RESUMEN
BACKGROUND: Gadolinium enhanced MRI (ge-MRI) is considered as gold standard for perfusion evaluation in case of scaphoid nonunion (SNU). However, its clinical value and specificity is still not clearly evaluated. This study compares preoperative ge-MRI-based perfusion assessment and intraoperative proximal pole (PP) perfusion after scaphoid reconstruction by vascularized bone grafts. In addition, the postoperative osseous consolidation (OC) was correlated to intraoperative perfusion findings. METHODS: Between 08/2010 and 01/2020, 60 of 271 patients with scaphoid nonunion received a vascularized radius bone graft for reconstruction. Medical reports were checked for intra-op perfusion findings. Consolidation rate was assessed at mean follow-up of 3 months by CT evaluation. In 50 cases (83.2%), complete medical and radiological history could be obtained. Preoperative ge-MRI was reevaluated by a blinded radiologist for advanced analysis of sensitivity and specificity. RESULTS: Preoperative ge-MRI (initial finding, IF) showed 23 avascular, 20 malperfused, and seven vital PP. Blinded radiological follow-up (second finding, SF) revealed 14 avascular, 28 malperfused, and 8 vital PP, with a concordance of 65.3% (n = 35). After correlation with the intra-op findings, a specificity of preoperative ge-MRI of 76.5% (IF) and 88.2 (SF), respectively, was revealed for exclusion of avitality. For detection of malperfusion, there was a sensitivity of 92.7% (IF) and 85.4% (SF), respectively. Complete OC was seen 12 weeks postoperatively in 37 (73.5%), partial OC in 9 (18.3%), and nonunion in 4 cases (8.2%) on CT-scans. Of the 41 malperfused/avascular PP, 31 (75.6%) progressed to complete and 6 (14.6%) to partial (at least 2 adjacent CT-layers of 2 mm) OC, with 4 nonunions. CONCLUSION: The sensitivity and specificity of ge-MRI for detection/ exclusion of malperfusion/avitality of the PP was lower than expected. Therewith, the intraoperative assessment of PP perfusion regains a high value in decision-making for the appropriate graft. We recommend preservation of the dorsal radial vascular plexus initially until the vascularity of the proximal pole has been estimated. Patient education for all contingencies and retraction options should be obtained.