Clinical outcome after treatment of single and multiple cartilage defects by autologous matrix-induced chondrogenesis

Abstract

Purpose:

Characterized cartilage lesions have a distinct impact on postoperative clinical outcome, which is still being evaluated. The purpose of this study was to assess the postoperative clinical outcome of autologous matrix-induced chondrogenesis (AMIC) for characterized cartilage lesions.

Methods:

Fifteen patients with articular cartilage (AC) defects of the knee were included in the study. AC defects were characterized intraoperatively by International Cartilage Repair Society score. Grade III–IV AC lesions were treated with AMIC; grade I–II lesions were left untreated. Patients were divided into subgroups and clinically evaluated by subjective autologous matrix-induced chondrogenesis (IKDC) and Tegner scores at median follow-up of 4.5 years.

Results:

Twenty-eight AC defects were diagnosed (1.9/patient). Multiple subgroup had larger diagnosed (7 ± 2.3 cm², p = 0.022) and untreated (3.1 ± 2.3 cm², p = 0.012) lesion areas than the single subgroup. Partly treated subgroup had larger untreated defect areas (3.6±2.3 cm², p = 0.025) than the Treated subgroup. Average subjective IKDC values of total group and individual subgroups improved significantly at follow-up. More patients restored their previous activity levels (p = 0.026) and had higher incremental subjective IKDC scores (p = 0.014) in the single subgroup than the multiple subgroup. Diagnosed defect size negatively correlated to subjective IKDC incremental (r = −0.624, p = 0.023) and postoperative scores (r = −0.545, p = 0.054) in total group.

Conclusions:

AMIC can have a clinically relevant outcome for patients with single or multiple knee AC lesions; however, clinical outcome is superior in patients with a single defect per knee. Patients with single defects returned to previous physical activity levels significantly faster than patients with multiple defects. Diagnosed AC defect areas negatively correlate to clinical improvement at follow-up.

Introduction

Articular cartilage (AC) defects still remain a significant problem, especially for population with a high level of physical activity. Incidental finding of chondral lesions in 63% of knee arthroscopic surgeries, present a big challenge to orthopedic surgeons. ¹ Single localized defects have been previously treated with established surgical techniques such as microfracture (MF), osteochondral graft transplantation (OAT), autologous chondrocyte transplantation (ACI), and matrix-induced ACI (MACI). Curl et al., however, concluded that people who lead an active lifestyle in sports have an increased risk for acquiring larger and often multiple lesions, which require personalized treatment approaches and have a higher risk of early osteoarthritis development. In addition, a tendency toward more than one lesion (multiple complex lesions) was found during knee arthroscopic surgery in a multitude of cases.

The treatment method is usually based on patient’s clinical status and description of a particular lesion. ² Local and multiple AC defects are often approached differently. This depends on the adopted standards of care, surgeon experience and preference, and patient expectations to return to active lifestyle.

Microfracture has been a first choice treatment for less physically active patients, who are diagnosed with small (≤ 2 cm²) focal chondral or osteochondral defects. ^3,4 However, clinical improvement gradually deteriorates after 2 years, usually because of partial fibrocartilaginous repair tissue, incomplete fill up of the defect and intralesional osteophytes. ⁵ Additionally, poorer clinical outcome was reported after microfracture of multiple AC defects, when compared to patients with a single defect per knee. ⁶

Osteochondral autograft transplantation (OAT) has been successfully used for treating small- to medium-sized focal chondral and osteochondral defects of the knee. ⁷ However, accompanying donor site morbidity prevents its application for larger or multiple defects of the knee. Laboratory culture-based techniques such as ACI and MACI have been applied for large, full-thickness, symptomatic AC defects ^8,9 ACI, MACI, and OAT may have clinical benefits over microfracture, superior same level return rate and durability after >3 years. ¹⁰ Albeit shown effective in long-term clinical data, a widespread use of ACI and MACI procedures is limited by a need for cell manufacturing plant and the cost of the procedure. ¹¹

AMIC has been shown to be clinically safe and effective in isolated, larger (>2 cm²) chondral defect cases. ^2,12,13 To sustain early mechanical stability and cartilage regeneration, collagen membrane is fixed over the microfractured defect. ¹⁴ Initially, indicated defect size for AMIC was 1.5 cm², nevertheless, subsequent studies attempted enlarging a defect area that could be covered with bioactive membrane. ^14,15 Due to increasing number of cases where more than one lesion and thus greater total defect areas are diagnosed, a need for improving treatment of severe cases has become more important recently. Several studies have reported an inferior clinical outcome of greater size lesions. ^{16 –18} However, some studies did not show influence of defect size on clinical effect. ^19,20 Still, there is scarcity of data about the actual return to previous physical activity levels and clinical outcome after grade I–II defects are diagnosed and left untreated.

