Interventional procedure overview of focal resurfacing implants to treat articular cartilage damage in the knee
Closed for comments This consultation ended on at Request commenting lead permission
Summary of key evidence on focal resurfacing implants to treat articular cartilage damage in the knee
Study 1 Elbardesy H (2021)
Study type | Systematic review and meta-analysis |
|---|---|
Country | Not described for individual studies |
Recruitment period | Not described for individual studies |
Study population and number | n=14 studies, 464 people People with a focal femoral condyle cartilage defect |
Age & sex | 47.9 years, 62.5% female (sex of 155 people was not reported) |
Patient selection criteria | Inclusion criteria: any clinical trials involving HemiCAP, UniCAP (Arthrosurface) or other focal resurfacing implant with mean follow up at least 2 years. Exclusion criteria: all cadaveric, biomechanical studies, and studies about partial resurfacing of the patellofemoral joint. |
Technique | Focal femoral condyle resurfacing with HemiCAP, UniCAP, Episealer (Episurf), or BioPoly (BioPoly) – see Other issues below. |
Follow-up | 1.56 to 11.7 years |
Conflict of interest/source of funding | Conflict of interest: the authors report that they have no conflict of interest. Source of funding: the authors report no funding source was received. |
Analysis
Follow up issues:The proportion of people lost to follow up ranged from 0 to 63.6%.
Study design issues: This systematic review and meta-analysis evaluated the outcomes of focal resurfacing of full-thickness cartilage defects. The study was reported to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The meta-analysis was conducted on KOOS. KOOS is a as well as subgroup analyses of KOOS component scores – pain, symptoms, activities of daily living, quality of life, and sport and recreational activities. Risk of bias of included studies was conducted according to the Cochrane risk of bias criteria.
Mean differences were computed with 95% CI discontinuous outcomes, using standard meta-analysis software (RevMan 5.3). Standardised mean differences were used to compute effect measures and a random-effects model for meta-analysis was used. I2 statistic was used to quantify heterogeneity.
Other issues: The publication is written in poor English with multiple typographical errors. One of the studies (Nathwani [2017]) is included in the meta-analysis but is incorrectly described as using the HemiCAP implant. Given Nathwani (2017) is a key publication for BioPoly, it is extracted as Study 6 in this overview. Another study, Stalmån (2018) is incorrectly described as using the HemiCAP implant. Stalmån (2018) used the Episealer implant and is included in the Appendix.
Key efficacy findings
KOOS
Number of people analysed: 127 (before surgery); 90 (after surgery)
Follow up at time of assessment: 2 years
There was a statistically significant improvement in each subset of the KOOS instrument from before surgery to after surgery, however, the studies were heterogenous, indicated by an I2 values of 96 and 97% for all outcomes.
Pain: standardised mean difference was -5.61 (95% CI -8.11, -3.11), showing lower pain after surgery.
Symptoms: standardised mean difference was -4.96 (95% CI -7.28, -2.63), inferring lesser symptoms after surgery.
Activities of daily living: standardised mean difference was -5.08 (95% CI -7.40, -2.76), showing lesser difficulty with performing activities of daily living after surgery.
Sport and recreational activities: standardised mean difference was -4.35 (95% CI -7.08, -1.61), showing lesser difficulty with performing sports and recreational activities.
Quality of life: standardised mean difference was -4.35 (95% CI -7.08, -1.61), showing lesser difficulty with performing sports and recreational activities.
Key safety findings
Revisions
Number of people analysed: 464
Follow up at time of assessment: variable, from 1.6 years to 11.7 years
The revision rate ranged from 0% (6 studies) to 61% (1 study)
The highest revision rates were seen in 2 studies using the UniCAP implant – 61% (Laursen [2019]) and 47% (Laursen [2016]).
These studies included people with relatively larger lesions.
