Dedicated Ankle AP and Long-Leg Radiographs Are Not Interchangeable for Assessing Coronal Ankle Alignment: A...

Long-leg radiographs (LLRs) are widely used to assess global lower limb alignment, and ankle alignment is frequently measured directly from the ankle region of these images in studies of knee and lower extremity realignment procedures. However, most ankle deformity assessment and treatment algorithms are based on dedicated ankle anteroposterior (AP) radiographs, which are centered on the tibiotalar joint with the beam projected parallel to the joint surface based on a protocol that differs from that used with LLR acquisition. Whether LLR-derived ankle measurements reflect the same coronal joint orientation as those obtained from dedicated ankle AP radiographs remains unclear. This study compared coronal ankle alignment measurements from dedicated ankle AP radiographs and LLRs using regional weightbearing computed tomography (WBCT) as a reference.

This retrospective study included 64 ankles in 62 patients who underwent LLRs, dedicated ankle AP radiographs, and ankle WBCT within 6 months. Coronal ankle alignment parameters, including medial distal tibial angle (MDTA) and talar tilt (TT), were measured on both LLRs and ankle AP radiographs using standardized techniques. Corresponding WBCT measurements were obtained at 25%, 50%, and 75% of the sagittal plafond dimension to represent anterior, middle, and posterior joint regions. Bland-Altman analysis was performed to compare each radiographic modality with regional WBCT measurements and to assess agreement between modalities. In addition, MDTA and TT from each radiographic modality were categorized by deformity direction to determine the proportion of ankles showing varus-valgus discordance between ankle AP radiographs and LLRs.

LLR-derived MDTA was significantly greater than ankle AP-derived MDTA (94.6° vs 89.7°; mean difference, +4.86°; 95% CI, 3.81°-5.92°; P < .001). LLR-derived TT was also greater than ankle AP-derived TT (3.7° vs 1.4°; mean difference, +2.31°; 95% CI, 1.34°-3.28°; P < .001). Ankle AP MDTA most closely approximated the middle WBCT level (mean bias, +0.57°), whereas LLR MDTA most closely approximated the posterior WBCT level (mean bias, +2.46°). For TT, ankle AP measurements showed low mean bias relative to both the middle (+0.65°) and posterior (+0.12°) WBCT levels, whereas LLR measurements showed the smallest bias relative to the posterior level (+2.43°). Varus-valgus orientation differed between modalities in 21.9% of ankles for MDTA and 14.1% for TT.

Dedicated ankle AP radiographs and LLRs are not interchangeable for coronal ankle alignment assessment. LLR-derived measurements appeared to preferentially reflect posterior plafond orientation, whereas dedicated ankle AP radiographs more closely approximated middle plafond region, particularly for MDTA. Reliance on LLRs alone may mischaracterize ankle deformity when joint-level assessment is required.

Level III, retrospective diagnostic.