Plain Ankle Radiographs Underestimate Valgus Tibiotalar Tilt in Progressive Collapsing Foot Deformity: A Comparative Analysis with WBCT

Abstract

Category: Ankle Arthritis Introduction/Purpose: Detection and accurate assessment of valgus tibiotalar tilt are important for staging and surgical planning of progressive collapsing foot deformity (PCFD). Plain ankle radiographs (XRs) have been used to evaluate this condition, but recent literature suggests that XRs may underestimate ankle deformities when compared to weightbearing computed tomography (WBCT), which is becoming increasingly utilized in PCFD evaluation. However, the discrepancies between these methods have not been explored. This study aims to: 1) determine the difference in tibiotalar tilt measurements between XRs and WBCTs; 2) compare the prevalence of significant valgus tibiotalar tilt (>4 degrees) between XRs and WBCTs in PCFD patients; and 3) determine the radiographic characteristics of patients who exhibit a discrepancy between XR and WBCT in the presence of significant valgus tibiotalar tilt. Methods: One hundred thirty-eight PCFD patients (Mean age: 55.1 years) having both WBCT and ankle XRs were reviewed. On ankle XRs, tibiotalar tilt (TTXR) and medial distal tibial angle (MDTAXR) were measured. On WBCT, the TTWBCT and MDTAWBCT were assessed at three locations (25% [anterior], 50% [middle], and 75% [posterior] of the sagittal dimension of the distal tibial plafond) in coronal views (Figure 1). The degree of tibiotalar tilt and prevalence of significant valgus tibiotalar tilt were compared between TTXR and the TTWBCT. Bland-Altman plots compared the mean TT and MDTA from XRs to those of the three coronal WBCT images to determine which region in WBCT is most similar to the XRs in each measurement. The radiographic characteristics of patients with and without a discrepancy between XRs and WBCTs for the presence of significant valgus tibiotalar tilt were examined via subgroup analysis. Results: Mean TTXR and TTWBCT were 1.5 (standard deviation [SD] 3) and 4.1(SD 3) degrees, respectively, indicating that XRs underestimate tibiotalar tilt by 63.4%. The prevalence of significant valgus tibiotalar tilt was 10.1% on XRs and 37.7% on WBCTs. Half the patients (4/8) who had bone-on-bone contact with tibiotalar tilt on WBCT didn’t exhibit such findings on XR (Figure 1). Bland-Altman plots showed that TTXR and MDTAXR best correlate with middle TTWBCT and MDTAWBCT (Figure 2). Subgroup analysis revealed that patients who exhibited significant valgus tibiotalar tilt on WBCTs, but not XRs, had smaller anterior TTWBCT and MDTAWBCT than those without such discrepancy (p <.05, Table 1), indicating that patients with a greater varus deformity on the anterior distal tibia are more likely to exhibit the discrepancy. Conclusion: This study demonstrates that ankle XRs underestimate valgus tibiotalar tilt compared to WBCTs in PCFD patients. The ankle XR parameters had the strongest correlation with those of the midpoint of the tibiotalar joint in WBCT and understated anterior side tilt and distal tibial deformity, likely due to beam projection angle and superimposition. Based on the findings of this study, WBCTs may be useful for detecting actual deformities in the ankle joint in PCFD patients; nevertheless, surgeons must be cognizant that they may underestimate the deformity in the anterior ankle when using conventional radiographs to assess ankle deformities in PCFD.