The main finding of this study confirmed our primary hypothesis that PFIA provides satisfying subjective outcomes at short-term follow-up in a selected group of patients. The overall failure rate of 11% within the first 2 years following implantation suggests reliability of the procedure and thus also confirms our primary hypothesis. Patella resurfacing at index surgery further lowered this failure rate to 4%. In general, certain pre- or perioperative characteristics, such as concomitant procedures addressing patellofemoral instability or malalignment, the lack of patellofemoral resurfacing at the index surgery or a high BMI, were predisposing factors for failure in our study, confirming our secondary hypothesis. Moreover, patients presenting with an increased BMI preoperatively and patients not undergoing patellar resurfacing at index surgery were significantly more likely to suffer from a worse postoperative outcome.
The results of this multicenter investigation, observing a favorable postoperative outcome, underscore the previously reported positive effect of the procedure per se in a large patient cohort for the first time [10, 12, 13]. While most of the results following implantation of the HemiCAP® Wave prosthesis range within the outcomes reported across multiple types of patellofemoral arthroplasties in a review of the literature, they surpass the collectively reported data in the transformed WOMAC-scoring .
The 2-year failure rate detected in this collective corresponds to the rates reported following implantation of comparable patellofemoral arthroplasty models [6, 30]. This demonstrates the validity of second-generation PFIA as a treatment option for isolated patellofemoral OA with prospects of favorable long-term survival rates. Studies investigating designs of first-generation patellofemoral arthroplasty, for which mid- and long-term follow-up is already available, show survival rates of 84% at a 10-year follow-up , 75–80% at 15-year follow-up [17, 30], and 69% at a 20-year follow-up . While the above-mentioned studies provide a possible range for long-term expectations for the HemiCap Wave model, the higher revision rates and lower survival rates of the first generation PFA-designs investigated in these long-term follow-up studies have to be taken into account [8, 25].
Regarding the results of the risk factor assessment, the presence of concomitant procedures addressing patellofemoral instability or malalignment as risk factors for failure are in line with previously published failure analyses. Moreover, previous investigations on patellofemoral malalignment in PFIA found patella alta and patellar subluxation  as well as patellofemoral maltracking  to be predictive for failure. In the large collective of this study, these findings could be extended to the general necessity for concomitant procedures addressing patellofemoral instability or maltracking. In these cases, concomitant surgery was performed according to an algorithm published by Imhoff et al. , to correct anatomical risk factors such as varus/valgus malalignment and insufficiency of the MPFL. The higher failure rate in these cases may root in the biomechanical principle of the patellofemoral inlay prosthesis a priori, as the possibility to intraoperatively correct patellofemoral maltracking is limited. As the medial and lateral trochlear edge are preserved, correction of rotation or alignment in a coronal plane are only possible to a limited extent . Thus, concomitant corrective procedures may fail to fully restore the physiological patellofemoral tracking desirable for optimal biomechanics of the PFIA—especially in cases of complex patellofemoral malalignment [21, 28, 34]. To address this malfunction, an implant design with a larger lateral dimension aimed at enhancing the tracking in complex maltracking pathologies is already available on the market .
Similar to our results, an increased BMI was identified as an independent factor predictive for an unfavorable outcome in another PFA model by Liow et al.. As an accepted risk factor for progression in knee OA , obesity may predispose for an early conversion to TKA—which remains the main cause for failure in PFA according to the current literature [3, 30, 32].
Not performing concomitant patellar resurfacing at the index surgery was identified as a further significant risk factor for failure. This may follow the rationale that additional patellar resurfacing mitigates the risk of progression of patellar OA and consequently pain—two main reasons for failed PFA treatment [3, 30]. This is supported by the finding, that secondary patellar resurfacing during follow-up of our cohort resulted in an elimination of the risk factor for failure “no patella resurfacing performed” at final evaluation. Indeed, biomechanical studies showed that implantation of a PFA significantly increases contact pressure of the patellofemoral compartment, creating a rationale for additional patella resurfacing [4, 33]. Biomechanical data from Vandenneucker et al. further demonstrated that superior restoration of the physiological kinematics of the patellofemoral joint can be achieved, when patella resurfacing is performed concomitantly . While studies addressing this question in PFA are scarce, extensive review of the literature in TKA demonstrated a lower revision rate when concomitant implantation of a patellar component was performed .
With the trend in surgery shifting to treatments of minimal invasiveness, results of modern PFIA treatment nevertheless have been benchmarked against TKA, the established treatment for OA of the knee joint. Biomechanically, PFA can sustain the physiological kinematics of the patellofemoral joint—in contrast to non-physiological conditions in the patellofemoral joint after TKA [27, 33]. Furthermore, it was shown that the ROM  and knee extension strength  are higher following PFA than TKA. Patient-reported outcomes following PFA were observed to be non-inferior to those reported after TKA while superior results were reported early after surgery  and in a young patient collective .
With comparable complication rates reported for both procedures in isolated patellofemoral OA , PFIA provides advantages over TKA including shorter rehabilitation, less morbidity, shorter intraoperative tourniquet time, preservation of the tibial/femoral bone stock [7, 31] and higher cost-effectiveness in younger patients .
While evidence investigating the outcome following patellofemoral arthroplasty has been mounting in recent years, patient satisfaction reporting is still scare . This multi-center study addresses this gap in knowledge the first time in a large patient collective, reporting a high patient satisfaction following PFIA.
While this study does demonstrate interesting findings, it is not without limitations. Firstly, while the data were collected prospectively, the study inherits the associated biases of a retrospective design. No statement about the pre- to postoperative changes could be made as no preoperative clinical scores were available and no control group could be established. Secondly, no radiographic evaluation at the final follow-up was conducted. Thirdly, as surgery was performed by specialists in the treatment of patellofemoral diseases in the respective centers, generalization to treatment with patellofemoral arthroplasty may be limited. Fourthly, there may be a performance bias in surgical technique across 11 different centers. However, benefitting from the comparative aspect of sampling in a multi-center approach may help better reflect general practice and reduce the selection bias of single-center design. Finally, to evaluate the outcome after successful PFIA treatment, failures were excluded from the outcome analysis. This potentially introduces a selection bias but avoids a confounding effect of TKA results. While this study reports outcomes and performs a failure analysis for a short- to mid-term follow-up period, further long-term follow-up is needed to conduct a meaningful comparison to different models of PFA and treatment with TKA.