Study | Condyle defect size and implant used | Follow up length, years | Revisions, n/N (%) |
Becher C, Kalbe C, Thermann H et al. (2011) Archives of orthopaedic and trauma surgery 131(8):1135-43 | ≤20 mm2 HemiCAP | 5.3 | 2/21 (10%) |
Becher C and Cantiller EB. (2017) Archives of orthopaedic and trauma surgery 137(9):1307-17 | NR HemiCAP | 11.7 | 0/2 (0%) |
Bollars P, Bosquet M, Vandekerckhove B et al. (2012) Knee surgery, sports traumatology, arthroscopy 20(9):1753-9 | 37% 20 mm2/63% 15 mm2 HemiCAP | 2.8 | 0/19 (0%) |
Dhollander AAM, Almqvist KF, Moens K et al. (2015) Knee surgery, sports traumatology, arthroscopy 23(8):2208-2212 | 50% 15 mm2/50% 20 mm2 HemiCAP | 2.2 | 0/14 (0%) |
Laursen JO. (2016) Knee surgery, sports traumatology, arthroscopy 24(5):1695-701 | >4 cm2 UniCAP* | 2 Clinical 7 Complications and reoperations | 30/64 (47%) |
Laursen JO and Lind M. (2017) Knee surgery, sports traumatology, arthroscopy 25(3):746-51 | <4 cm2 HemiCAP** | 2 Clinical 7 Complications and reoperations | 9/36 (25%) |
Laursen JO, Mogensen CB, Skjøt-Arkil H. (2019) Knee Surgery, Sport traumatology, arthroscopy 27(5):1693-7 | >4 cm2 UniCAP | 7.2 | 36/59 (61%) |
Miniaci A. (2014) Clinical Sports Medicine 33(1):57-65. | NR UniCAP* | 1.6 | 0/35 (0%) |
Nahas S, Monem M, Li L et al. (2020) The journal of knee surgery 33(10):966-70 | NR HemiCAP | 9.8 | 2/14 (14%) |
Nathwani D, McNicholas M, Hart A et al. (2017) JB&JS Open Access 2(2) | 2.7 cm2 ± 0.6 cm2 BioPoly* | 2.0 | 0/33 (0%) |
Pascual-Garrido C, Daley E, Verma NN, and Cole BJ. (2017) Arthroscopy: the journal of arthroscopic & related surgery 33(2):364-73 | NR HemiCAP | 2.0 | 8/32 (25%) |
Stålman A, Skoldenberg O, Martinez-Carranza N et al. (2018) Knee surgery, sports traumatology, arthroscopy 26(7):2196-204 | ≤3.2 cm2 Episealer* | 2.0 | 0/10 (0%) |
Çepni Ş, Veizi E, Tahta M et al. (2019) Archives of orthopaedic and trauma surgery 140(2):209-18 | 3.6 cm2 ± 0.5 cm2 HemiCAP (62%) or BioPoly (38%)*** | 2.0 | 13/118 (11%) total 5/73 (7%) HemiCAP 8/45 (18%) BioPoly |
Hobbs H, Ketse-Matiwane N, van der Merwe W et al. (2013) SA Orthopaedic Journal | <4 cm2 HemiCAP | 4.7 | 2/7 (29%) |
*Incorrectly listed in Elbardesy (2021) as HemiCAP.
**Incorrectly listed in Elbardesy (2021) as UniCAP.
***Incorrectly listed in Elbardesy (2021) as HemiCAP only.
Study 2 Australian Orthopaedic Association National Joint Replacement Registry (2020)
Study details
Study type | Registry analysis – not published in a peer-reviewed journal |
Country | Australia |
Recruitment period | 1999 to 2020 No partial knee resurfacing procedures were conducted in 2020, data here was extracted from the 2020 report (with data up to 2019). |
Study population and number | n=245 procedures All partial knee resurfacing procedures reported to the Australian Orthopaedic Association National Joint Replacement Registry. |
Age & sex | Not reported |
Patient selection criteria | All partial knee resurfacing procedures reported to the registry. Partial resurfacing was defined as 'involves the use of one or more button implants to replace part of the natural articulating surface on one or more sides of the joint, in one or more articular compartments of the knee.' |
Technique | All procedures used the HemiCAP range of implants. Most (n=145) were implanted on the femoral articular surface. There were 85 procedures that involved resurfacing the patella/trochlear joint either on 1 side (n=27) or both sides (n=58). As the specific implant type was not specified, some of these implants may have been the HemiCAP Wave or Wave Kahuna (see Rapid review of literature). |
Follow-up | Up to 12 years |
Conflict of interest/source of funding | Conflict of interest: Not reported Source of funding: Not reported |
Analysis
Study design issues: This multicentre, prospective registry analysis reported the revision rate of partial knee resurfacing procedures conducted in hospitals in Australia.
Other issues: The data was published as a report and not in a peer-reviewed journal.
Key safety findings
Revisions
Number of people analysed: 220 procedures
Follow up at time of assessment: up to 12 years
The cumulative percent revision of partial knee resurfacing for osteoarthritis was 49.5% at 12 years.
Most primary partial resurfacing implants were revised to either a total knee replacement (65.0%) or a unicompartmental knee replacement (19.0%).
