Accuracy of Magnetic Resonance Imaging for Grading of Subglottic Stenosis in Patients with Granulomatosis with Polyangiitis: Correlation with Pulmonary Function Tests and Laryngoscopy

Abstract

OBJECTIVE: To compare magnetic resonance imaging (MRI)-based and laryngoscopy-based subglottic stenosis (SGS) grading with pulmonary function testing (PFT) in patients with granulomatosis with polyangiitis (GPA).

METHODS: In this retrospective study, we included 118 examinations of 44 patients with GPA and suspected SGS. All patients underwent MRI, laryngoscopy, and PFT. Stenosis was graded on a 4-point scale by endoscopy and MRI using the Meyer-Cotton (MC) score (score 1: ≤50%, 2: 51-70%, 3: 71-99%, and 4: 100%) and as percentage by MRI. Results were compared with peak expiratory flow (PEF) and maximum inspiratory flow (MIF) from PFT, serving as objective functional reference.

RESULTS: In MRI, 112 of 118 examinations (95%) were rated positive for SGS (grade 1 [n = 82], grade 2 [n = 26], and grade 3 [n = 4]), whereas in laryngoscopy 105 of 118 examinations (89%) were rated positive for SGS (grade 1 [n = 73], grade 2 [n = 24], and grade 3 [n = 8]). MRI and laryngoscopy agreed in 75 of 118 examinations (64%). MRI determined higher scores in 20 examinations (17%) and lower scores in 23 examinations (19%) compared to laryngoscopy. MC scores as determined by both MRI and laryngoscopy showed comparable correlations with PEF (r = -0.363, P = 0.016, and r = -0.376, P = 0.012, respectively) and MIF (r = -0.340, P = 0.024, and r = -0.320, P = 0.034, respectively). The highest correlation was found between MRI-based stenosis grading in percentage with PEF (r = -0.441, P = 0.003) and MIF (r = -0.413, P = 0.005).

CONCLUSION: MRI and laryngoscopy provide comparable results for grading of SGS in GPA and correlate well with PFT. MRI is an attractive noninvasive and radiation-free alternative for monitoring the severity of SGS in patients with GPA.

Bibliographical data

Original languageEnglish
ISSN2151-464X
DOIs
Publication statusPublished - 05.2018
PubMed 28772006