Accuracy of enhanced and unenhanced MRI in diagnosing scaphoid proximal pole avascular necrosis and predicting surgical outcome

Determine the sensitivity, specificity and accuracy of unenhanced and enhanced MRI in diagnosing scaphoid proximal pole (PP) avascular necrosis (AVN) and correlate whether MRI can help guide the selection of a vascularized or nonvascularized bone graft. The study was approved by the IRB. Two MSK radiologists independently performed a retrospective review of unenhanced and enhanced MRIs from 18 patients (16 males, 2 females; median age, 17.5 years) with scaphoid nonunions and surgery performed within 65 days of the MRI. AVN was diagnosed on the unenhanced MRI when a diffusely decreased T1-W signal was present in the PP and on the enhanced MRI when PP enhancement was less than distal pole enhancement. Surgical absence of PP bleeding was diagnostic of PP AVN. Postoperative osseous union (OU) was assessed with computed tomography and/or radiographs. Sensitivity, specificity and accuracy for PP AVN were 71, 82 and 78 % for unenhanced and 43, 82 and 67 % for enhanced MRI. Patients with PP AVN on unenhanced MRI had 86 % (6/7) OU; 100 % (5/5) OU with vascularized bone grafts and 50 % (1/2) OU with nonvascularized grafts. Patients with PP AVN on enhanced MRI had 80 % (4/5) OU; 100 % (3/3) OU with vascularized bone grafts and 50 % (1/2) OU with nonvascularized grafts. Patients with viable PP on unenhanced and enhanced MRI had 91 % (10/11) and 92 % (12/13) OU, respectively, all but one with nonvascularized graft. When PP AVN is evident on MRI, OU is best achieved with vascularized grafts. If PP AVN is absent, OU is successful with nonvascularized grafts.