Cumulative percent revision of primary partial resurfacing knee replacement (primary diagnosis osteoarthritis) | ||||||||
N total | N revised | 1 year | 2 years | 3 years | 5 years | 8 years | 12 years | |
Partial knee resurfacing | 220 | 93 | 6.4 (3.8, 10.5) | 16.4 (12.1, 22.0) | 17.8 (13.3, 23.6) | 27.6 (22.1, 34.1) | 37.9 (31.4, 45.2) | 49.5 (41.8, 57.8) |
Number at risk | - | - | 206 | 181 | 174 | 140 | 87 | 42 |
Study 3 Christensen BB (2021)
Study details
Study type | Registry analysis |
Country | Denmark |
Recruitment period | 1997 to 2020 |
Study population and number | n=379 implants All knee resurfacing procedures reported to the Danish Knee Arthroplasty Registry. |
Age & sex | Mean 50 years; 57% female |
Patient selection criteria | Reporting to the registry is mandatory for both public and private hospitals and data was collected directly by the surgeon through standardised forms. |
Technique | A resurfacing implant was defined as UniCAP, HemiCAP, or Episealer. Most implants were isolated HemiCAPs (n=231; 61%). Other implants were isolated UniCAPs (n=112; 30%), isolated Episealers (n=20; 5%), or more than 1 implant in combination (n=16; 4%). As the specific implant type was not specified, some of these implants may have been the HemiCAP Wave or Wave Kahuna (see Rapid review of literature). |
Follow-up | Median 8 years (IQR 3 to 10 years) |
Conflict of interest/source of funding | Conflict of interest: The authors declared no conflict of interest. Source of funding: The authors declared that funding was not received. |
Analysis
Study design issues: This retrospective analysis of prospectively collected data from the multicentre Danish Knee Arthroplasty Registry assessed the survival of knee resurfacing implants. Survival of the resurfacing implants was primarily analysed by the Kaplan-Meier method with observations included at the date of index surgery and with date of revision surgery as endpoint.
Study population issues: The majority of people (87%) were Charnley A class, indicating an issue with a single joint only. The predominant indication for resurfacing implants was secondary osteoarthritis (42%), followed by primary osteoarthritis (32%), and osteochondral lesions (20%). Most of the observations had prior knee surgery in the affected knee with arthroscopy (60%) and microfracture procedures (22%) being the most frequent.
Key safety findings
Revisions
Number of people analysed: 379 implants
Follow up at time of assessment: median 8 years, up to 12.5 years
A total of 70 implants (19%) were revised to arthroplasties. This corresponded to estimates of:
1-year revision-free survival: 0.95 (95% CI 0.93 to 0.97)
5-year revision-free survival: 0.84 (95% CI 0.80 to 0.88)
10-year revision-free survival: 0.80 (95% CI 0.75 to 0.84)
Median time to revision: 2 years (IQR 1 to 4 years).
The revised implants were mainly UniCAP (n=35, 50%) and HemiCAP (n=33, 47%), with Episealer (n=1,1.5%) and combined implants (n=1, 1.5%) less common.
Most revisions (n=61; 87%) were converted to cruciate-retaining total knee arthroplasties.
Study 4 van Buul GM (2021)
Study details
Study type | Single centre, before-and-after study |
Country | Ireland |
Recruitment period | 2009 to 2013 |
Study population and number | n=132 People who had focal articular cartilage resurfacing. |
Age & sex | Mean 35.7 years; 58.3% male |
Patient selection criteria | Inclusion criteria: not reported. Exclusion criteria: any tibial articular wear. |
Technique | Implant: HemiCAP (n=102), UniCAP (n=11), and HemiCAP PF Classic (n=19) Technique: implant was fixed into a titanium screw via morse taper after preparation of the defect with a unique mapping technique, according to the manufacturer's instructions for use. After surgery care: non-weight-bearing, limited flexion for 2 weeks. Then weight-bearing as tolerated, and light activities at 3 weeks. People were advised to avoid running until 3 months after surgery. Contact sports were encouraged after 6 months. |
Follow-up | 4 years |
Conflict of interest/source of funding | Conflict of interest: The authors declare that there was no conflict of interest. Source of funding: Not reported |
Analysis
Study design issues: This single centre, before-and-after study assessed the outcomes of resurfacing in people with focal cartilage defects. The primary outcome was WOMAC score, where the maximum score, 96, represents the best outcome. Other outcomes included revisions and complications.
Paired t-tests were used for analysing consequential data within groups. p<0.05 was considered statistically significant.
Study population issues: All people included in the study had previous treatment, with the majority (92.4%) having microfracture.
Key efficacy findings
Patient-reported outcomes
Number of people analysed: 132
Follow up at time of assessment: 4 years
There was a statistically significant increase in WOMAC score from 75.3 ± 2.3 before surgery to 87.3 ± 6.6 at 4-year follow up (p<0.001).
At the 4-year follow up, there were no differences in WOMAC score in the following subgroups: age (younger than 40 years vs. 40 years or older), BMI (less than 25 vs. 25 or greater), gender (male vs. female), previous anterior cruciate ligament surgery (yes vs. no), and extra cartilage lesion (present vs. absent).
