240. The Clinical Utility of Molecular Testing in the Diagnosis and Management of Infectious Diseases: Plasma-based Next-Generation Sequencing (PNGS)

Abstract

BackgroundMolecular diagnostic tests can provide microbiologic results rapidly and with greater sensitivity than traditional methods. However, these tests come with considerable costs, so thoughtful diagnostic stewardship is essential to ensure that resources and outcomes are optimized. We sought to evaluate the impact of PNGS testing on patient management.MethodsFrom February 2017 to January 2019, physicians in our group ordered 164 PNGS tests (Karius, Redwood City CA) on 125 patients. A retrospective chart review was performed to determine the clinical indication and utility of the test.ResultsThe assay failure rate was 4.9% (8/164). Positive (pos) results were noted in 34% (53/156), of which 23 (43.4%) represented false pos results; 28 were true pos (52.8%) but 2 were unnecessary (also had pos blood cultures). The most common reason for testing was to assess for Mycobacterium chimera (Mc) infection, representing 94 of 156 (60.3%) tests. Of the 21 patients with known Mc, only 10/21 had pos initial tests (47.6%); if patients with Mc localized to the sternum were excluded (8 patients), 76.9% with deep organ involvement had pos initial tests. Five patients with deep Mc infection had persistently pos results while on therapy; 4 of these had not had surgery; 1 was 6 months s/p valve replacement for Mc. The next most common indication was to r/o endocarditis in 18/156 (11.5%) and had an impact in 8/18 (44.4%), including 4 patients whose PNGS result identified a likely pathogen in culture negative endocarditis (CNE). Of the 62 tests done for non-Mc patients, 33.9% (21/62) were useful for management decisions. Among patients who eventually had a diagnosis made but had negative PNGS results included patients with Whipple’s (1), CNS infection (2), and fungal infections (5).ConclusionOverall, only 17.9% (28/156) of tests yielded true pos results. The most common reason was to evaluate for Mc infection. PNGS did not detect Mc in patients with proven local disease and was pos in >75% with deep/disseminated disease. However, a negative result did not exclude significant Mc infection. Repeat testing can be considered if clinical suspicion is high but should not be done before standard blood cultures are negative. While more than 60% of the non-Mc tests were not clinically useful, there was modest added utility where infection is high on the differential especially patients with CNE.Disclosures All authors: No reported disclosures.