Accuracy of CT Derived Patient Specific Instrumentation for Total Ankle Arthroplasty: The Impact of the Severity of Preoperative Varus Ankle Deformity

CATEGORY: Ankle; Ankle Arthritis INTRODUCTION/PURPOSE: Significant preoperative varus tibiotalar deformity was once believed to be a contraindication to performing total ankle arthroplasty (TAA). While correction of ankle deformity has been shown to be possible, the ability to restore neutral alignment in more severe ankle arthritis is still a concern. Preoperative CT derived navigation is gaining popularity for its ability to accurately restore neutral alignment in TAA, in addition to decreasing operative and fluoroscopic times. The primary goal of this study was to evaluate whether accuracy and reproducibility of final implant positioning using patient specific guides were influenced as varus deformity became more severe in patients undergoing TAA. METHODS: A retrospective review of 32 patients with varus end-stage ankle arthritis who underwent TAA utilizing total ankle implants with CT derived patient-specific guides was conducted. Preoperative computed tomography (CT) scans were obtained to assess coronal plane deformity, mechanical and anatomic alignment, and build patient-specific guides that referenced bony anatomy. Preoperative weightbearing radiographs were used to measure the tibiotalar and talar tilt angles to assess coronal plane deformity severity and congruency, respectively. Patients were subcategorized into varying degrees of deformity based on their tibiotalar angles (<0-5° neutral, 6-10° mild, 11-15° moderate, and >15° severe). The first postoperative weightbearing radiograph following surgery was used to measure coronal plane alignment of the tibial implant relative to the target axis determined by the preoperative CT template. RESULTS: The average preoperative varus deformity, as determined by the tibiotalar angle, was 6.06° (range: 0.66- 16.3°) for the entire cohort. Average preoperative varus deformity was 2.63° for the 'neutral' group, 7.64° for the mild group, 11.5° for the moderate group and 16.3° for the severe group. Postoperative weightbearing alignments for 96.9% (30/31) of patients demonstrated neutral implant alignment (< 5°). Relative to the CT derived target axis, the average tibial implant coronal deviation for all patients was 1.54° (range: 0.17-5.7°). The average postoperative coronal alignment relative to the target axis was 1.61° for the neutral group, 1.78° for the mild group, 0.94° for the moderate group, and 1.41° for the severe group. There was no statistically significant difference in final implant alignment relative to the target axis between deformity subgroups (p= 0.256). CONCLUSION: The results of our study further support that neutral postoperative TAA alignment can repeatably be obtained using preoperative CT derived patient specific instrumentation (PSI). Furthermore, the accuracy and reproducibility of achieving neutral postoperative radiographic alignment with PSI does not appear to be impacted by worsening varus ankle deformity. All but one patient (96.9%) achieved neutral postoperative alignment relative to the predicted target axis, with the only outlier being from the mild deformity subgroup. Future studies are warranted assessing the effect that even greater coronal plane deformity (>20 °) has on the accuracy of TAA implantation using PSI.