2294. Evaluation of the Karius Plasma Next-Generation Sequencing Cell-free Pathogen DNA Test to Determine the Etiology of Infection and Impact on Anti-Microbial Management in Patients with Severe Neutropenia and Fever

BACKGROUND: Standard microbiological testing (MT) fails to identify a pathogen in most chemotherapy recipients with febrile neutropenia (FN), who therefore receive prolonged empiric courses of broad-spectrum antimicrobials (AM). We evaluated the ability of the Karius next-generation sequencing plasma test (KT) to identify infectious etiologies of NF and its impact on AM management. METHODS: This prospective, observational study enrolled 57 patients with ≤500 neutrophils/mm(3). Samples were collected within 24 hours of fever onset (T0) and every 2–3 days. Cell-free plasma DNA was prepared and sequenced in a CLIA/CAP laboratory, human reads excluded, and remaining sequences aligned to a curated pathogen database that includes bacteria, viruses, fungi and parasites. Positive agreement (PA) was defined as KT identification of ≥1 isolate also seen by blood culture (BC). Discordant results were adjudicated by 3 infectious disease specialists as: Definite: KT identified ≥1 organism also seen by MT± 7 days of enrollment; Probable: KT result was a likely cause of NF compatible with clinical diagnosis; Possible: KT result was consistent with an infection but not a common cause of NF. RESULTS: 56 results (55 subjects) with valid KT and BC results were analyzed. Compared with BC, KT had a PA of 90% (9/10) and negative agreement of 31% (14/45). KT identified >1 organism in 61% (25/41) of the cases. Definite (13), Probable (24) and Possible (4) cases were classified as True Positives. Using clinical adjudication, KT had a sensitivity of 98% (41/42) and specificity of 100% (14/14). The committee would have changed AM therapy 68% (27/40) of the time, had the KT results been available in real-time (~T52–100h). In 8/19 cases (42%) vancomycin would have been discontinued; in 6/27 cases (22%) and in 5/27 cases (19%), anaerobic coverage or antivirals would have been added earlier. Serial analysis of a Pneumocystis jirovecii infection indicated that earlier diagnosis and treatment may have prevented morbidity and eventual ICU transfer. CONCLUSION: The absence of infectious etiology in NF often leads to broad AM therapy or delay of targeted treatment. Given its sensitivity and ability to detect a breadth of pathogens, the KT can provide useful data for diagnosis and management of NF and may allow for optimization of AM therapy. DISCLOSURES: H. Seng, Karius, Inc.: Employee, Salary. R. Aquino, Karius, Inc.: Employee, Salary. D. Hollemon, Karius: Employee, Salary. D. Hong, Karius, Inc.: Employee, Salary. T. Blauwkamp, Karius, Inc.: Board Member, Employee and Shareholder, Salary. M. Kertesz, Karius Inc.: Board Member, Employee and Shareholder, Salary. L. Blair, Karius: Employee, Salary. S. Zompi, Karius, Inc.: Employee, Salary.