Research Article: PCR with electrospray ionization-mass spectrometry on bronchoalveolar lavage for detection of invasive mold infections in hematological patients

Date Published: February 22, 2019

Publisher: Public Library of Science

Author(s): Anders Krifors, Volkan Özenci, Måns Ullberg, Malin Ackefors, Martin Jädersten, Kristoffer Strålin, Ola Blennow, Vishnu Chaturvedi.

http://doi.org/10.1371/journal.pone.0212812

Abstract

Invasive mold infections are life-threatening complications in patients with hematological malignancies. Conventional microbiological methods for diagnosing invasive pulmonary mold infections have low sensitivity, and molecular methods are being developed. Detection of molds using PCR with a narrow spectrum has been reported, but data with broad-spectrum PCR are lacking. In this study, the diagnostic performance and utility of a broad-spectrum PCR (broad-spectrum PCR with subsequent electrospray ionization-mass spectrometry, PCR/ESI-MS) for detection of molds in bronchoalveolar lavage (BAL) in 27 hematological patients with a new pulmonary infiltrate was analyzed. Using the revised EORTC/MSG criteria, PCR/ESI-MS was the only positive microbiological test in patients with proven invasive mold infection (n = 2) and correctly identified all cases of probable invasive pulmonary aspergillosis (n = 5). In patients with a possible invasive mold infection (n = 5), PCR/ESI-MS was positive in three patients. Mucorales was identified with PCR/ESI-MS in four patients that were all culture negative. The PCR/ESI-MS results had a clinical impact on antifungal therapy in 12 (44%) of the patients: modification of treatment in 6 (22%) patients and discontinuation in 6 (22%) patients. This study provides proof of concept that routine use of a broad-spectrum PCR for molds in bronchoalveolar lavage in immunocompromised patients is sensitive, fast, and has an impact on clinical decision-making

Partial Text

Invasive mold infections are life-threatening complications in patients with hematological malignancies, and adequate early treatment has been shown to be essential for a successful outcome [1,2]. Identification of non-aspergillus molds, especially molds from the order Mucorales causing mucormycosis, is of particular concern because of their intrinsic resistance to antifungal agents and the often aggressive course of the infections [3,4]. Conventional microbiological methods for detection of molds, i.e. microscopy and culture, are limited by low sensitivity and unacceptable long turn-around time and more sensitive tests targeting biomarkers have been developed. However, of the tests most commonly used today, galactomannan (GM) and Aspergillus PCR can only detect Aspergillus spp. and β-D-Glucan, although having a broader mold spectrum, cannot detect Mucorales [5]. Thus, new broad-spectrum and molecular-based, diagnostic tests are much needed. One such test is broad-spectrum PCR with subsequent electrospray ionization-mass spectrometry (PCR/ESI-MS) which can identify more than 200 fungal species with a turnaround time of under 7 hours [6–10]. Between February 2016 and May 2017, PCR/ESI-MS was available as a routine diagnostic method at Karolinska University Hospital, as one of only two hospitals worldwide. In the present study, the diagnostic performance and clinical utility of routine use of a broad-spectrum PCR for detection of molds in bronchoalveolar lavage (BAL) in hematological patients with a new pulmonary infiltrate was evaluated. We hypothesized that PCR/ESI-MS would be a sensitive and fast technique that would also have an impact on clinical decision making.

This retrospective cohort study includes all hematological patients at Karolinska University Hospital Huddinge Stockholm, Sweden, that had a diagnostic bronchoscopy performed between February 2016 and May 2017 due to a new pulmonary infiltrate diagnosed by computer tomography. During this period PCR/ESI-MS was included as an optional routine test when a diagnostic BAL was performed. If a patient had more than one BAL performed, only the results from the first BAL were included. Low volume BAL was performed using a standard protocol of 10–20 mL of 0.9% saline solution inserted and subsequently retrieved for analysis. Microscopy, bacterial and fungal cultures, and GM were performed in all patients while other diagnostic tests were ordered per the discretion of the referring physician. All microbiological analyses presented were conducted at the Department of Clinical Microbiology, Karolinska University Laboratory, Karolinska University Hospital, Stockholm. GM testing was performed using the Bio-Rad Platelia kit per manufacturers´ instructions. A positive cut-off level of 0.5 for serum and 1.0 for BAL fluid was used per the manufacturers’ recommendation. Aspergillus PCR was performed using a method compliant with the European Aspergillus PCR initiative guidelines [11,12]. The method is a qualitative DNA detection by amplification of a 483–505 bp fragment (depending on the pathogen) of the 18S rRNA gene of Candida and Aspergillus spp. with singleplex real-time PCR. Hybridization probes labeled with LC670 and LC640, respectively, allow PCR product detection and subsequent melting curve analyses.

In total, 36 patients had at least one BAL performed during the study period. The most common underlying hematological diseases were acute myeloid leukemia (n = 15) and myelodysplastic syndrome (n = 8) (Fig 1). The median age was 61 years (range 21–77). During the week preceding the bronchoscopy, 22 patients (61%) an absolute neutrophil count of <0.5 x 109/L at least once, of which 13 patients (36% of all patients) had <0.1 x 109. Seventeen patients (47%) received mold-active prophylaxis or empiric treatment at the time of BAL (Fig 1). Microscopy, fungal culture, and GM were performed on all BAL samples (n = 36), PCR/ESI-MS PCR was performed on 75% (n = 27), and Aspergillus PCR on 47% (n = 17) of the BAL samples. Serum GM was performed in 47% (n = 17) of the cases (Fig 1). There was no significant difference in proven or probable invasive mold infections between patients in whom PCR/ESI-MS was performed (7/27; 26%) and not performed (1/9;11%; p = 0.65, Fischer´s exact test). In this retrospective cohort study, the routine use of PCR/ESI-MS on BAL in hematological patients was found to have an impact on antifungal treatment in 44% of patients. Because of the low sensitivity of conventional microbiological tests focus has shifted to PCR-based techniques for identifications of non-Aspergillus molds. Several studies reporting successful development of non-Aspergillus PCR have been published, but the tests have suffered the disadvantage of being in-house analyses and/or having a narrow diagnostic spectrum [14]. In the present study, use of a standardized broad-spectrum PCR was able to accurately identify all five cases of probable IA, as well as four otherwise undetected cases of non-Aspergillus molds belonging to the order Mucorales. An important advantage of PCR-based methods compared to culture-based methods is the potential to identify molds after empirical antifungal therapy has been initiated, a situation which is common in clinical practice. Interestingly, PCR/ESI-MS could identify clinically important molds in three of the five patients that were already receiving mold-active treatment and had lung filtrates compatible with an invasive mold infection.   Source: http://doi.org/10.1371/journal.pone.0212812

 

Leave a Reply

Your email address will not be published.