Introduction

Treatment of isolated localized, full-thickness femoral condyle cartilage defects in middle-aged active patients is a challenge due to poor cartilage healing capacity and frequent disabling symptoms [15]. Also such cartilage lesion can progress into osteoarthritis (OA) [7]. Biological treatment options such as marrow stimulation and chondrocyte transplantation are influenced by patient age and have less favourable outcomes with increasing patient age [20, 21]. Total and unicompartmental knee arthroplasty are typically not indicated for these patients due to insufficient response to arthroplasty treatment in younger patients when osteoarthritis is not fully developed and increased risk of revision in young age groups [11, 14].

An anatomic metallic implant for femoral resurfacing (HemiCAP® Focal Femoral Condyle Resurfacing Prosthesis, Arthrosurface Inc, Franklin, MA, USA) was introduced in 2003 for treatment of full-thickness femoral chondral lesions. Both femoral condyle and a trochleal implant have been designed (Fig. 1) The indication for treatment with femoral resurfacing implants is symptomatic focal traumatic or degenerative osteochondral defects in femoral condyles or trochlea [16]. The implant has also been used for revision of failed chondrocyte transplantations [8, 12]. Biomechanical studies have demonstrated that the implants did not result in deleterious loading to opposing cartilage surface [2]. Biological response to femoral resurfacing implants has been tested in animal studies. In a goat model, implantation resulted in moderate synovitis, slight tibial chondral fibrillation and subchondral bone remodelling. Typically edge synovial overgrowth and good trabecular bone ingrowth to the fixation anchors were demonstrated [19].

Fig. 1
figure 1

The HemiCAP implants. a The femoral condyle implant. b The trochlear implant

Full size image

Only limited evidence of clinical outcome and failure rates has been presented. Two case series both with around 20 patients with varying osteochondral pathology have demonstrated reduced pain and improved knee function after treatment based on improvements in subjective outcome scores [3, 5]. No studies have present clinical outcome with both condylar and trochleal HemiCAP® implantations. There is a need for better clinical evidence of outcome in larger patient materials treated with both condylar and trochlea implants.

The aim of the present study was to present the clinical outcome in a prospective cohort of 61 patients treated with HemiCAP® femoral resurfacing implants for both condylar and trochleal cartilage lesions. It was hypothesized to find reduced pain and improved knee function after femoral resurfacing treatment.

Materials and methods

From 2007 to 2012, 61 patients were treated at a single centre with femoral resurfacing using the HemiCAP implant. Thirty-six patients received the femoral condyle implant and 25 patients the trochlea implant. Indication for treatment with femoral resurfacing implant was symptomatic cartilage lesion at the femoral condyles or trochlea demonstrated by MRI or arthroscopy, which was ICRS grade 3–4 and size less than 4 cm2. Patients were not offered treatment in case of age below 35 and above 65, valgus or varus malalignment of >5 degrees ligament instability, more than 50 % meniscus removal, failed previous cartilage procedures (microfracture, debridement) and BMI > 40. There were 28 males and 33 females, with median age of 49 (35–65). Patients were followed for 2–7 years. The demographic characteristics of the total patient material and the subgroups of femoral condyle and trochlea implant treatments are given in Table 1.

Table 1 Demographic characteristics of the patient material
Full size table

The study was approved by the regional ethical committee # 1-10-72-219-34.

Outcome evaluation

Patients were assessed after 3 months, one and 2 years by radiologic status, Knee Society Scores (KSS) objective and function subscales [17] and pain score using numerical rank scale (0–10), with 10 being worst possible level of pain. For the KSS function score, we defined a response rate of good outcome as having an increase of 20 points or greater in the KSS function preoperatively to follow-up time.

Complications and reoperations were registered. Analgetic consumption preoperatively and at follow-up time points was assessed to document improvements in pain level.

Radiographic evaluation

Osteoarthritis (OA) grade was evaluated by Kellgren–Lawrence grade for both medial and lateral tibiofemoral compartments preoperatively and at 2-year follow-up to investigate OA development after surgery [18]. The fractions of patient with different degrees of Kellgren–Lawrence grading are presented as well as the fraction of patient that has changed to a more severe OA grade.

Device description

The HemiCAP resurfacing implant consists of two components: a fixation component and a modular articular component, connected with a Morse Taper (HemiCAP® Focal Femoral Condyle Resurfacing Prosthesis, Arthrosurface Inc, Franklin, MA, USA). The fixation component is a titanium cancellous screw with full-length cannulation. The cobalt chrome articular component is available in 15 or 20 mm diameter size and comes in 16 different offset configurations which correspond to the superior/inferior and medial/lateral radius of curvatures at the implant site. The trochlear implant is 15 or 20 mm in diameter with a single concave shape fitting the trochleal sulcus (Fig. 1). A polyethylene inlay is available for the patella and is recommended to be used in combination with the trochlear implant.

