Is Dynamic Gadolinium Enhancement Needed in MR Imaging for the Preoperative Assessment of Scaphoidal Viability in Patients with Scaphoid Nonunion?

Abstract

To compare the accuracy of dynamic gadolinium-enhanced magnetic resonance (MR) imaging with that of standard MR imaging for assessing the viability of the proximal pole of the scaphoid in patients with nonunion. The study protocol was submitted to the institutional review board, and the need to obtain additional approval was waived for this retrospective study. Twenty-eight patients (mean age ± standard deviation, 24.3 years ± 6.4) with nonunion of a scaphoid fracture underwent dynamic gadolinium-enhanced MR imaging of the wrist 28 days ± 19 before surgery. Dynamic gadolinium-enhanced MR imaging consisted of acquisition of 40 consecutive coronal T1-weighted images over 1 minute. Two readers retrospectively evaluated MR images obtained with a standard protocol and rated the viability of the proximal scaphoid pole. The steepest upslope of gadolinium uptake was calculated in a region of interest placed in the proximal scaphoid pole by a third reader. Receiver operating characteristic curves were calculated, and the areas under the receiver operating characteristic curve (A(z) values) were compared. Diagnostic performance in determining scaphoid viability was calculated for readers 1 and 2. Histologic findings in 11 patients and surgical findings in all patients served as the standard of reference. The sensitivity, specificity, and accuracy of standard MR imaging in the detection of scaphoid necrosis were 54%, 93%, and 75%, respectively, for reader 1 and 62%, 93%, and 78% for reader 2. Interreader reliability was excellent (κ = 0.92). The A(z) was 0.82 for reader 1 and 0.87 for reader 2. The diagnostic performance of dynamic gadolinium-enhanced MR imaging, determined with the steepest upslope value, was inferior to that of standard MR imaging, with an A(z) of 0.57. Findings at histologic examination (viable bone, necrotic bone, callus formation) did not correlate with those at dynamic gadolinium-enhanced MR imaging. Because the diagnostic performance of dynamic gadolinium-enhanced MR imaging in the evaluation of scaphoid viability was inferior to that of a standard MR imaging protocol, dynamic acquisition may not be needed in patients with nonunion of scaphoid fractures.