Key safety findings
Revisions and reoperations
Number of people analysed: 132
Follow up at time of assessment: 4 years
A total of 3 people were revised during the study, leading to an overall 4-year implant survival of 97.7%.
One person was revised to UniCAP (from HemiCAP), 1 person was revised to unicompartmental knee arthroplasty, and 1 person was revised to total knee arthroplasty.
A total of 16 reoperations were performed, leading to an overall reoperation rate of 12.1%.
Complications
No complications were reported.
Study 5 Çepni Ş (2020)
Study details
Study type | Single centre, retrospective cohort study |
Country | Turkey |
Recruitment period | 2014 to 2017 |
Study population and number | n=118 People with focal full-thickness knee cartilage lesions. |
Age & sex | Mean 56.3 years; 81.4% female |
Patient selection criteria | Inclusion criteria: people aged 40–65 years with a focal chondral lesion or defect of the medial or lateral femoral cartilage. Only lesions of ICRS grades 3 or 4 were included, with a maximum lesion area of 4 cm2 or less. Exclusion criteria: people with BMI 35 kg/m2 or more, generalised degenerative arthritis, chronic malalignment of the knee, ligamentous instability, symptomatic meniscal tear or total meniscectomy, kissing lesion on the tibia, metal allergies, inflammatory joint diseases, history of local and/or systemic corticosteroids, or the use of immunomodulating agents. |
Technique | Implant: HemiCAP (61%) or BioPoly (38%). Technique: under either general or spinal anaesthesia, diagnostic arthroscopy was performed to assess the defect. The diameter was measured, a guide wire inserted, and the defect was reamed. The implant was placed and recessed 1 mm deeper than surrounding cartilage. After surgery care: knee exercises on the 1st postoperative day, followed by gradually increasing weight-bearing. Full range of motion and weight bearing was permitted in the 6th week after surgery. |
Follow-up | Mean 4.7 years |
Conflict of interest/source of funding | Conflict of interest: The authors declared no conflict of interest. Source of funding: The authors declared that funding was not received. |
Analysis
Follow up issues: A total of 143 people were initially treated, of which 11 dropped out and 14 were lost to follow up.
Study design issues: This single centre retrospective cohort study assessed resurfacing for people with focal cartilage defects of the knee. Patient-reported outcomes included KOOS QoL, a visual analogue scale (VAS) score for pain, and the Tegner activity score. The scoring of KOOS QoL, VAS for pain, and Tegner have been described in Outcomes measures. Two-year follow up data was used for the analysis of KOOS QoL, VAS, and Tegner scores. Clinical and radiological outcomes were also assessed. These included the pain, loss of range of motion, allergies, infection, and wound problems.
Several significance tests were used depending on the type of data analysed. Univariate and multivariate regression were used to identify predictors for revision surgery, with all variables p<0.10 in univariate analysis entered into multivariate analysis. p<0.05 was considered statistically significant. There was no adjustment for multiple comparisons.
Study population issues: People who had BioPoly had a statistically significantly higher BMI than people who had HemiCAP implanted. People who had HemiCAP had statistically significantly higher preoperative scores on the VAS pain scale, indicating worse baseline pain. There were no other statistically significant differences in baseline characteristics between the groups.
Key efficacy findings
Patient-reported outcomes
Number of people analysed: 118
Follow up at time of assessment: 2 years
In both treatment groups, there were statistically significant improvements in VAS pain (p<0.001 for both groups), KOOS QoL (p<0.001 for both groups), and Tegner activity score (p<0.001 for both groups) from assessment before surgery to 2 years after surgery.
The following analyses compare outcomes between the treatment groups:
VAS Pain: At 2 years after surgery, people who had HemiCAP experienced a statistically significantly greater improvement in VAS pain than people who had BioPoly (p<0.001).
KOOS QoL: At 2 years after surgery, there was no difference in the change in KOOS QoL between people who had HemiCAP and people who had BioPoly (p=0.150).
Tegner Activity score: At 2 years after surgery, people who had HemiCAP experienced a statistically significantly greater improvement in Tegner Activity score than people who had BioPoly (p<0.001).