Surgical procedure

The procedure is initiated with a standard arthroscopy to identify cartilage status and indication and treat any concomitant intraarticular pathology. The defect is exposed using a small para-patellar incision. The cartilage lesion is sized. A special centralized drill guide is used to place a k-wire perpendicular and central to the articular cartilage surface. Over the k-wire, reaming for fixation screw is performed and the fixation screw is implanted into bone. Mapping instruments measure the surface curvature, and a matching surface reamer prepares the inlay implant bed. Sizing trials are used to confirm an accurate fit to the surrounding cartilage. The resurfacing implant is fixed press-fit onto the fixation screw and was seated flush or slightly recessed 0.5 mm to the surrounding articular cartilage surface. A standardized rehabilitation protocol with free range of motion was allowed immediately postoperatively. For the first 2 weeks, patients used touch weight bearing where after full weight bearing was allowed.

Statistical analysis

(Demographics and baseline characteristics were analysed with the use of descriptive statistics presented as mean value and standard deviations. Chi-square and Student’s t test were used for comparison of proportional and paired data. The Kaplan–Meier survival analysis was used with revision or dead as endpoint and with 95 % confidence interval (CI). For the statistical analysis, the software program Stata 13 (StataCorp, College Station, Texas, USA) was used. p values <0.05 were considered statistically significant.

Results

Clinical and radiographical evaluation at 12-week, 1- and 2-year follow-up was completed by 100, 85 and 67 % of patients, respectively. The reason for lost to follow-up was patients not returning for scheduled follow-up visit. The evaluation of the failure endpoint revision to arthroplasty was completed of all patients, since a national registry for knee arthroplasty exists in our country ensuring tracking of all arthroplasty procedures.

Objective and subjective outcomes (Table 2)

Objective outcome based on objective KSS objective knee score improved significantly for both condylar and trochlear implants. KSS improved for condylar implants from 54 (6.4) to 92 (8.4) (p < 0.001). For trochlear implants, KSS score improved form 50 (5.3) to 88 (7.5) (p < 0.001). The subjective KSS function and pain scores also improved significantly for both implant types. There was a tendency to more pronounced pain relief for condylar implants compared to trochlear implants.

Table 2 Objective and subjective outcomes
Full size table

The proportion of patients, which had a clinical relevant improvement in HSS function score of >20 point, was above 90 % at one- and 2-year follow-up. There was no tendency to any difference in this response rate between trochlea and condyle implant patients.

Analgetic consumption

Preoperatively, all patients used some type of oral analgetic medicine. Paracetamol, NSAID and morphine were used preoperatively in 100, 89 and 23 % of patients, respectively. At 2-year follow-up, analgetic consumption was reduced to 31, 23 and 0 %, respectively. No analgetic medicine was used in 31 % of patients at 2-year follow-up.

Radiographic osteoarthritis evaluation

Preoperatively, 28 and 13 % of patients have Kellgren–Lawrence OA grade 2 in medial and lateral compartments, respectively. The remaining patients had KL grade 1. At 2 years, Kellgren–Lawrence OA grade 2 was seen in 40 and 25 % of patients in medial and lateral compartments, respectively. For the medial and lateral compartments, 29 and 25 % of patients demonstrated a worsening in OA from Kellgren–Lawrence OA grade 1 to grade 2.

Failures and complications

Fourteen (23 %) of implants were revised to arthroplasty due to progression of cartilage lesions over the 7-year follow-up period (Fig. 2). Nine of 36 femoral condyle implants (25 %) and 5 of 25 trochlear implants (20 %) were revised. In seven cases, periimplant unicompartmental cartilage injury progression was seen. In four cases, multicompartmental cartilage injury progression was seen. In three cases, periimplant cartilage injury progressing and implant screw loosening were seen. No difference between females and males and between condylar and trochlear implants was found for reoperation rate.

Fig. 2
figure 2

Kaplan–Meier survival curve of implant survival with revision surgery to hemi or total knee arthroplasty as endpoint

Full size image

Discussion

The most important finding of the present study was that condylar resurfacing treatment with the HemiCAP implant has demonstrated clinically relevant improvements in both function and pain reduction in patients with unicompartmental localized degenerative cartilage lesion. Also no difference in clinical outcome has been found between femoral condyle lesions or trochlear lesion treated with resurfacing. Another important finding was that reduction in pain resulted in a significant reduction in analgesic medicine usage.

The results are in agreement with the few previously published studies with condylar resurfacing treatment. Bollars et al. [5] found significant improvement in subjective score in 27 patients treated with the HemiCAP implant. The subjective outcome in that study demonstrated similar level as a normal matched population.

