Introduction

The three biomechanical consequences of anterior cruciate ligament (ACL) deficiency are increased anterior translation of the tibia [1], loss of the synchronization between the lateral femoral condyle and tibial plateau (pivot shift) [2], and the medialisation of the centre of rotation due to an internal rotation of the tibia [3]. The natural history of ACL deficiency can lead to femoro-tibial osteoarthritis (OA) due the biomechanical changes, cartilage and meniscal injuries caused by the initial injury or through repetitive episodes of instability. Conversely, tibio-femoral OA can also lead to ACL deficiency [4], typically progressing anterior to posterior in the medial compartment and destroying the ACL and damaging the lateral compartment [5].

The classic surgical options for OA secondary to ACL deficiency are either a high tibial osteotomy (HTO) combined with ACL reconstruction in early to moderate stages [6, 7] or a TKA in advanced stages. However, patients are often young, physically active and TKA may limit activity. Furthermore, it carries a high risk of revision within five years (5% under 55 years old [8]). Isolated UKA on the other hand for ACL deficient monocompartmental knee OA has been shown to have an intolerable rate of 21% failure at two years [9], with majority of failures occurring due to early tibial loosening [10].

Several studies have challenged the classic indication of not performing UKA in ACL deficient knees, published encouraging results with survival rates over 90% at 5 years follow-up [11,12,13]. Despite these good results, the Oxford team in their series of 52 patients at five years reported a 10% rate radiolucencies, one conversion to TKA and one PE dislocation [13] reflecting the challenging nature of this surgery. Combined surgery is technically difficult as adjustment of the posterior tibial slope (PTS), hip-knee-ankle (HKA) angle, height of the polyethylene (PE) and finally the tension of the ACL reconstruction influence gap balancing in both flexion and extension. Robotic assistance system allows for precise control of the aforementioned difficulties [14]; however, results when used for this specific indication to the best of our knowledge have not been published.

The purpose of this study was to evaluate the early clinical and radiological outcomes of combined single-stage robotic-assisted UKA and ACL reconstruction in patients with an ACL deficiency and concomitant symptomatic medial or lateral compartment knee OA.

Methods

Consecutive patients from a single-centre undergoing robotic-assisted UKA with combined ACL reconstruction between October 2016 and April 2020 were included and retrospectively reviewed. Inclusion criteria was isolated medial or lateral femorotibial OA with post-traumatic ACL deficiency confirmed on MRI imaging, a reducible deformity and BMI < 35. Exclusion criteria were a lower limb coronal plane deformity greater than 10° of varus or 15° of valgus associated procedure (osteotomy, controlateral UKA, patellofemoral arthroplasty) and a fixed anterior subluxation of the tibia (Fig. 1). ACL degenerative deficiency secondary to OA was considered as a contraindication to surgery as in that situation degenerative changes are, most of the time, present in the lateral compartment [5]. In that situation, a TKA was then performed. All cases were performed by the same surgeon who performs more than 50 UKA per annum.

Fig. 1
figure 1

Patient with a medical history of ACL rupture with a fixed anterior subluxation of the tibia that contra-indicate the surgery

Full size image

Assessed data

Pre-operatively, all patients completed an IKS score (functional and knee scores). Radiographic examination was performed before surgery, at two months, then once a year post-operatively (weight bearing antero-posterior and lateral knee radiographs, patellar axial view and full-length standing radiographs). A pre-operative MRI was routinely performed to assess the integrity of the ACL as well as cartilage wear of the two other compartments. Radiological evaluation was assessed by an independent orthopaedic surgeon and included the HKA angle and the PTS. The PTS was measured as the angle between the articular surface of the tibia and the posterior cortex of the tibia [15]. All measurements were performed with the software Centricity Universal Viewer Zero Footprint (version 6.0 SP7.0.2—GE Healthcare, Barrington, USA). Outlier rates were determined for mechanical axis alignment (HKA objective: 178° ± 2° for varus deformation, 182° ± 2° for valgus deformation), and PTS (5° ± 3°) [14, 16]. Radioluciencies were assessed and classified as physiological or pathological (progressive, poorly defined, > 2 mm thick, no matching radiodense line). Radiolucencies beside the vertical wall of the femur were not assessed, as this is not a site where the component is fixed and few cement is placed [17]. Clinical results were assessed post-operatively, at the last follow-up, by Tegner level activity scale [18], IKS score (divided into functional and knee scores) [19], forgotten joint score (FJS) [20] and by a satisfaction score (divided into very satisfied, satisfied, disappointed). Complications, all re-operations and revision were recorded.

