Are three-dimensional patient-specific cutting guides for open wedge high tibial osteotomy accurate? An in vitro study

Open wedge high tibial osteotomy (OWHTO) has been described as an efficient conservative surgical treatment preserving the bone stock for patients with moderate medial gonarthrosis and lower leg malalignment [1, 2]. The objective of the OWHTO is to correct lower leg malalignment in both the frontal and sagittal tibial planes to limit the overload of the medial compartment. Accurate correction is essential to its success as under correction leads to persistent pain and over correction to functional limitations [3, 4].

Different methods and instrumentations have been developed to help the surgeon to achieve the planned correction [3, 5‐7].

Conventional methods use standard instrumentation to open the osteotomy [5, 6]. Frontal correction can then be managed intraoperatively by measuring the opening angle or opening gap and comparing it to the preoperative plan [8‐10], or by using a radiopaque cable under fluoroscopy to control lower limb alignment [9, 11‐13]. The control of the correction for both the frontal and the sagittal planes at the same time with standard instrumentations remains challenging [14‐16]. Computer-assisted surgery (CAS) for OWHTO was validated experimentally in vitro by Hankemeier et al. [7] and subsequently used in clinical studies [3, 17‐21]. CAS allows real-time control of the correction. All these studies reported better accuracy and reliability for CAS than for conventional or cable methods, but with increased surgical time and a control of the global alignment of the limb but not of tibia only and not the posterior tibial slope (PTS).

While frontal plane correction is most often described for osteotomy management, PTS management in the sagittal plane is essential to preserve biomechanics [22‐24]. Song et al. [19] reported that PTS is unchanged if the anterior opening is equal to 67% of the medial opening. However, this evaluation is complex to perform during surgery, and patient-specific instruments may help the surgeon to manage both the sagittal and the frontal plane correction during surgery [25, 26].

Using an experimental setup, three-dimensional patient-specific cutting guides were reported to be more accurate than free-hand technique to perform an osteotomy cut and drill in a synthetic bone [27]. The use of patient-specific cutting guides for OWHTO was also reported recently in three studies realized on small series of patients [25, 26, 28]. All three clinical studies reported good accuracy and reliability of the procedure. Planning procedures for OWHTO were carried out either on 2D long-leg radiographs [28] or from a 3D model [25, 26]. Pérez-Mañanes et al. used CT images to obtain the 3D tibial surface and to position two K-wires to lead the cut, and the correction was planned using two additional wedges.

To avoid the disappointing clinical results observed with patient-specific instrumentation for total knee arthroplasty, we wanted to evaluate in vitro this new technique. After developing a specific patient-specific guide to control both the cut and the correction adapted for a specific OWHTO plate, it was our aim to evaluate the accuracy of the system in an in vitro CT scan-controlled study. Our hypothesis was that the patient-specific cutting guide for OWHTO can provide an accurate correction in both the frontal and sagittal planes.

The mean pre-OWHTO mMPTA was 87.7° (SD 2.2°), and the mean pre-OWHTO PTS was 5.1° (SD 2.6°). A mean 7.3° (SD 1.3°) correction in the frontal plane and a 3.3° (SD 2.7°) correction in the sagittal plane were thus planned. For all specimens, the positioning of the patient-specific cutting guides on the bone was possible in accordance with the planning. All patient-specific cutting guides fitted the tibial surface, so all OWHTO were performed as planned.

Differences between the post-OWHTO and planned-OWHTO configurations were calculated. The mean difference between the post-OWHTO and planned-OWHTO configurations was 0.2° (from − 0.3° to 0.5°, SD 0.3°) in the frontal plane, and − 0.1° (from − 0.7° to 0.8°, SD 0.5°) in the sagittal plane (Table 2).

According to Pearson’s correlation tests, a statistically significant correlation was found between the planned-OWHTO and post-OWHTO configurations, for the mMPTA measurements with a correlation coefficient of 0.99 (95% CI, 94–99%, p < 0.0001), and for the PTS measurements with a correlation coefficient of 0.97 (95% CI, 86–99%, p < 0.0001) (Fig. 4).

Bias coefficient C_b was 0.99 in both the frontal and sagittal planes (Table 3).

OWHTO success remains on accurate correction in order to avoid persistent pain or functional limitations [3, 4]. This in vitro study aimed to investigate the accuracy in both the frontal and sagittal planes provided by patient-specific cutting guides with respect to 3D planning of OWHTO. The hypothesis that patient-specific cutting guides can provide the planned correction in both the frontal and sagittal planes was confirmed in this CT scan-controlled in vitro study.

In this patient-specific procedure for OWHTO, bone deformation angles were measured with respect to the anatomical reference planes determined using the procedure of Lee et al. [31] in order to mimic clinical practice. Clinical studies report several methods of finding anatomical reference points and measuring deformation angles [25]. Seeking to achieve maximum reliability and to measure only the correction provided by the patient-specific cutting guide, all measurements were performed relative to the same axis and in the same planes for the pre-OWHTO, planned-OWHTO, and post-OWHTO configurations. Unlike Munier et al., who performed measurements separately on the preoperative and postoperative 3D configurations, and then compared how far the two configurations differed from the planned correction, we performed our measurements relative to the preoperative mechanical axis. This was made possible by the registration process we used to superimpose the post-OWHTO configuration on the planned-OWHTO configuration. As our objective was to assess the amount of correction, it was vital to keep the reference definition the same for both planned-OWHTO and post-OWHTO configurations.

