. Galactomannan (GM) is a major component of the Aspergillus cell wall. BAL GM is one of the mycologic criteria for diagnosis of probable IA, but it is frequently positive in patients with Aspergillus airway colonization, and its specificity has not been well-studied. Our goal was to estimate the specificity of a positive BAL galactomannan value in a contemporary cohort of consecutive patients with BAL GM checked as part of their workup for potential IA. Methods: We reviewed clinical and microbiologic data of patients who had at least one positive BAL GM (≥0.5), at Brigham and Women’s Hospital from November 2009 to March 2016. We applied EORTC/MSG IMI definitions to classify patients as having possible, probable or proven IMI, excluding BAL GM result as mycologic criterion. Results: We studied 134 patients. Median age was 58; 49% were women. 54% had hematologic malignancy (HM), 10% were solid organ (SOT) and 34% hematopoetic stem-cell transplant (HSCT) recipients. 60% received mold-active antifungal treatment. 4 patients (3%) had proven, 60 (45%) probable, 15 (11%) possible, and 55 (41%) no IMI. One had proven mucormycosis, making at least 42% of positive BAL GM results falsely positive (specificity 58%, 95% confidence interval 47–69%). 6-week mortality was 35% overall: 75% for proven, 47% probable, 33% possible, and 20% for no IMI (χ2 for trend P = 0.008). In patients with no IMI, 6-week mortality was comparable in those who did not receive mold-active treatment (13%, 5/38) and those who did (38%, 6/16, Fischer’s P = 0.07). Conclusion: In this study, at least 42% of positive BAL GM values were falsely positive, potentially exposing these patients to unnecessary antifungals. The number of ‘probable’ IMI cases (which, along with proven IMI are typically included in clinical trials of new antifungals) would be falsely increased by 25%, using a positive BAL GM alone to adjudicate IMI status. Accurate noninvasive tests for diagnosis of IMI are urgently needed. Table: % +BAL GM values in patients with no IMI at various cutoff values. BAL GM ≥ 0.5 BAL GM ≥ 0.8 BAL GM ≥ 1.0 Overall 42 22 17 HM 28 14 7 SOT 100 50 50 HSCT 30 11 7 Disclosures All authors: No reported disclosures.