Surgical techniques

The implant used was a cemented, cutting type unicompartmental prosthesis with a metal back fixed-bearing tibial component (Journey Uni, Smith and Nephew®). Surgery was performed in the supine position with a tourniquet. A standard skin incision for a mid-vastus approach was performed and the gracilis and semitendinosis tendons were harvested and prepared to form a four-strand graft with the pedicle left attached to the tibia. Next, arthroscopy was performed to assess all three compartments to confirm suitability for the UKA. The notch was cleared and the femoral and tibial tunnel prepared using an outside-in drilling technique. A standard femoral tunnel was created for all cases; however, for medial UKA the tibial tunnel was placed slightly more laterally than usual in order to avoid impingement of the graft on the prosthetic tibial plateau. Planning of bone cuts, implant positioning and balancing were performed using the BlueBelt Navio robotic surgical system (Smith and Nephew®) according to a previously described technique by Lustig et al. [21]. With the trial components in place, the ACL graft was pulled into the tunnels and checked for impingement. Next, the graft was fixed into the tibial tunnel with a screw, then final implant was cemented and finally graft fixation on the femoral side performed with a screw (Fig. 2). A significant advantage of robotic assistance is the control of the HKA angle to avoid any over-correction (Fig. 3a), control of the tibial slope (Fig. 3b) (to avoid excess strain on the ACL graft) and to control the graft tension whilst targeting gap balancing with fine adjustments with thickness of the PE (Fig. 4).

Fig. 2
figure 2

Pre-operative and 2 years follow-up X-Rays of a robotic medial UKA (Journey Uni, Smith and Nephew®) with ACL reconstruction in a patient who underwent 3 Lemaire interventions

Full size image
Fig. 3
figure 3

Contribution of Navio® during the planning. (a) Planning of residual HKA angle ← and gap balancing  , (b) Planning of PTS

Full size image
Fig. 4
figure 4

Contribution of Navio® during the testing with trials. The dotted line represents the planning, the continued area represents the actual testing: (a) Without tightening the ACL gaps are too loose between 0° and 90° of flexion, (b) Perfect testing corresponding to the planning

Full size image

Statistics

All statistical analysis was performed using XLstat (version 2015.1, Addinsoft, Paris, France). Descriptive data analyses such as means, standard deviations and ranges were performed. Comparisons were made using independent t tests for normally distributed variables and the Mann–Whitney U test for none normally distributed variables. Categorical variables were summarized by their percentages. Significance was set at alpha = 0.05; P values < 0.05 were considered statistically significant.

Results

Two lateral UKA and eight medial UKA were performed. Patient demographic details are described in Table 1. Mean follow-up was 45 ± 13 months [24–66], no patients were lost to follow-up.

Table 1 Demographic data
Full size table

Clinical results

IKS knee and function scores were significantly improved following surgery (Table 2). The mean post-operative Tegner score was 4.5 ± 1.4 [3–6]. Nine patients (90%) returned to the desired level of sport. Five patients (50%) were able to ski and four (40%) to practice running again. No anteroposterior instability was observed. Only one person (10%) was disappointed by the surgery due to a limitation of sport, with no pain in daily life.

Table 2 Comparison of pre- and post-operative clinical and radiographic data
Full size table

Complications

Two (20%) arthrolysis under arthroscopy for stiffness were required at two and three months after surgery. PTS correction was acceptable in both cases(4° and 5°), ROM improved to 120° and 130° and both patients were very satisfied with the end-result of the surgery.

No complication due to the robotic-assisted surgery occurred: Specifically, no pin site related problems or conversion no mechanical technique were required. The mean operative time was 128 min. ± 21 [96–150].

Radiological results (Table 2)

No patients demonstrated progress radiolucency’s or OA in the contralateral compartment at the last follow-up. There were no outliers concerning the radiological accuracy (Fig. 5).

Fig. 5
figure 5

Repartition of the postoperative radiographic results objectives mean median

Full size image

Discussion

The most important finding of this study was that satisfactory patient reported outcomes and a high level (90%) of return to sport in a young and active population (mean Tegner = 4.5) was achieved following robotically assisted UKA with combined ACL reconstruction. Radiographic results demonstrated a high level of accuracy using this image-less robotic platform, with the alignment objectives being achieved in 100% of cases.

