Long leg radiographs underestimate the degree of constitutional varus limb alignment and joint line obliquity in comparison with computed tomography: a radiographic study

Abstract

Purpose The purpose of this study was to understand if diferences exist between computed tomography (CT) and long leg radiographs (LLR) when deining coronal plane alignment of the lower limb in total knee arthroplasty (TKA). It aimed to identify any such diferences between the two imaging modalities by quantifying constitutional limb alignment (arithmetic hip-knee-ankle angle (aHKA), joint line obliquity (JLO) and Coronal Plane Alignment of the Knee (CPAK) type within the same population. Methods A retrospective radiographic study compared pre-operative LLR and CT measurements in patients undergoing robotic-assisted TKA. The aHKA, JLO and CPAK types were calculated after measuring the medial proximal tibial angle (MPTA) and lateral distal femoral angle (LDFA). The primary outcomes were the mean diferences in aHKA (MPTA− LDFA), JLO (MPTA + LDFA) and proportions of CPAK types between LLR and CT groups. The secondary outcomes were the diferences in CT-derived MPTA values based on four diferent tibial sagittal landmarks. Results After exclusions, 465 imaging sets were analysed in 394 patients. There was a statistically signiicant mean diference between LLR and CT, respectively, for both MPTA (87.5° vs. 86.2°; p < 0.01) and LDFA (88.7° vs. 87.3°; p < 0.01). There were also statistically signiicant diferences for aHKA (− 0.2° vs. − 1.1°) and JLO (175.1° vs. 173.4°) for LLR and CT, respectively (both p < 0.01). CT increased the proportion of patients with CPAK Type I (constitutional varus aHKA, apex distal JLO) and CPAK Type II (neutral aHKA, apex distal JLO), and decreased numbers of CPAK Types III-VI. There were signiicant mean diferences in the MPTA using varying sagittal landmarks. Conclusion Alignment determined by LLRs underestimates the magnitude of both constitutional varus alignment and joint line obliquity compared to CT, diferences that notably increase the proportions of patients included in CPAK Types I and II. These distinctions are primarily due to underestimation of proximal tibial varus when measured on LLRs compared to CT, which more speciically deines articular weight-bearing points. Level of evidence III.

Large multiplanar changes to native alignment have no apparent impact on clinical outcomes following total knee arthroplasty

“…As they are descriptive integrators of alignment variables, we used the Functional Knee Phenotypes (FKP) and CPAK systems to classify our cohort, both of which have been validated by studies using CT imaging [40–42]. Our preoperative alignments are strikingly similar to these (Table 5), thereby strengthening the hypothesis that plain radiography may underestimate the degree of deformity and constitutional joint line obliquity, and that the variation of phenotypes that exists is much broader than previously considered [35, 41].…”

Imageless robotic total knee arthroplasty determines similar coronal plane alignment of the knee (CPAK) parameters to long leg radiographs

“…There are several available methods for identifying native coronal plane morphology, including generic cartilage wear assumptions, use of long leg radiographs, or methods based on advanced imaging [ 12 – 15 ]. Some of the methods based on CT or MRI use proprietary algorithms that may not be accessible to all surgeons [ 14 – 17 ].…”

“…Our standard deviation was 3.4° with literature-based cartilage assumptions, which decreased to 3.1° with optimized cartilage assumptions. Tarassoli et al reported, in a subsequent study, that long leg radiographs underestimated proximal tibial varus in comparison to CT-based assessments by 1.3° on average [ 17 ]. The authors described that the MPTA, as measured from CT images, is sensitive to the sagittal position of the tibial landmarks [ 17 ].…”

“…Tarassoli et al reported, in a subsequent study, that long leg radiographs underestimated proximal tibial varus in comparison to CT-based assessments by 1.3° on average [ 17 ]. The authors described that the MPTA, as measured from CT images, is sensitive to the sagittal position of the tibial landmarks [ 17 ]. Our data are consistent with this observation, which is reflected by increased standard deviations for the MPTA compared to the LDFA for both the literature-based cartilage-assumptions (2.6° vs. 1.7°) and the optimized cartilage-assumptions (2.5° vs. 1.6°).…”

CPAK classification detect the real knee joint apex position in less than half of the knees

“…All LSRs were taken with the patient barefooted, feet positioned to be shoulder width apart, and the knee in full extension. The patella was facing forward, and the hip and ankle joints were fully visible, which allowed the joint line measurement points to be clearly identified [27]. The x-ray tube was at 2 m perpendicular to the patient and was aimed at the centre of the knees.…”