Navigation in MOWHTO provided reliable information about bony correction, however MA on standing radiograph tended to be overcorrected, and the difference of JLCA between supine and standing radiographs was the most important preoperative factor that could predict the coronal correction discrepancy after MOWHTO.
High tibial osteotomy (HTO) has been widely used for the treatment of medial compartment arthritis with varus deformity of the knee. It is well known that the achievement of ideal limb alignment after high tibial osteotomy is closely related to the favorable clinical outcomes, and common target is slight overcorrection to 3° of valgus of mechanical axis (MA). While undercorrection is related to early failure leading to revision osteotomy or arthroplasty and poor clinical outcomes, overcorrection causes cosmetic problem, overloading of lateral compartment and persistent pain, which can also result in inferior clinical outcomes. To achieve accurate and reproducible MA after HTO, various techniques including preoperative geometric methods like Miniaci method, intraoperative cable method, grid board method and navigation system are employed. Navigation system in HTO is known to reduce outliers from predicted postoperative MA, however overcorrection occasionally occurs after medial opening wedge (MOW) HTO even with utilization of intraoperative navigation.
The incidence and causes of coronal correction errors after navigation-assisted MOWHTO are still unclear. Thus, we designed the study to figure out the accuracy of navigation in MOWHTO, and to identify factors affecting the coronal correction error after navigation-assisted MOWHTO.
One-hundred and fourteen knees in 98 patients who underwent navigation-assisted MOWHTO from May 2010 to July 2016 were enrolled and retrospectively reviewed. MA on standing whole-leg radiograph and medial proximal tibial angle (MPTA) were measured preoperatively and at postoperative 6 months, and the differences (dMA and dMPTA) were calculated. Joint line convergence angle (JLCA) on supine and standing radiographs were measured preoperatively, and the difference (dJLCA) was calculated. JLCA on varus and valgus stress radiographs were also measured preoperatively. The difference of JLCA between standing and varus/valgus stress radiographs were defined as varus/valgus openings. To assess the accuracy of navigation, dMA and dMPTA were compared with the coronal correction by navigation (dNMA) using intraclass correlation coefficients (ICCs). Univariable and multivariable regression analyses were used to identify factors affecting coronal correction discrepancy (dMA–dNMA). The knees with coronal correction discrepancy exceeding ±3° were classified as navigation outliers.
Mean dMA was 11.6° (range, 5.5° to 22.0°) and mean dMPTA was 9.9° (range, 4.4° to16.3°). Mean dNMA was 9.7° (range, 4.0° to 15.0°). Reliability of navigation was good in terms of bony correction (ICC between dNMA and dMPTA, 0.844), fair in terms of MA correction (ICC between dNMA and dMA, 0.706). Mean coronal correction discrepancy was 2.0° ± 2.4° and 32 knees (28%) showed discrepancy exceeding ±3° (navigation outliers). All the 32 knees were overcorrected. The navigation outlier group had more preoperative varus deformity, greater JLCA on both standing and varus stress radiographs, greater dJLCA, lesser varus opening and greater valgus opening. Univariable analysis revealed coronal correction discrepancy correlated with the preoperative MA, JLCA on standing radiograph, dJLCA, varus opening, and valgus opening. In multivariable analysis, the difference of JLCA between supine and standing radiographs was shown to be a predictive factor for coronal correction discrepancy (unstandardized coefficients, 1.026; R^2, 0.470).
Navigation in MOWHTO provided reliable information about bony correction, however MA on standing radiograph tended to be overcorrected. The difference of JLCA between supine and standing radiographs was the most important preoperative factor that could predict the coronal correction discrepancy after MOWHTO.
Level of Evidence: Case series, Level IV