Erschienen in:
04.03.2020 | Scientific Article
Predicting osteomyelitis in patients whose initial MRI demonstrated bone marrow edema without corresponding T1 signal marrow replacement
verfasst von:
Alessandra J. Sax, Ethan J. Halpern, Adam C. Zoga, Johannes B. Roedl, Jeffrey A. Belair, William B. Morrison
Erschienen in:
Skeletal Radiology
|
Ausgabe 8/2020
Einloggen, um Zugang zu erhalten
Abstract
Purpose
We endeavored to determine which characteristics of diabetic ulcers portend the strongest risk for osteomyelitis in patients whose initial T1-weighted imaging was normal. By determining which features have a greater risk for osteomyelitis, clinicians can treat patients more aggressively to reduce the sequela of inadequately treated osteomyelitis.
Materials and methods
We performed a retrospective analysis of MR imaging from 60 pedal ulcers with suspected osteomyelitis. Ulcer dimensions and depth were measured. Ratios of marrow ROI/joint fluid ROI on T2/STIR sequences were obtained. Progression to osteomyelitis on subsequent MRI was characterized by loss of normal marrow signal on T1-weighted images. Statistical analysis was performed with a two-sample t test and Cox proportional hazard model. A p value < 0.05 was used as the threshold for statistical significance.
Results
Sixty MR exams were identified. Thirty-four progressed to osteomyelitis. Marrow ROI/joint fluid ratios averaged 65% in the osteomyelitis group, and 45% in the non-osteomyelitis group, p < 0.001. ROI ratios > 53% had a 6.5-fold increased risk of osteomyelitis, p < 0.001. Proximity to bone averaged 6 mm in the osteomyelitis group and 9 mm in the non-osteomyelitis group, p = 0.02. Ulcer size averaged 4 cm2 in the osteomyelitis group versus 2.4 cm2 in the non-osteomyelitis group, p = 0.07. Ulcers greater than 3 cm2 has a 2-fold increase in the risk of osteomyelitis, p = 0.04.
Conclusion
Increasing bone marrow ROI signal/joint fluid ratios on T2/STIR images were the strongest risk factors for developing osteomyelitis, while ulcer size and depth are weaker predictors.