Outcome | Total (N=118) | BioPoly (n=45) | HemiCAP (n=73) | p-value |
VAS pain | ||||
Before surgery | <0.001 | |||
Mean ± SD | 6.63 ± 0.90 | 5.89 ± 0.53 | 7.08 ± 0.77 | |
2 years after surgery | 0.134 | |||
Mean ± SD | 1.16 ± 0.70 | 1.29 ± 0.78 | 1.09 ± 0.64 | |
Difference (before vs. after) | <0.001 | |||
Mean ± SD | -5.46 ± 1.22 | -4.60 ± 0.91 | -5.99 ± 1.07 | |
KOOS QoL Score | ||||
Before surgery | 0.828 | |||
Mean ± SD | 54.97 ± 3.48 | 54.83 ± 3.68 | 55.05 ± 3.37 | |
2 years after surgery | 0.052 | |||
Mean ± SD | 86.45 ± 7.26 | 84.79 ± 8.58 | 87.47 ± 6.16 | |
Difference (before vs. after) | 0.150 | |||
Mean ± SD | 31.48 ± 8.14 | 29.96 ± 9.16 | 32.41 ± 7.35 | |
Tegner Activity Score | ||||
Before surgery | 0.062 | |||
Mean ± SD | 1.07 ± 0.73 | 1.22 ± 0.77 | 0.97 ± 0.69 | |
2 years after surgery | <0.001 | |||
Mean ± SD | 2.96 ± 0.74 | 2.44 ± 0.66 | 3.27 ± 0.61 | |
Difference (before vs. after) | <0.001 | |||
Mean ± SD | 1.89 ± 0.99 | 1.22 ± 0.73 | 1.30 ± 0.89 |
Abbreviations: KOOS, Knee injury and Osteoarthritis Outcome Score; QoL, quality of life; SD, standard deviation; VAS, visual analogue scale.
Key safety findings
Revisions
Number of people analysed: 118
Follow up at time of assessment: 4.7 years
A total of 13 people (11.0%) required revision surgery.
A numerically higher proportion of people who had BioPoly implanted required revisions (17.8%) than people who had HemiCAP implanted (6.8%; p=0.077).
In multivariate regression analysis, the BioPoly implant was a statistically significant risk factor for revision surgery:
Adjusted HR (95% CI): 6.90 (1.04 to 45.73; p=0.045)
Note: this analysis may be confounded because 3 of the revisions of the BioPoly implant were in people who experienced trauma leading to pain and loosening (see below).
Reasons for revision surgery were listed as follows:
BioPoly (2 revised to unicompartmental knee arthroplasty, 6 to total knee arthroplasty):
Progressive pain after trauma, implant loosening, chondral lesions in the patellofemoral compartment, n=2
Progressive pain after trauma, implant loosening, n=1
Progressive arthritis in the lateral compartment, n=2
Progressive arthritis in the medial compartment, n=2
Infection, n=1
HemiCAP (2 revised to unicompartmental knee arthroplasty, 3 to total knee arthroplasty):
Generalized arthritis involving all compartments, n=2
Progressive arthritis in the lateral compartment, n=1
Progressive pain, implant loosening, generalized chondrolysis in all compartments, n=1
Infection, n=1
Study 6 Nathwani D (2017)
Study details
Study type | Single arm, multicentre, before-and-after study* |
Country | UK |
Recruitment period | Not reported |
Study population and number | n=33 People with symptomatic femoral condyle lesions who had BioPoly implanted. |
Age & sex | Mean 42.7 years; sex not reported |
Patient selection criteria | Inclusion criteria: Age 21 years or older; symptomatic femoral condyle lesions classified as ICRS grade 2, 3, or 4; femoral condyle lesion size 3.1 cm2 or less circumscribed by normal or nearly normal (ICRS grade-0 or 1) cartilage with an overall depth 4 mm or less from the articulating surface; sufficient subchondral bone quality to support implant. Exclusion criteria: BMI 30 kg/m2 or more; generalized degenerative or autoimmune arthritis; gout; uncorrected chronic malalignment of the knee†; uncorrected ligamentous instability†; uncorrected mechanically symptomatic meniscal tear or total meniscectomy†; kissing lesion on tibia; 1 implant or more required to accommodate lesion; patient-reported allergy to titanium alloy, ultra-high molecular weight polyethylene, or hyaluronan/hyaluronic acid; use with opposing articulating tibial components; any concomitant painful or disabling disease of the spine, hips, or lower limbs that would interfere with evaluation of the affected knee; pregnant, prisoner, vulnerable population, unable to provide informed consent. †Could be included if corrected during surgery. |
Technique | Implant: BioPoly (100%) Technique: performed using the approved BioPoly surgical technique. The implantation site was prepared with use of a bone-sparing technique to establish the correct implant orientation and depth relative to surrounding anatomy. Once the implantation site was deemed appropriate, the BioPoly implant was press-fit into the site. After surgery care: immediately after surgery, active motion and weight-bearing as tolerated were permitted. At 1 to 3 weeks, and continuing through 4 to 7 weeks, weight-bearing as tolerated and unrestricted range of motion as tolerated were permitted. At 8 to 11 weeks – return to full activity as tolerated. |
Follow-up | Up to 2 years |
Conflict of interest/source of funding | Conflict of interest: three authors reported receiving personal fees from BioPoly LLC, the manufacturers of the BioPoly implant. Source of funding: the study was funded by BioPoly LLC. |
*This study was included in Elbardesy (2021).
