Reply to Herbrecht et al

Raoul Herbrecht, Blandine Guffroy, François Danion, Aïna Venkatasamy, Célestine Simand, and Marie-Pierre Ledoux Department of Hematology, University Hospital of Strasbourg, Strasbourg, France, Institut de Cancérologie Strasbourg Europe (ICANS), Strasbourg, France, Université de Strasbourg, Institut National de la Recherche Médicale, Unité Mixte de Recherche S1113/Interface de Recherche Fondamentale et Appliquée en Cancérologie, Strasbourg, France, Department of Infectious Diseases, University Hospital of Strasbourg, Strasbourg, France, Department of Radiology, University Hospital of Strasbourg, Strasbourg, France, and Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC), Illkirch, France


Reply to Herbrecht et al
To the Editor-We thank Herbrecht and colleagues for their valuable comments regarding application of the updated radiologic criteria of the revised and updated consensus European Organization for Research and Treatment of Cancer and the Mycoses Study Group Education and Research Consortium (EORTC/ MSGERC) definitions for invasive fungal disease [1]. The authors point out that an important limitation of the first definitions in 2002 [2] and the 2008 iteration [3] was an overly strict definition of radiologic signs of pulmonary disease. In the most recent version of the definitions, the finding of a wedge-shaped, lobar or segmental consolidation was added to the existing criteria of dense, well-circumscribed lesion(s) with or without a halo sign, air crescent sign, or cavity, which, together with a host criterion and mycologic evidence, constitute probable invasive aspergillosis [1].
Herbrecht et al report the radiologic findings in a cohort of 727 patients with proven or probable invasive pulmonary aspergillosis, of whom 621 had initial computed tomographic (CT) imaging available. Using the revised radiologic definitions, 1 or more wedge-shaped, lobar or segmental consolidations were present in 23.7% of patients. Moreover, nearly one-third of patients with proven invasive aspergillosis presented with a consolidation pattern but without a nodule on initial CT scan. Indeed, nonneutropenic patients were more likely to have consolidation than neutropenic patients.
We applaud the authors for producing these data so soon after the EORTC/ MSGERC definitions were published and welcome the fact that their findings support the radiologic definitions proposed [1]. Determining appropriate radiologic criteria for invasive fungal disease is challenging, as the radiologic lesions are nonspecific, especially among nonneutropenic patients [4,5]. During our deliberations, we considered other radiographic findings such as bronchial thickening with tree-in-bud lesions, ground glass opacities, micronodules, and pleural effusions as diagnostic criteria. However, the consensus was to exclude these because of their lack of specificity for pulmonary aspergillosis. Nevertheless, more data such as those presented by Herbrecht et al are clearly needed as they will be instrumental in helping reevaluate the radiologic criteria and inform future revisions of the EORTC/MSGERC definitions.

Notes
Disclaimer. The content is solely the responsibility of the authors and does not necessarily represent the official views of their institutions.
Potential conflicts of interest. S. C.-A. C. reports grants from Meck Sharp & Dohme (MSD) Australia, outside the submitted work. J. P. D. reports personal fees from F2G Ltd, Gilead Sciences, and Pfizer, outside the submitted work. C. O. M. reports grants from Gilead Sciences and MSD, and advisory board fees from MSD, outside the submitted work. All other authors report no potential conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. First, in at least 20 instances of positive tests, mNGS testing confirmed results obtained through conventional microbiologic testing. In some of the cases in which new organisms were identified, therapy (either empiric or targeting another identified organism) that would be expected to cover the identified organism was continued. Is this a failure of the assay, or a failure of appropriate application of the assay? Second, in several instances of negative tests, other infectious testing was also negative, and a noninfectious diagnosis was subsequently made. In others, a negative test occurred after empiric treatment had been started and the patient was improving. Prior studies have shown the test is often negative in this setting [3]. Is it a limitation of the mNGS test to be negative in the setting of a noninfectious diagnosis? Should it have ever been sent on an improving patient on empiric treatment?
Any medical intervention, whether it be a test, a drug, or a procedure, should involve a careful assessment by the treating physician as to whether it can lead to some benefit to the care of the patient. Too often, however, providers become enamored with the "new thing, " whether it be a drug, a test, or some other novel intervention. They may want to be the first to use it, or the first on their team to suggest it be used, potentially leading to lack of appropriate critical thinking about whether it is the correct thing to do for a given patient.
We agree with Babady that diagnostic stewardship should be considered for any testing, particularly when it applies most specifically to a given subdiscipline. However, appropriate stewardship of medical resources should be part of every physician's job, and any medical intervention should be undertaken with careful consideration of the possible outcomes and within the expertise of the provider, independent of cost. The higher reported utility of mNGS testing at our center [4] is likely heavily influenced by the active role ID takes in overseeing the use of the test, and includes both careful patient selection and proper timely utilization of the test. Juice is only "worth the squeeze" when someone first puts some thought into selecting the fruit.

Reply to Muller and Chaudhury
To the Editor-We thank Drs Muller and Chaudhury for sharing their perspectives on our study evaluating the clinical impact of plasma metagenomic nextgeneration sequencing (mNGS) [1,2]. There are several reasons why our studies reached diverging conclusions. First, in our study, we applied impact criteria based on the treating team's evaluation of clinical utility. Given that plasma mNGS is currently a send-out test and results are available after conventional microbiological tests, most providers did not consider mNGS to provide additional value when it simply confirmed conventional results. This contrasts with the definition of "clinical relevance" in their study, which included cases where mNGS confirmed conventional positive and negative test results rather than producing a new positive result or diagnosis [2,3]. Indeed, had we applied our criteria across both studies, the overall conclusions would likely have been similar. Future studies