Influence of the anterior notch in mobile-bearing UKA on patellofemoral radiotracer uptake and clinical outcome

In mobile-bearing unicondylar knee arthroplasty (UKA) creation of an osteochondral notch adjacent to the femoral implant is mandatory to prevent anterior impingement of the bearing (Fig. 1a). [1]. A notch size of 3 to 5 mm was recommended, but in daily practice the size necessary to prevent impingement in full extension can vary considerably, dependent on implant sizing, positioning and surgeon preference (Fig. 1b).

However, the lateral aspect of that notch is part of the deep femoral trochlea, where the patella articulates in deep knee flexion (Fig. 2). It was reported that even these ‚transition zones’ are exposed to relevant contact pressure [2]. Others emphasized the importance of a smooth transition between the femoral implant and the anterior region of the femur [3‐5] (trochleocondylar junction). This is, however, not possible in mobile-bearing UKA as the surgeon has to create the above-mentioned notch to provide space for the anterior lip of the polyethylene insert.

It was the purpose of the study to investigate whether the surgically created, anterior notch in mobile-bearing UKA was associated with clinical PFJ outcome and PFJ radiotracer uptake. Consequently, we hypothesized that there would be significant correlations between the parameters 1) notch size, 2) clinical outcome in terms of PFJ sensitive knee scores and 3) PFJ radiotracer uptake.

Patients scheduled for routine mobile-bearing UKA implantation due to osteoarthritis or osteonecrosis in the medial compartment were considered for enrollment. Contraindications were: 1) failed upper tibial osteotomy, 2) insufficiency of the collateral or anterior cruciate ligaments, 3) fixed varus or valgus deformity greater than 15°, 4) flexion deformity greater than 15°, and 5) rheumatoid arthritis. The study was conducted in accordance with the Declaration of Helsinki and informed consent was obtained from all subjects prior to participation.

All surgical procedures were performed under general or spinal anaesthesia under tourniquet control and after standard antibiotic prophylaxis. Patients received the ‘Oxford 3’ medial UKA (Zimmer Biomet Inc., Warsaw, Indiana, USA). The surgical technique was as recommended in the manufacturer’s surgical manual. Accordant to these recommendations an osteochondral resection was performed anterior to the implant to prevent impingement of the polyethylene inlay in full knee extension. After implantation the wound was closed over drainages. All operations were performed by an experienced consultant surgeon, and prophylaxis against venous thrombosis was administered in all cases. Patients underwent a standard rehabilitation program after surgery, consisting of continuous passive motion, and active and passive exercise under the guidance of a physiotherapist. Subsequently, most of the patients attended an outpatient rehabilitation program.

With regard to the above-mentioned hypotheses we determined the following outcome parameters: 1) notch size, 2) clinical PFJ outcome and 3) PFJ radiotracer uptake.

The size of the notch adjacent to the femoral implant was measured intraoperatively by always the same observer using a sliding calliper. This was done at 10:30 o’clock in right knees and at 1:30 o’clock in left knees. This location was chosen to determine how far the notch extends in the direction of the deep femoral trochlea (where the patella articulates in deeper knee flexion) (Fig. 3). Measurements were done at the bony level (bottom of the notch) and at the cartilage level (surface of the notch) to determine the parameters ‘notch size bone’ and ‘notch size cartilage’, respectively (Fig. 4).

Clinical knee score outcome was assessed with the ‘Knee Injury and Osteoarthritis Outcome Score (KOOS)’ [6] and the Kujala Score [7]. High sensitivity for PFJ disorders, and favourable findings with regard to validity, reliability and responsiveness were reported for these scores [8‐10]. On both the Kujala Score and the KOOS zero represents extreme knee problems, and 100 no knee problems. The knee scores were assessed preoperatively and six months postoperatively, always in a self-administered fashion.

PFJ bone remodelling around the notch was determined by ‘single photon emission computed tomography – computed tomography’ (SPECT-CT) six months after implantation. Three hours after administration of 500 MBq 99mTc-3,3-diphospho-1,2-propandicarbonacid (DPD; Teceos®) a whole body planar scan was acquired followed by a SPECT/CT of the knee region, covering a field of view of 45 cm. The exams were performed on a dedicated SPECT-CT system equipped with a dual-head gamma camera and a low-dose CT device (Philips Brightview XCT®). The particular value of SPECT-CT in assessing postoperative knee disorders was already demonstrated [11, 12].

Data processing was performed as follows. The Kujala Score and the KOOS were evaluated as described in the original publications [6, 7]. For analysis of SPECT-CT data the software ‘Hybrid Viewer’ (Hermes Medical Solutions, Stockholm, Sweden) was used. Two areas were of interest: the area of the notch and the entire medial facet of the patella. Around both areas volumes of interest (VOI) were determined with always the same standardized technique. We determined regions of interests (ROIs) in 2D manually e.g. in transverse slices of the medial patella. At the level of the patella’s maximum mediolateral width the mediolateral axis of the patealla was defined for reference purposes. A perpendicular line (anteroposterior direction) was defined and always positioned through the patellar ridge at the patella’s articular surface. A ROI was defined that included all aspects of the patella medial of that line (Fig. 5). Many ROIs then comprised the VOI. Within the VOI the maximum tracer uptake was determined. The maximum uptake was chosen as appropriate parameter a) because it was less dependent on variations in VOI definition and b) it was also chosen by previous studies dealing with SPECT-CT and knee disorders as appropriate outcome parameter [13‐16].

Because of inter-individual tracer uptake differences caused by patient weight or variable tracer dispersion a ratio was calculated between the maximum uptake within the two above-mentioned VOIs and a reference VOI from cancellous bone from the contralateral femur (normalized maximum uptake).