Analysis
Follow up issues: This is an interim publication of a planned 5-year study. At the time of analysis, 24 people had completed the 6 month follow up, 22 had completed the 1 year follow up, and 12 had completed the 2 year follow up.
Study design issues: This multicentre, before-and-after study assessed the 2-year outcomes of BioPoly for people with focal cartilage defects of the femoral condyles. The primary outcome measures were KOOS overall and subscores, a VAS score for pain, the SF-36 physical component score, and the Tegner activity score.
Before vs. after surgery comparisons were performed using 2-sample, 1-tailed t-tests. p<0.025 was considered significant.
Study population issues: Most people had a history of knee surgery (75.8%).
Key efficacy findings
Patient-reported outcomes
Number of people analysed: 33
Follow up at time of assessment: up to 2 years
There were statistically significant improvements in KOOS overall (and component subscale scores), VAS pain, and the SF-36 physical component from before surgery to 6 months, 1 year, and 2 years after surgery (all p<0.025).
There was a statistically significant improvement in Tegner activity score at 2 years after surgery compared to before surgery (p<0.025). There were no statistically significant differences in Tegner activity score at 6 months or 1 year follow up.
Before surgery (N=33) | 6 months follow up (N=24) | 1 year follow up (N=22) | 2 years follow up (N=12) | |
KOOS overall | 44.9 ± 18.0* | 67.9 ± 15.9 | 67.3 ± 18.9 | 77.6 ± 16.6 |
Pain | 51.9 ± 20.4* | 76.9 ± 14.9 | 75.6 ± 18.5 | 81.2 ± 16.2 |
Quality of life | 22.2 ± 18.4* | 50.8 ± 26.7 | 48.9 ± 26.7 | 68.2 ± 22.5 |
Sports | 30.0 ± 27.4* | 56.9 ± 22.1 | 56.4 ± 29.3 | 69.2 ± 25.8 |
Activities of daily living | 64.2 ± 24.3* | 84.9 ± 14.3 | 84.4 ± 16.1 | 89.0 ± 15.7 |
Symptoms | 56.2 ± 20.6* | 70.2 ± 18.0 | 71.3 ± 19.2 | 80.4 ± 12.9 |
VAS pain | 4.1 ± 2.5* | 2.4 ± 2.4 | 2.0 ± 2.0 | 1.4 ± 2.2 |
SF-36 physical | 42.3 ± 32.0* | 69.7 ± 28.2 | 71.0 ± 27.7 | 81.9 ± 30.8 |
Tegner activity | 2.5 ± 1.7** | 3.3 ± 1.4 | 3.1 ± 1.9 | 4.0 ± 1.9 |
*p<0.025 compared with scores at 6 months, 1 year, and 2 years.
**p<0.025 compared with score at 2 years.
Key safety findings
Revisions
Number of people analysed: 33
Follow up at time of assessment: Up to 2 years
One person was revised after the 2-year follow up because of the failure of osseointegration.
This person was managed with an alternative biological treatment.
Complications
Number of people analysed: 33
Follow up at time of assessment: Up to 2 years
No device-related adverse events were reported.
The following adverse events were reported as non-device-related:
Knee pain (arthralgia), n=9
Pain was localised to the contralateral compartment in 4 of these people.
Wound infection, n=1
Stiffness, n=1
Swelling, n=2
Crepitation, n=3
Loose cartilage body, n=1
Identified in the operatively treated knee 4 months postoperatively. Necessitated arthroscopic surgery but was not related to the implant. The implant was assessed and was deemed to be functioning and well fixed.
Study 7 Ryd L (2021)
Study details
Study type | Consecutive case series |
Country | Not reported |
Recruitment period | 2012 to 2020 |
Study population and number | n=612 knees, 682 implants People who had Episealer implanted for focal cartilage defects of the knee |
Age & sex | Mean 48.6 years; sex not reported |
Patient selection criteria | All post-market surveillance records from all people in the manufacturer-held database without any exclusions from December 2012 (first implant) to June 2020. |
Technique | Implant: Episealer (100%) Technique: Approved Episealer surgical approach. A 3-dimensional magnetic resonance imaging (MRI) scan was taken of the affected knee. The implant was then custom-made to fit the defect. Implantation was performed using a person-specific instrumentation and drill guide which allowed precise placement and fine-tuning of the implant recession depth. |
Follow-up | Up to 7 years |
Conflict of interest/source of funding | Conflict of interest: 2 authors are paid consultants to Episurf Medical AB, the manufacturer of Episealer. The other author is an employee of Episurf Medical AB. Source of funding: Not reported, likely company funded. |
Analysis
Follow up issues: One patient operated upon in 2018 died from unrelated causes in 2019 but was included in the analysis.
