18-FDG PET in the Evaluation of Therapeutic Response of Necrotizing Otitis Externa.

Abstract

The duration of antibiotic treatment for necrotizing otitis externa (NEO) and the criteria for discontinuation are currently uncodified. Our study evaluated the performance of [18F]FDG PET to assess the therapeutic response of NEO. We retrospectively included patients treated for NEO who underwent a PET scan at baseline and 4 to 9 weeks after antibiotic initiation. PET images were analyzed visually and quantitatively (SUVmax, Total Lesion Glycolysis [TLG], and Total Metabolic Volume [TMV]). An intensity-weighted extent score (extent) that we created was also calculated. For each parameter, delta values ("∆": percentage change from baseline for each parameter) were also calculated. Thirty-nine patients and 40 ears were enrolled. Pseudomonas aeruginosa was isolated from 32 (86%) samples. Twenty-nine (72.5%) patients were treated with CEFTAZIDIME and CIPROFLOXACIN for a median of 6.9 weeks. On PET scan re-evaluation, 15 (27.5%) patients were visually in complete response, 17 (42.5%) in partial response, 5 (12.5%) stable, and 3 (7.5%) in progression. The sensitivity and specificity of visual analysis for the diagnosis of recovery were 100% and 43%, respectively. The best performing parameters for predicting recurrence were initial TLG and TMV (sensitivity and specificity of 80% and 100% for respective thresholds of 50.1 g and 8.3 mL). ∆extent showed sensitivity and specificity of 80% and 97.1% using a threshold of -4.3%, and ∆TLG showed sensitivity and specificity of 80% and 94.3% using a threshold of -37.4%. PET performed well in assessing the therapeutic response of NEO, with excellent sensitivity but low specificity of visual analysis. A complementary quantitative analysis was useful to distinguish between satisfactory and insufficient partial responses. Level 3.