We evaluated the performance characteristics of the FilmArray respiratory panel and the eSensor respiratory viral panel on clinical and spiked lower respiratory tract specimens (LRTS). The overall agreement between the two methods was 89.5% (51/57). The lower limit of detection of both assays for all targets in LRTS was comparable to that for nasopharyngeal swab specimens.
Lower respiratory tract infections, especially pneumonia, can be particularly severe and a cause of high morbidity and mortality in cancer and hematopoietic stem cell transplant (HSCT) patients (1). In HSCT recipients, progression of respiratory viral infections from the upper to the lower respiratory tract occurs most frequently with respiratory syncytial virus (RSV), human metapneumovirus (hMPV), parainfluenza viruses (PIV), and influenza viruses (2, 3). Two of the several multiplexed molecular assays cleared by the Food and Drug Administration (FDA) for the detection of respiratory pathogens include the FilmArray respiratory panel (FA RP) test (BioFire Diagnostics, Salt Lake City, UT) and the eSensor respiratory virus panel (eS RVP) test (GenMark Diagnostics, Carlsbad, CA). However, both assays are cleared only for testing on nasopharyngeal swab (NPS) specimens. A few studies evaluating both the FA RP and the eS RVP have included a small number of other specimen types, including bronchoalveolar lavage (BAL) fluids, bronchial washes (BW), sputum (SPT), throat swabs, and tracheal aspirates (4-8). In this study, we compare the performances of both the FA RP and the eS RVP exclusively with lower respiratory tract specimens (BAL fluids, BW, and SPT) from cancer patients with symptoms of pneumonia. In addition to accuracy, the lower limit of detection (LOD) of the assays for each target in BAL fluids was compared to the LOD for NPS specimens.This was a retrospective study performed with specimens collected from July 2010 to March 2013 and stored at Ϫ80°C following testing by direct fluorescent antibody assay and viral culture as previously described (4). A total of 52 samples (31 BAL fluid, 17 BW, and 4 SPT) from 45 patients were included: 32 consecutive positive samples and 20 randomly selected negative specimens. Patients' ages ranged from 25 to 89 years old, with 25 liquid-tumor patients (12 with leukemia and 13 with lymphoma) and 20 solidtumor patients (12 with lung cancer and 8 with other types of cancer). The most common symptoms included radiological evidence of lung abnormality (infiltrates or nodules) and/or fever and cough. The study was granted a waiver of the HIPPA authorization and informed consent by the Memorial Sloan-Kettering Cancer Center (MSKCC) institutional review board.To increase the number of positive specimens tested, BAL fluids (n ϭ 14) negative for all 20 targets detected by the FA RP were combined, split into 4 pools, and spiked with organisms from the Zeptometrix respiratory panel (Zeptometrix, Buffalo, NY). This panel includes all the targets listed in Table 1. The concentration of each viral stock solution varied fr...