Surgical accuracy of patient-specific instruments in high tibial osteotomy and distal femoral osteotomy: an alignment study of the lower-limbs, pelvis and lumbar spine

Introduction: The high tibial osteotomy (HTO) and distal femoral osteotomy (DFO) are common surgical procedures for the treatment of unicompartmental knee osteoarthritis in combination with varus/valgus knee malalignment. Patient-specific instruments (PSI) are novel 3D-printed surgical guides that aim to improve the coronal and sagittal plane accuracy of these procedures, which are important factors for patient outcomes. Investigation of PSI accuracy in HTO/DFO is ongoing, as different PSI designs could determine the accuracy of the corrections. Furthermore, pelvic/spinal alignment changes following HTO/DFO has received limited attention, despite studies demonstrating correlations between change in lower-limb anatomy and alignment of the pelvis/spine. Methods: X-ray and EOS radiographic images were analysed in a retrospective single-centre clinical study to compare the planned HTO/DFO surgical corrections to post-operative measurements of the hip-knee-ankle angle (HKA) and tibial slope (TS). Secondary observational measurements included the functional leg length (FLL), anatomical tibia and femur length, mechanical medial proximal tibial angle, mechanical lateral distal femoral angle, knee joint line convergence angle and knee flexion angle. Pelvic and spinal parameters included pelvic obliquity (PO), pelvic tilt, sacral slope, lumbar scoliosis and lumbar lordosis. Intra/inter-observer and methodological (inter-image) reliability were investigated to identify potential sources of measurement error – analysed using coefficient of variation (CV), with an acceptable variability threshold of 30%. Results: HKA error – the difference between the planned HKA and post-op measured HKA – was -2.05° ± 3.03 for HTO and 0.50° ± 2.90 for DFO. TS error was -0.97° ± 1.46. FLL change was 4.4 mm ± 4.7, and PO change was 1.33° ± 1.26 on the operative sides. HKA measurements showed acceptable variability (CV=5.17–26.76%). TS measurements showed acceptable variability for intra-observer reliability (CV=25.64%), but unacceptable variability for inter-observer and inter-image reliability (CV=179.18% and 45.88%, respectively). All inter-image measurements except for HKA showed unacceptable variability (CV=42.55–118.26%). Conclusion: This study indicated that performing HTO/DFO using PSI can result in accurate corrections of varus/valgus knee malalignment. However, intra/inter-observer and inter-image reliability data of the two-dimensional methods utilised in this study showed measurement variability that limits the confidence in our findings. Three-dimensional surgical planning and post-operative evaluation of HTO/DFO using weight-bearing computed tomography or sterEOSTM software should be investigated, as these are likely to address the limitations of this study and provide more accurate results on surgical accuracy and post-operative alignment changes.