Plain Radiographs Underestimate Tibiotalar Tilt in Progressive Collapsing Foot Deformity

Abstract

Background:

Accurate evaluation of tibiotalar alignment is essential for staging and surgical planning in progressive collapsing foot deformity (PCFD), yet plain radiographs are limited by their 2-dimensional nature and variability in beam angulation. This study pursued 3 objectives: to quantify discrepancies in tibiotalar tilt (TT) between plain radiographs (XRs) and weightbearing CT (WBCT), to compare the prevalence of significant valgus tilt between modalities, and to identify features associated with underrecognized deformity on XRs.

Methods:

One hundred thirty-eight PCFD patients (mean age: 55.1 years) who underwent both WBCT and ankle XRs were reviewed. On weightbearing ankle XRs, tibiotalar tilt (TT_XR), medial distal tibial angle (MDTA_XR), and talus-ground surface angle (Tal-GS_XR) were measured. On WBCT, corresponding parameters (TT_WBCT, MDTA_WBCT, and Tal-GS_WBCT) were assessed at 25% (anterior), 50% (middle), and 75% (posterior) of the joint. The degree of tibiotalar tilt and prevalence of significant valgus tibiotalar tilt (≥4°) were compared. Bland-Altman plots evaluated agreement between XRs and WBCTs. Patients were classified into Discrepancy (TT_XR < 4°, TT_WBCT ≥ 4°) and No-discrepancy groups, and group characteristics were compared. Interobserver and intraobserver reliability were assessed using intraclass correlation coefficients (ICCs).

Results:

Mean TT_XR was 1.5°, and mean TT_WBCT was 4.1°, indicating a 63% underestimation on radiographs. Significant valgus tilt (≥4°) was present in 10.1% of radiographs vs 37.7% of WBCT scans. Radiographs correlated best with the middle WBCT slice. Discrepancy group patients showed greater anterior TT_WBCT values. On multivariable logistic regression, anterior tibiotalar tilt on WBCT was the predictor of XR-WBCT discrepancy (odds ratio, 33.0; 95% CI, 5.6-114.7; P < .001), suggesting anterior deformities visualized on WBCT are not reliably captured on radiographs. Interobserver reliability ranged from 0.80 to 0.86 and intraobserver reliability from 0.88 to 0.89 (ICC).

Conclusion:

Plain radiographs underestimate valgus tibiotalar tilt compared with WBCT, mainly because of limited visualization of the anterior joint. Surgeons should recognize that reliance on XRs alone may lead to underrecognition of ankle valgus alignment in patients with PCFD. As WBCT is not universally available, optimizing XR projection angles may better capture anterior tibiotalar alignment and improve preoperative planning.

Level of Evidence:

Level III, retrospective diagnostic.