Becher et al. [3] demonstrated similar significant improvement in subjective outcome in 21 patients. Similar to the present study, none of these studies found any complication with aseptic implants loosening. A recent review of outcomes after condylar resurfacing treatment concluded good early to midterm results, which is supported by the results of the present study [6]. The present study represents the largest case series with HemiCAP femoral resurfacing treatment presented until now.

A relatively high proportion of 23 % was found that needed reoperation with revision to some type of knee arthroplasty. Such reoperation rate has also been demonstrated in national knee arthroplasty registries. The Australian Arthroplasty Registry has presented revision rates of 28 % [1]. The causes for these reoperations were symptom or osteoarthritic progression. This could indicate that the cartilage pathology of early degenerative changes in some instances is a progressive condition that is not haltered by local implant resurfacing. This is to some degree supported by the radiographic findings in the present study with 29 and 25 % OA grade deterioration after 2-year follow-up in medial and lateral compartments, respectively.

Other cartilage treatment modalities for large cartilage lesions such as autologous chondrocyte transplantation and allogenic osteochondral grafting have demonstrated various failure rates. Autologous chondrocyte transplantation has demonstrated graft failure in 26 % of cases, but failure after this procedure is normally new biological cartilage restoration and not arthroplasty [4]. Allogenic osteochondral grafting has demonstrated revision to arthroplasty in 10 % after 10 years [13]. However, both of these treatments are typically offered to patients younger than 35 years in which clinical treatment failure does not automatically lead to an arthroplasty reoperation offer.

The middle-aged patient (35–65 years) with symptomatic early degenerative knee cartilage pathology is a demanding patient group, expecting treatments to result in pain-free activities of daily living and high activity levels at work and for recreation. Since conventional arthroplasty treatment is often not offered to patients when only early osteoarthritic changes exist despite severe symptoms, there exists a treatment gap for these patients. Recent studies have estimated that 20 % of symptomatic knee OA in USA falls into this treatment gap (London and Miller 2011).

The good results of the present study with clinically relevant improvement in function and pain reduction suggest that femoral resurfacing treatment can be used with predictive outcome in patients between 35 and 65 years with localized femoral condyle cartilage lesions. Since degenerative cartilage changes appear to progress in some patients resulting in need for revision surgery in 23 % of patients within 7 years, it is important to consult with patients that the treatment primarily is aimed at providing temporary pain relief and function improvement rather than a permanent solution for early degenerative cartilage pathology. Also the high revision rate is indicating a need for careful indication selection to HemiCAP treatment. The implant should only be used for younger patients with isolated deep cartilage lesions not suitable for biological repair in knees without clear degenerative changes. Most likely deep osteochondral pathology such as old osteochondritis dissecans and osteonecrosis should not be treated with a resurfacing implant due to insufficient subchondral pathological bone removal. In patients with valgus malalignment and in cases of bone-on-bone cartilage lesion in one tibiofemoral compartment, surgical treatment with axis correctional high tibial osteotomies or unicompartment arthroplasty results in good outcome with low revision rates to arthroplasty of 7 % after 15 years for unicompartmental knee arthroplasty and of 14 % for proximal tibial osteotomy [9, 10].

The findings of clinically relevant improvement in function and pain reduction suggest that femoral resurfacing treatment can be used in patients between 35 and 65 years with localized femoral condyle cartilage lesions. The condylar implants should be considered a temporary treatment for early OA, since the disease in a significant proportion of patient will progress and arthroplasty treatment then become necessary.

An advantage of the present study is that patient number is large enough to present subgroup outcome data on both condylar and trochlear implants, which has not been presented in previous studies. The present study is limited by investigating only the short-term to mid-term clinical results of condylar resurfacing treatment. The average follow-up was 41 months, and 66 % of the patients had more than 2 years of follow-up, leaving 33 % of patients being lost to clinical and radiographic follow-up. However, arthroplasty reoperation follow-up was complete due to national registry data. During the first 2 years, postoperative patients have typically returned to the level of activity they may possibly achieve, and symptoms from remaining surgical morbidity typically do not improve any further. Therefore, a 2-year follow-up time is clinically relevant. A prospective case series as the present study yields a heterogeneous patient material with respect to cartilage pathology and previous surgery. This heterogeneity results in a study population that is typical for the patient population suffering from significantly symptomatic cartilage lesions. Taking into account the follow-up rate and the two different implants presented in our case series, then the present case series size is not superior to previous studies.

Conclusions

The present study demonstrated improved subjective outcome and reduced pain after femoral resurfacing using the HemiCAP implant in a relatively large cohort of patients with symptomatic cartilage lesions. A concerning 23 % reoperation rate with conversion to arthroplasty was found. Femoral resurfacing implantation treatment can be a temporary treatment for cartilage lesions expected to develop into osteoarthritis and for younger patients not eligible for arthroplasty treatment.