UKA combined with ACL reconstruction is a technically more demanding procedure; however, the utilisation of robotic-assistance for this indication has not been described. Whilst early evidence suggests robotically assisted UKA may have superior survival outcomes compared to mechanical UKA [22], the outcomes of combined ACL reconstruction with UKA are limited to surgery performed with manual instrumentation. A systematic study by Volpin et al. in 2019 [11] reported pooled data from eight studies with 186 patients that underwent combined mechanical UKA and ACL reconstruction at a mean follow-up of 37.6 months. The global clinical results were very satisfying. Complications are summarized in Table 3. The most frequent complication was PE dislocation (n = 3, 1.6%) in two series using a mobile bearing [13, 23]. Achieving balancing may reduce the incidence of liner dislocation [24], and robotically assisted surgery aids this goal by providing anticipated gaps with implant panning. Additionally in this study, one (0.5%) patient had contralateral OA and 37 (20%) were observed to have radiolucency’s. Stephanie et al. [25] reported that these complications could be decreased by avoiding over or under correction, which is also aided by the assistance of the robotic system. In the current study, neither of these complications were observed.

Table 3 Comparison of our series to the literature
Full size table

None of those studies reported the radiological results in term of global accuracy. This criteria was yet studied by Batailler et al. [31] when performing standard UKA and found a significant difference between the mechanical and the robotic (16% of outliers versus 32% for medial UKA) with a lower rate of revision in the robotic group (5% versus 9%, NS).

The rate of arthrolysis under arthroscopy was high (n = 2, 20%), without any subsequent consequences and good recovery of flexion at the last follow-up. Arthroscopic arthrolysis for stiffness with UKA is not widely described in the literature, Fournier et al. [32] in a series of 22 arthrolysis and found that robotic-assistance was a protective factor against requiring surgery for stiffness. Derreveaux et al. [33] found that combined procedures with UKA (HTO, ACL, bicompartmental arthroplasty) had a high risk of stiffness (22% of arthrolysis). In the current study, stiffnesses was not correlated with excessive correction of the PTS (PTS = 6° each, correction = 4° and 5°). Several studies agreed to say that a greater PTS in TKA is correlated to a better flexion [34, 35]. However, those results cannot be extrapolated to UKA [36, 37]. The less a prosthesis is constraint the more antero-posterior stability depends on ligaments and a lower PTS [38]. Thus, Hernigou and Deschamps recommended a PTS between 3° and 7° with a significant increased risk of anterior tibial translation and no gain on the flexion beyond [36].

The average age of patients in this study was 57 years. An important objective of the intervention in this patient population is to facilitate ongoing participation in sporting activities. In the current study, one patient was disappointed and whilst not experiencing pain, was not able to return to any sport. Literature suggests that TKA is a less satisfying option for patients, wishing to return to sport. Witjes et al. in a literature review found a return to sport between 36 and 89% for the TKA versus 75 to 100% for UKA [39]. The contribution of robot-assistance for a retrun to sport was previously reported by Canetti et al. who found a faster return to sport in favour of the robotic group after a lateral UKA [40].

Monocompartmental arthritis may be treated with either UKA or HTO. A previous study in 2016 by Mancuso et al. comparing either HTO or UKA with ACL reconstruction found a higher complication rate with HTO (21%) than UKA (2.8%) and a comparable rate of revision (2.6% vs 2.8%) [41]. However this comparison may not be valid as HTO is often utilised in younger patients with moderate stage of OA [42]. Furthermore, HTO has been shown in some studies to allow a faster return to sport [43] and with less risks of conversion to TKA [44].

Most of the limitations mentioned by Figueroa et al. [45] in their state of the art concerning new technologies were not observed here: there were no pin-site related problems, the Navio® system does not need any pre-operative CT and the duration of the intervention (mean = 128 min) seems reasonable given the complex nature of the intervention. Previous studies have not reported their operating time.

The limits of this study are that it is retrospective in nature with a low number of patients and relatively short-term follow-up. However, to the best of our knowledge this is the first study of its kind reporting outcomes of robotically assisted UKA and combined ACL reconstruction. Besides, among the eight studies of the literature review of Volpin et al. [11], only four had more than 15 patients and only two had a mean follow-up greater than three years.

Conclusion

Imageless robotically assisted surgery is a precise and accurate tool to assist with the challenges faced when performing UKA with combined ACL reconstruction. This is the first series reporting its results for this specific indication and demonstrates excellent accuracy, good clinical results and an excellent rate of return to sport. Longer term follow-up and a larger comparative series are required to further understand the benefits gained from its use.