Jonathan H Chung1, Gwen Huitt2, Kunihiro Yagihashi3, Stephen B Hobbs4, Anna V Faino5, Bradley D Bolster6, Jürgen Biederer7, Michael Puderbach8, David A Lynch1. 1. 1 Department of Radiology and. 2. 2 Department of Medicine, National Jewish Health, Denver, Colorado. 3. 3 Department of Radiology, St. Marianna University School of Medicine, Kanagawa, Japan. 4. 4 Department of Radiology, University of Kentucky, Lexington, Kentucky. 5. 5 Department of Biostatistics and Bioinformatics, National Jewish Health, Denver, Colorado. 6. 6 Siemens Medical Solutions USA, Inc., Salt Lake City, Utah; and. 7. 7 Radiologie Darmstadt, Gross-Gerau County Hospital and. 8. 8 Hufeland Klinikum GmbH, Abteilung für Diagnostische und Interventionelle Radiologie, Heidelberg, Germany.
Abstract
RATIONALE: Computed tomographic (CT) radiography is the reference standard for imaging Mycobacterium avium complex (MAC) lung infection. Magnetic resonance imaging (MRI) has been shown to be comparable to CT for characterizing other pulmonary inflammatory conditions, but has not been rigorously tested for imaging MAC pneumonia. OBJECTIVES: To determine the feasibility of pulmonary MRI for imaging MAC pneumonia and to assess the degree of agreement between MRI and CT for assessing the anatomic features and lobar extent of MAC lung infections. METHODS: Twenty-five subjects with culture-confirmed MAC pneumonia and no identified coinfecting organisms were evaluated by thoracic MRI and then by chest CT imaging performed up to 1 week later. After deidentification, first the MRI and then the CT scans were scored 2 weeks apart by two chest radiologists working independently of one another. Discrepancies were resolved by a third chest radiologist. The scans were scored for bronchiectasis, consolidation or atelectasis, abscess or sacculation, nodules, and mucus plugging using a three-point lobar scale (absent, <50% of lobe, and >50% of lobe). Agreement analyses and ordinary least products regressions were performed. MEASUREMENTS AND MAIN RESULTS: A fixed bias was found between total CT and MRI scores, with CT scoring higher on average (median difference: 4 on a scale of 48; interquartile range: 3, 6). Fixed biases were found for bronchiectasis and consolidation or atelectasis subscale scores. Both fixed and proportional biases were found between CT and MRI mucus plugging scores. No bias was found between CT and MRI nodule scores. There was nearly perfect lobar percent agreement for more conspicuous findings such as consolidation or atelectasis and abscess or sacculation. CONCLUSIONS: In this exploratory study of 25 adult patients with culture-proven MAC lung infection, we found moderate agreement between MRI and CT for assessing the anatomic features and lobar extent of disease. Given the feasibility of chest MRI for this condition, future work is warranted to assess the clinical impact of MRI compared with CT in assessing progression of untreated MAC infection and response to treatment over time.
RATIONALE: Computed tomographic (CT) radiography is the reference standard for imaging Mycobacterium avium complex (MAC) lung infection. Magnetic resonance imaging (MRI) has been shown to be comparable to CT for characterizing other pulmonary inflammatory conditions, but has not been rigorously tested for imaging MACpneumonia. OBJECTIVES: To determine the feasibility of pulmonary MRI for imaging MACpneumonia and to assess the degree of agreement between MRI and CT for assessing the anatomic features and lobar extent of MAClung infections. METHODS: Twenty-five subjects with culture-confirmed MACpneumonia and no identified coinfecting organisms were evaluated by thoracic MRI and then by chest CT imaging performed up to 1 week later. After deidentification, first the MRI and then the CT scans were scored 2 weeks apart by two chest radiologists working independently of one another. Discrepancies were resolved by a third chest radiologist. The scans were scored for bronchiectasis, consolidation or atelectasis, abscess or sacculation, nodules, and mucus plugging using a three-point lobar scale (absent, <50% of lobe, and >50% of lobe). Agreement analyses and ordinary least products regressions were performed. MEASUREMENTS AND MAIN RESULTS: A fixed bias was found between total CT and MRI scores, with CT scoring higher on average (median difference: 4 on a scale of 48; interquartile range: 3, 6). Fixed biases were found for bronchiectasis and consolidation or atelectasis subscale scores. Both fixed and proportional biases were found between CT and MRI mucus plugging scores. No bias was found between CT and MRI nodule scores. There was nearly perfect lobar percent agreement for more conspicuous findings such as consolidation or atelectasis and abscess or sacculation. CONCLUSIONS: In this exploratory study of 25 adult patients with culture-proven MAClung infection, we found moderate agreement between MRI and CT for assessing the anatomic features and lobar extent of disease. Given the feasibility of chest MRI for this condition, future work is warranted to assess the clinical impact of MRI compared with CT in assessing progression of untreated MACinfection and response to treatment over time.
Authors: Christopher Vinnard; Alyssa Mezochow; Hannah Oakland; Ross Klingsberg; John Hansen-Flaschen; Keith Hamilton Journal: Front Microbiol Date: 2018-11-20 Impact factor: 5.640