The purpose of this study was to evaluate clinical improvement after AMIC technique to treat symptomatic single or multiple complex AC lesions. In addition, untreated defect impact on clinical outcome was assessed.

Materials and methods

Patients

A retrospective study was carried out from July 2007 to September 2017 in the Hospital of Lithuanian University of Health Sciences in Kaunas, Lithuania. Inclusion criteria for patients were symptomatic knee articular AC defects and verification by magnetic resonance imaging (MRI) examination. Exclusion criteria were tibiofemoral malalignment, osteoarthritis of any joint, unstable knee and other systemic medical conditions. Clinical outcome scores were used for in-person evaluation. Subjective IKDC and Tegner scores were used before the operation and at a median follow-up of 4.5 years (range 1–10 years). Follow-up MRI screening was not conducted because the patients felt well, and our study was not aimed at quantifying MRI data. The study was approved by the hospital ethics committee; patients gave informed consent to participate in the study.

Overall, we treated 15 patients (10 male and 5 female), with the median age of 30.7 years (range 23–45 years). Patients were actively engaged in recreational sport activities at least five times per week. Preoperative level according to Tegner score was 5.93 ± 1.1. The right-to-left side ratio was 8:7. All defects were classified according to the International Cartilage Repair Society (ICRS) classification. Grade III–IV lesions accordingly to MRI investigations were treated using AMIC, grade I–II defects were left untreated (Figure 1).

Figure 1.

Arthroscopic images of osteochondral defect in the medial femoral condyle (a, b) and patellofemoral groove (c, d) taken during AMIC procedure. Microfracture (a, c) is performed initially, followed by a membrane application (c, d) to cover the defect. AMIC: autologous matrix-induced chondrogenesis.

Additionally, all patients (total group, n = 15) were divided into four subgroups:

patients with one diagnosed AC defect (Single subgroup, n = 8);

patients with two or three diagnosed AC defects (Multiple subgroup, n = 7);

patients with grade III–IV AC defects treated (Treated subgroup, n = 10); and

patients with grade III–IV AC defects treated and grade I–II AC defects untreated (Partly treated subgroup, n = 5).

Surgical procedure

All cases were performed by a single experienced surgeon. During the procedure, loose chondral flaps, degenerated cartilage and calcified layer were debrided down to the subchondral bone until the stable cartilage rim was reached. AC defect area was estimated with arthroscopic hook probe and calculated in square centimeter

AMIC was carried out by a mini arthrotomy for 12 patients and arthroscopically in 3 cases. Calcified chondral layer was removed with a burr and microfracture with awl was performed to penetrate the subchondral bone as previously described. ³ A commercially available collagen I/III membrane (Chondro-Gide, Geistlich Surgery, Wolhusen, Switzerland) was trimmed to be slightly undersized in relation to the defect to avoid displacement. Collagen membrane was implanted into the prepared defect with a porous surface facing downward to support the outflow of the bone marrow content. Membrane was fixed to the defect site with commercially available fibrin glue (Tisseel, Baxter, Westlake Village, California, USA). Knee flexion and extension were checked intraoperatively to ensure stability of the graft.

Statistical analysis

Statistical analysis was performed using SPSS software (22.0 SPSS Inc. Chicago, Illinois, USA). Values of quantitative variables between two independent groups were compared using the nonparametric Mann–Whitney test. Two paired groups were compared using the nonparametric Wilcoxon test. The χ ² test for independence (homogeneity) or Fisher’s exact test (for small samples) was used for the analysis of two qualitative variables. Spearman’s rank correlation coefficient was applied to test the relationship between two quantitative variables. The differences and relationships between analyzed variables were taken as statistically significant if p < 0.05.