Study design issues: This multicentre, consecutive case series analysed the rate of revisions of the Episealer implant for people with focal condyle or trochlear cartilage defects. The outcome of this analysis was revision rates.
Implant revision rate was calculated using a Kaplan-Meier analysis. To calculate the cumulative revision rate, the number of knees 'at risk' of revision at the beginning of each year of follow up was determined. The number of revisions during each year of follow up was registered and divided by the numbers 'at risk' during that period and expressed as a percentage. These yearly percentages were added together to give a cumulative revision rate.
Key safety findings
Revisions
Number of people analysed: 612 knees
Follow up at time of assessment: up to 7 years
A total of 14 of the 612 operated knees were revised, giving a crude revision rate of 2.3%.
Using Kaplan-Meier analysis, implant survivorship at 7 years of follow up was 95.95%.
[Note: in the publication, it is stated that 'Kaplan-Meier survivorship at 6 years of follow-up was 96.3%'. It is unclear why this figure is different to the data provided in the table.]
Time (months) | Number at risk | Cumulative number revised | Implant survival | ||
Estimate, % | 95% confidence lower bound | 95% confidence upper bound | |||
0 | 612 | - | - | - | - |
12 | 478 | 4 | 99.25 | 98.02 | 99.72 |
24 | 327 | 7 | 98.57 | 97.01 | 99.32 |
36 | 191 | 14 | 95.95 | 93.08 | 97.64 |
48 | 97 | 14 | 95.95 | 93.08 | 97.64 |
60 | 42 | 14 | 95.95 | 93.08 | 97.64 |
72 | 15 | 14 | 95.95 | 93.08 | 97.64 |
84 | 4 | 14 | 95.95 | 93.08 | 97.64 |
Reasons for revision (some people had multiple reasons):
Pain, n=6
Disease progression, n=2
Multiple lesions, n=2
Implant was too small, n=2
Trauma after surgery, n=1
Metal allergy, n=1
Borderline indication, n=1
Tibial cartilage wear, n=1
Infection, n=1
Unknown, n=1
High tibial osteotomy failed, n=1
Study 8 Holz J (2021)
Study details
Study type | Single arm, multicentre, before-and-after study* |
Country | Belgium, Denmark, Germany, The Netherlands, Sweden, UK |
Recruitment period | 2013 to 2017 |
Study population and number | n=75 People who had Episealer for symptomatic cartilage defects in the knee |
Age & sex | Mean 48 years; 59% female |
Patient selection criteria | Inclusion criteria (selected): 18 years of age or older; focal femoral chondral or osteochondral lesions: ICRS III–IVb; symptoms of pain and disability; failed conservative treatment. Exclusion criteria (selected): younger than 35 years of age or older than 70 years; bone on bone disease; multifocal chondral defects; severe chondral lesion (ICRS III–IV) on opposing surface; systemic and/or inflammatory joint disease; inflammatory arthritis or radiographic osteoarthritis; joint instability or malalignment that is not correctable at the time of treatment. |
Technique | Implant: Episealer (100%). A total of 60 people had medial condyle lesions, 5 had lateral condyle lesions, and 10 had trochlear lesions. Technique: Approved Episealer surgical approach. A 3-dimensional MRI scan was taken of the affected knee. The implant was then custom-made to fit the defect. Implantation was performed using a person-specific instrumentation and drill guide which allowed precise placement and fine-tuning of the implant recession depth. After surgery care: Full unrestricted motion was allowed from the outset. Protected touch weight bearing for 2 weeks followed by gradual progression to full weight bearing over the subsequent 2 weeks. Cycling and strength work could commence at 6 weeks building up proprioception and core control over a 6-month period before allowing return to activities. |
Follow-up | 2 years |
Conflict of interest/source of funding | Conflict of interest: 4 authors are consultants for Episurf Medical AB, the manufacturer of Episealer. Source of funding: The authors report that no additional funding was received for this study. |
*The people in this study were likely also included in Ryd (2021).
Analysis
Follow up issues: Of 80 people who had Episealer, 1 declined to participate in the final analysis. A further 4 people were excluded from the final analysis: 2 because of failure to complete outcome measures at set points, and 2 because they were revised during the study period.
Study design issues: This multicentre, before-and-after study evaluated the 2-year outcomes of a custom-made implant for people with focal chondral defects of the knee. Outcome measures included overall KOOS and component subscale scores, and VAS for pain. KOOS was evaluated against a minimal clinically important difference of 10. This is recommended by the publisher of the KOOS, however, they note that 'Recent publications highlight that there is probably no such thing as one MIC for KOOS'. Outcomes were assessed at 3 months, 1 year, and 2 years.