For statistical analysis the software programme SPSS (International Business Machines Corporation, Armonk, NY, USA) was used. Because the case number was <30 the data were regarded as not normally distributed. Accordingly, medians and interquartile ranges (IQR) were determined as descriptives. For inferential statistics Spearman correlation coefficients were calculated and alpha was defined as 0.01 because of multiple analyses. Correlation strength was defined as follows: very high from 0.90 to 1.00, high from 0.70 to 0.90, moderate from 0.50 to 0.70, low from 0.30 to 0.50 and negligible from 0.00 to 0.30 [17]. The sample size was chosen pragmatically.

The 24 patients (11 male, 13 female) had a median age of 64.5 years (IQR 14) and a body mass index of 27.84 (IQR 3.99). 11 patients underwent right-sided UKA and 13 patients left-sided UKA. The patients’ mechanical tibiofemoral angle was 2.5° varus preoperative and 2.0° postoperative.

The KOOS and the Kujala Score improved from preoperative to 6 months postoperative (Table 1), but this was not linked to a hypothesis or inferential statistical tests. The notch size was 7 mm at the cartilage level (IQR 3.5, range 3–12 mm, 95%: 6.3–8.3) and 4.25 mm at the bone level (IQR 3.5, range 2–12 mm, 95% CI: 4.4–6.5).

With regard to our hypotheses the size of the anterior notch showed no significant correlations with the maximum tracer uptake at the patella or the trochleocondylar junction (p = n.s.). Similarly, no significant correlations were seen between the maximum tracer uptake (patella or trochleocondylar junction) and the knee scores (KOOS or Kujala Score) (p = n.s.). However, \several low and moderate positive correlations were found between the notch size and the knee scores 6 months postoperatively (see Table 2 for details).

The most important findings were the absence of significant correlations between a) notch size and PFJ radiotracer uptake or b) knee score outcome and PFJ radiotracer uptake and the presence of significant correlations between c) ‚PFJ-sensitive’ knee scores and notch size.

Although the PFJ was already issued in the context of UKA in previous research [18‐22], the specific issue of the current study has so far not been dealt with. No previous researchers applied SPECT-CT in a case series of mobile-bearing UKA to determine PFJ bone remodelling in the trochleocondylar junction. The findings of this study suggest that the size of the notch is not related to the amount of radiotracer uptake either at the notch itself or at the articulating part of the patella. Previous research suggested that increased radiotracer uptake can be interpreted as increased bone stress [23]. Therefore, it may be concluded that greater notch sizes are not associated with increased stress at the notch itself or the articulating aspects of the patella.

Low and moderate correlations were found between ‚PFJ-sensitive’ knee scores and notch size. Surprisingly, the coefficients indicated an inverse relationship to what had been assumed by us: the bigger the notch the better the knee score. Thus, based on the findings it can be concluded that a larger notch size in UKA does not negatively influence PFJ function. Indeed, it seems that a larger notch was associated with superior clinical outcome. It might be speculated whether, when in doubt, a larger anterior notch should be resected in medial mobile-bearing UKA. However, due to the small sample size our findings should be interpreted with caution.

Several authors reported on fixed-bearing UKA and emphasized the importance of a smooth transition between the anterior part of the femoral implant and the trochlea [3‐5]. With regard to mobile-bearing UKA previous researchers reported anterior impingement of the bearing on the femur as a possible reason for surgical reintervention [24, 25]. We believe that this specific mechanism might also be applicable when interpreting the findings of the current study. This means that a larger notch might better prevent bearing impingement in full extension However, the prevention of anterior impingement of the bearing might not be the only potential reason for the positive correlations between knee scores and notch size. It might as well be speculated that notch size is just a surrogate variable for alignment of the femoral component in the sagittal plane. Implantation of the femoral component in slight flexion usually leads to more protuberance of the anterior lip of the bearing (over the anterior margin of the femoral component) in full knee extension. Consequently, a larger notch is created by the surgeon to accommodate for this protuberance and to prevent impingement of the bearing at the femoral bone. In other words, it could be that positioning the femoral component in slight flexion is the true factor that leads to superior score outcome. With this regard, it is also interesting that the manufacturer just recently recommended positioning the femoral component in slight flexion.

It is surprising that the notch size did not correlate with a biomechanically more closely related parameter, i.e. the radiotracer uptake, whereas it did correlate with a more high-level outcome parameter such as the knee scores. The exact reason for that counterintuitive constellation is not clear to us.

It is also surprising that the correlations between notch size and knee scores were observed heterogenously over type of subscores (KOOS symptoms and KOOS quality of life) (Tab. 2). We cannot provide explanations for the inhomogenous incidence of correlations over the different subscores.

Interestingly, the manufacturer of the Oxford mobile-bearing UKA recently introduced a special milling system to create the notch adjacent to the femur in a standardized fashion (Art.No. 32–423,238 and 32–423,239). Thus, the anterior notch in mobile-bearing UKA would appear to be a technical detail that is currently receiving increased attention.

Nevertheless, the study at hand is the first publication to investigate the consequences of the requisite anterior notching in mobile-bearing UKA on the PFJ. We therefore believe it contributes to the current scientific knowledge. The strengths of the study also lie in the in-vivo measurement of PFJ bone remodelling with SPECT-CT fusion, as stimulated by previous publications [26, 27].

From the findings made in our study it is concluded that a larger size of the anterior notch in mobile-bearing medial Oxford UKA is not associated with increased osteochondral remodelling processes at the patella or the trochleocondylar junction. Neither is a larger sized notch associated with worse clinical PFJ outcome. Surprisingly, a larger notch was associated with superior clinical outcome. The exact mechanism for this contraintuitive finding remains unclear but may be the basis for future research.