Intraoperative methods to control correction are varied, have limited accuracy, and mainly focus on the frontal plane correction. In their review of the literature, Van Den Bempt et al. reported ranges of accuracy in the frontal plane from several clinical studies [4]. For conventional methods, the mean amplitude of the range of accuracy was 5.6° (from 4° to 8°) [33‐37], whereas using the CAS method, it was 5.5° (from 4° to 7°) [10, 17, 38‐41]. Irrespective of the range of accuracy chosen by the authors, the literature reports a mean 32% of outlier patients when authors used conventional methods [9, 10, 17, 21, 34‐37, 40, 42‐45] and 22% when they used CAS [10, 17, 21, 38‐42, 44, 46]. Some in vitro studies find CAS to be accurate for OWHTO. On a single synthetic bone and with a statistical model, Keppler et al. [47] evaluated a mean error of 0.7° in the frontal plane between the postoperative results and the target. Wang et al. [48] performed OWHTO with several amounts of correction in a synthetic bone using CAS and reported 0.4° accuracy in the frontal plane. Both authors validated their experimental work in a preliminary clinical study, on five and four patients. They both found a mean error of 1° in the frontal plane. On a single cadaveric specimen, Lützner et al. [49] evaluated the accuracy of CAS to measure lower limb alignment. They found a mean error of 0.6° in the frontal plane. Hankemeier et al. [7] performed OWHTO on 20 legs randomly assigned to CAS or a conventional method. They reported better accuracy and less variability than with a cable method. Among clinical studies using patient-specific cutting guides, Menetrey et al. [20] evaluated the frontal plane correction, based on 2D measurements for planning. Their patient-specific cutting guide only incorporates the cutting plane position, the correction being guided by two additional wedges. They found 0.5° accuracy (from 0° to 1.2°) in the frontal plane. Patient-specific cutting guides containing both the cutting plane planning and the correction were used in two other studies [19, 21]. In one, Munier et al. performed a postoperative 3D evaluation of their patients. Overall, they found similar accuracy for both 2D and 3D measurements by reproducing the measurement protocol on the postoperative model [26] around 0.0° (from − 1.7° to 1.8°, SD 1.1°) in the frontal plane. In the other study, Victor and Premanathan [25] reported a mean difference of 0.1° (from − 1° to 1°, SD − 0.1°) in the frontal plane.

Controlling the sagittal plane correction is essential to preserve knee biomechanics [23‐25], but conventional methods have limited accuracy, remaining on gap measurements [15, 19]. Using CAS on synthetic bones, a mean error of 0.9° [47] and a 0.5° accuracy [48] were reported in the sagittal plane. Using patient-specific cutting guides, an accuracy around 0.3° (from − 2° to 3.2°, SD 1.4°) [26] and a mean difference of − 0.1° (from − 3° to 2°, SD 1.2°) [25] in the sagittal plane were reported.

In this study, the patient-specific OWHTO procedure includes 3D preoperative planning of the surgery and the design of a patient-specific cutting guide which takes into account the tibial anatomy, the position of the cutting plane, the amount of correction planned in all planes, and the plate location. The mean differences between the planned-OWHTO and post-OWHTO models are 0.2° (from − 0.3° to 0.5°, SD 0.3°) in the frontal plane and − 0.1° (from − 0.7° to 0.8°, SD 0.5°) in the sagittal plane. OWHTO outcomes are compared with the planning by superimposing the post-HTO 3D reconstruction on the planning 3D model. This enabled to assess the accuracy of patient-specific cutting guides. Findings of the present study suggest that PTS is managed with accuracy, which is important for the management of cruciate ligament balance [23, 25].

One limitation of our study is the small number of specimens used. Moreover, not all specimens had a sufficient varus deformation in their proximal tibia to be considered as candidates for OWHTO. However, different degrees of correction consistent with clinical practice were planned in both the frontal and sagittal planes. We also departed from clinical practice in performing all the measurements on 3D models. This was possible thanks to the cadaveric nature of the specimens. This enabled us to assess precision and accuracy by comparing post-OWHTO and planned-OWHTO 3D models, thereby avoiding measurement errors related to 2D radiography. The preoperative CT scan required for this procedure could be another limitation for direct clinical application, not being necessary in conventional or CAS procedures. However, Menetrey et al. [20] reported that the patient-specific cutting guide reduced the use of intraoperative fluoroscopy from 55 images on average (range, 41–73) in conventional methods to 8 (range, 6–14), as well as requiring less surgical time.

OWHTO is demanding surgery, and accuracy of the correction in both frontal and sagittal planes is essential to its success. This study shows that combining 3D planning with patient-specific cutting guides for OWHTO is a reliable and accurate method of achieving multiplanar correction in both the frontal and the sagittal planes. Further randomized clinical studies should be carried out to validate these experimental results and evaluate the risk-benefit ratio of the preoperative CT scan, and the reduction in surgical time.