Results

Fifteen patients were followed up for 54.4 months (range 12–120 months). At the end of follow-up, a total of 28 knee AC defects (1.9 per patient; 2–9 cm²) were included. Twelve patients (80%) were diagnosed with medial femoral condyle lesions, eight patients (53%) with patella lesions, five patients (33%) with lateral femoral condyle lesions, and three patients (20%) with patellofemoral groove lesions. Mean size of grade I–IV lesions was 5.3 ± 2.3 cm². Grade III–IV defects (n = 16.57%), treated area was 3.7 ± 1.1 cm², which was significantly smaller compared to diagnosed defect (p < 0.05). Eleven patients had medial (seven cases) or lateral (three cases) partial meniscectomies and anterior cruciate ligament reconstruction (one case) performed during the same procedure (Table 1).

Table 1.

Patient characteristics in different subgroups.^a

	Patients	Average age (years)	Male:female	Defect size	MFC	LFC	P	PFG
Single	8	29 (6.48)	6:2	3.1 (2.3)	6 (75%)	1 (12.5%)	1 (12.5%)	0
Multiple	7	32.7 (7.8)	4:3	2.3 (1.1)	6 (30%)	4 (20%)	7 (35%)	3 (15%)
Treated	10	29 (6.28)	8:2	2.9 (2.1)	8 (53%)	3 (20%)	3 (20%)	1 (6%)
Partly treated	5	34 (7.98)	2:3	2.4 (1.21)	3 (25%)	2 (16%)	5 (42%)	2 (16%)

^aValues are represented as mean (standard deviation). No significant differences were found among subgroups as assessed by nonparametric Mann–Whitney test. MFC, Medial femoral condyle; LFC, Lateral femoral condyle; P, patela; PFG, patelofemoral groove.

We analyzed lesion characteristics of individual subgroups. There was no difference between subgroups sex, age, and defect sizes. Total defect area which was diagnosed in the multiple subgroup (2.9 defect per knee) was larger than the single subgroup lesion area (p = 0.022). In addition, untreated grade I–II lesions were also found to be significantly larger in the multiple subgroup, when compared to the single subgroup (p = 0.012).

Total defect area was comparable in the treated subgroup and the partly treated subgroup (p = 0.139). However, a smaller defect area in the latter group received treatment for grade III–IV lesions, thus leaving a greater part untreated (p = 0.025), when compared to the treated subgroup (Table 2).

Table 2.

Defect area characteristics in all patient subgroups (cm²).

	Single subgroup	Multiple subgroup	p Valueb	Treated subgroup	Partly treated subgroup	p Valuec
Total defect area diagnosed	4.0 (1.31)	7 (2.3)	0.022	4.6 (2.1)	6.6 (2.3)	n.s.
Total defect area treated	4.0 (1.31)	3.3 (0.82)	n.s.	4.6 (2.1)	3.0 (0.0)	0.03
Total defect area left untreated	0	3.1 (2.3)	0.012	0	3.6 (2.3)	0.025

^aValues are represented as mean (standard deviation).

^bDefect areas that were diagnosed and left untreated were significantly larger in the multiple subgroup compared to the single subgroup.

^cDefect area that was left untreated was significantly larger in the partly treated subgroup compared to the treated subgroup.

Clinical outcome

We observed a significant increase of mean subjective IKDC score in all patients at follow-up. The rate of return to the same level of sporting activities was reported in 73% of studied patients, as assessed by the Tegner score.

Additional sensitivity analysis of individual subgroups confirmed consistency of subjective IKDC score improvement at follow-up. Substantially superior increment of subjective IKDC score (p = 0.014) was reported in the single subgroup when compared to the multiple subgroup (Figure 2).

Figure 2.

IKDC score median improvements from preoperative score to IKDC score in single and multiple subgroups at the follow-up. IKDC score improvement in the single subgroup was more significant than improvement in the multiple subgroup at the follow-up (p = 0.014). IKDC: International Knee Documentation System.

Inferiority of clinical outcome in the multiple subgroup was further supported by a tendency toward lower subjective IKDC score postoperatively (p = 0.1). In addition, patients in the multiple subgroup reported significantly lower activity levels than patients in the single subgroup at follow-up (43% return rate; p = 0.026). Contrary to that, every one of eight patients in the single subgroup have reported a complete return to the same level of sporting activities, as was evident by the Tegner score (Figure 3).

Figure 3.

Return rate to the activity levels before injury as assessed by the Tegner score in single and multiple subgroups at follow-up. Patients from the multiple subgroup reported significantly lower return to the previous activity levels at follow-up (43% return rate), when compared to all patients who returned in the single subgroup (Fisher’s exact test p = 0.026).

