Advances in imaging for pre-surgical planning in hip resurfacing arthroplasty

Background

Accurate preoperative templating is essential for the success of hip resurfacing arthroplasty (HRA). While digital radiograph is currently considered the gold standard, stereoradiograph and CT converted 3D methods have shown promising results. However, there is no consensus in the literature regarding the preferred modality for HRA templating, and angular measurements are often overlooked. Thus, this study aimed to: (1) compare the performances of different modality in implant sizing and angle measurements, (2) evaluate the measurement reproducibility, (3) assess the impact of severe osteoarthritis on femoral head sizing, and (4) based on the analysis above, explore the optimal imaging and planning strategy for HRA.

Hypothesis

An optimal imaging modality exists for HRA planning regarding implant sizing and angular measurements.

Materials and methods

Preoperative imaging data from seventy-seven HRA surgeries were collected. Three raters performed templating using digital radiograph, stereoradiograph, and CT converted 3D models. Measurements for femoral head size, neck-shaft angle, and calcar-shaft angle were obtained. The femoral head sizing was compared to the intraoperative clinical decision. The reproducibility of measurements was assessed using the intraclass correlation coefficient (ICC). Correlations were examined between sizing disagreement and osteoarthritis grade (Tonnis Classification).

Results

Digital radiograph, stereoradiograph, and 3D techniques predicted one size off target in 27/77 (35%), 49/70 (70%), and 75/77 (97%) of cases, respectively, corresponding to 1.8 ± 1.6 (0 to 5.67), 0.9 ± 0.7 (0 to 2.67), and 0.4 ± 0.4 (0 to 1.67) sizes off target, indicating statistically significant differences among all three modalities, with p-values all below 0.01. There were no statistically significant differences among the different modalities for angular measurements. Measurements showed moderate to excellent reproducibility (ICC = 0.628–0.955). High-grade osteoarthritis did not impact image sizing in any modality (r = 0.08–0.22, all p > 0.05).

Discussion

CT converted 3D models were more accurate for implant sizing in HRA, but did not significantly outperform other modalities in angular measurements. Given the high costs and increased radiation exposure associated with CT, the study recommended using CT scans selectively, particularly for precise femoral head sizing, while alternative imaging methods can be effectively used for angular measurements.

Level of evidence

III; retrospective comparative diagnostic study.