Each clinical outcome score at different time points was compared against before surgery values, using 1‑sample, 2-tailed paired t-tests. p<0.05 was considered statistically significant. Linear mixed-effects models were used to analyse the progression of outcome scores over the study. Categorical data were analysed against other variables using Chi2 test, and significance of variation in clinically important difference using 2-tailed Fisher's exact test. There was no adjustment for multiple comparisons.
Study population issues: 48 people had previous treatment for their cartilage lesions. Thirty-one of these people had microfracture.
Key efficacy findings
Patient-reported outcomes
Number of people analysed: 75
Follow up at time of assessment: up to 2 years
There were statistically significant improvements in all outcome measures from before surgery to 3-month, 1-year, and 2-year follow up (p<0.0001 to p=0.024).
There were no statistically significant differences in outcome score between groups according to implant type, lesion size, or previous repair surgery (p>0.05).
Clinical outcome | Time (months) | Mean change from before surgery | SD | 95% CI | p-value at each time point | p-value over time |
KOOS-Pain | 3 | 16.73 | 2.77 | 11.20 to 22.25 | <0.0001 | |
12 | 22.92 | 3.46 | 16.02 to 29.82 | <0.0001 | ||
24 | 26.28 | 3.22 | 19.86 to 32.70 | <0.0001 | 0.002 | |
KOOS-Symptoms | 3 | 8.05 | 2.79 | 2.48 to 13.62 | 0.005 | |
12 | 17.58 | 2.95 | 11.70 to 23.46 | <0.0001 | ||
24 | 18.30 | 2.81 | 12.70 to 23.90 | <0.0001 | <0.0001 | |
KOOS-ADL | 3 | 16.42 | 2.85 | 10.73 to 22.10 | <0.0001 | |
12 | 24.75 | 3.28 | 18.20 to 31.29 | <0.0001 | ||
24 | 22.75 | 3.18 | 16.42 to 29.09 | <0.0001 | <0.0001 | |
KOOS-Sport | 3 | 9.47 | 4.1 | 1.28 to 17.66 | 0.024 | |
12 | 23.06 | 4.53 | 14.02 to 32.09 | <0.0001 | ||
24 | 25.27 | 4.13 | 17.04 to 33.50 | <0.0001 | 0.002 | |
KOOS-QoL | 3 | 14.60 | 3.17 | 8.27 to 20.92 | <0.0001 | |
12 | 24.05 | 3.76 | 16.54 to 31.55 | <0.0001 | ||
24 | 25.26 | 3.57 | 18.14 to 32.37 | <0.0001 | <0.0001 | |
VAS pain | 3 | 26.31 | 3.22 | 19.89 to 32.73 | <0.0001 | |
12 | 27.12 | 4.29 | 18.57 to 35.67 | <0.0001 | ||
24 | 30.22 | 3.95 | 22.34 to 38.11 | <0.0001 | <0.0001 |
Abbreviations: ADL, activities of daily living; CI, confidence interval; KOOS, Knee injury and Osteoarthritis Outcome Score; SD, standard deviation; VAS, visual analogue scale.
Key safety findings
Revisions
Number of people analysed: 75
Follow up at time of assessment: 27 months
A total of 3 people were revised, 2 during the 24-month follow up, and 1 at 27 months. This resulted in an overall revision rate of 4%.
Reasons for revision were as follows:
Within the 24-month study period:
1 person had an atypical lesion with significant bone marrow oedema condyle before surgery. Did not improve following implantation and re-presented with increased pain at 15 months. The implant was revised to a unicompartmental arthroplasty.
1 person had cysts persisting from previous OATS plugs. Symptoms did not improve, and the implant was revised at 19 months to bone grafting and coverage with a chondrogide membrane.
At 27 months follow up:
1 person reported severe pain. Implant was found to be loose, and cultures suggested infection. Revised to unicompartmental arthroplasty.
Complications
Number of people analysed: 75
Follow up at time of assessment: 27 months
Complications are poorly described. The publication states:
'Two patients underwent arthroscopy for painful mechanical clicking and for debridement of scar tissue, with both improving, and one patient developed a [deep vein thrombosis].'
It is not clear whether the people having arthroscopy both had painful mechanical clicking and required debridement of scar tissue, or whether 1 person had painful mechanical clicking and the other required debridement of scar tissue.
It is also unclear whether the deep vein thrombosis occurred in a person who had arthroscopy or another person.
The cause, severity, or treatment of the deep vein thrombosis is not described.
How are you taking part in this consultation?
You will not be able to change how you comment later.
You must be signed in to answer questions