Despite significantly improved clinical outcomes in treated and partly treated subgroups, we did not find a difference between improvement of clinical outcome among these subgroups, according to the subjective IKDC score (p = 0.56). However, a clear tendency for superior rate of return to sporting activities was noted for patients who had all their defects treated (Table 3).

Table 3.

Clinical outcomes for patients in subgroups.

Group/subgroup	IKDC score preoperative	IKDC score postoperative	IKDC score improvement	Returned	Did not return
All patients	55.5 (5.3)	86.5 (6.4)	31 (5.4)	11 (73%)	4 (27%)
Single	55.2 (6.0)	89.7 (4.1)	34.4 (4.7)b	8 (100%)c	0
Mutliple	55.8 (4.9)	82.8 (6.9)	27 (2.6)b	3 (42.9%)c	4 (57.1%)
Treated	54.6 (5.4)	87.1 (6.3)	32.4 (5.7)	8 (80%)	2 (20%)
Partly treated	57.5 (5.2)	85.2 (7.5)	27.7 (2.8)	3 (60%)	2 (40%)

IKDC: International Knee Documentation Committee.

^aValues are represented as mean (standard deviation). Significant clinical improvement, as expressed by IKDC score was evident in every individual group, exact Wilcoxon test, p < 0.001.

^b Single subgroup versus multiple subgroup, Mann–Whitney’s exact test, p = 0.014.

^c Single subgroup versus multiple subgroup, Fisher’s exact test, p = 0.026.

Larger area of diagnosed defect (which comprised 57% grade III–IV lesions) was associated with inferior clinical outcome and negatively correlated to the subjective IKDC score at follow-up (r = −0.624, p = 0.023; Figure 4).

Figure 4.

IKDC median score difference correlation to total diagnosed defect area. Patients with larger diagnosed defects exhibited inferior clinical improvement at follow-up (r = −0.624; R ² = 0.45; p = 0.023).

Arthroscopic approach did not show superior outcome when compared to mini-arthrotomy. Concomitant surgeries did not influence clinical outcome. No postoperative complications were observed during follow-up period.

Discussion

AC lesions are becoming increasingly prevalent, as a result of improved diagnosis and increased physical activity. In addition to single osteochondral lesions, multiple defects are being increasingly diagnosed as well. In a study including 565 patients, two AC lesions were found in 25% and three or more in 12% of patients. ²¹

The aim of this study was to evaluate clinical outcome of treated grade III–IV and untreated grade I–II AC defects after AMIC.

In the present study, we used subjective IKDC and Tegner scores to evaluate symptoms and return to the pre-injury activity level. IKDC questionnaire has been validated as a score by ICRS and is often used in cartilage repair studies after AMIC, ^13,22,23 OAT, ^24,25 microfracture ^7,26 treatments. Tegner score has been demonstrated as a functional method to evaluate patients, engaged in recreational sport activities at least five times per week, and their ability to return to the pre-injury activity level. ^{27 –29}

Study patients were evaluated being at recreational sports level, thus representing an active cohort of population. Higher preoperative IKDC scores, similarly documented in the studies by Buda et al., ³⁰ Kusano et al., ¹³ Panni et al., ²² and Pascarella et al., ²³ might be indicative of a less symptomatic patient or represent a higher tolerance toward disturbed daily activities. Due to scarcity of data containing lesion characteristics in these studies, smaller diagnosed defect areas or lower grade cartilage lesions might also be the reason for higher preoperative subjective IKDC scores in our study. In addition, different clinical valuation scores used in studies makes adequate direct comparison among them more ambiguous.

With 4.5 years of median follow-up, the study included 15 patients that exhibited significant clinical improvement. A 73% return rate to previous activity levels after AMIC was superior to previously reported outcomes after microfracture, ³¹ however inferior to ACI, with 87% of their patients able to return to physical activity at 5 years after surgery. ³² Chalmers et al. reported a faster return to sports following microfracture, when compared to ACI and OAT. ³³ A decreased physical activity in our study may be explained by a deliberate action taken by a patient to prevent a failure of initially successful treatment.

Previous studies have reported increased clinical scores after AMIC treatment at different follow-up periods. ^34,13 Kusano et al. ¹³ reported 38 patients treated with AMIC at 29 months of follow-up. Significant clinical improvements in IKDC, Tegner, Lysholm, and VAS scores were found. Dhollander et al. ³⁴ reviewed the results of 10 patients treated with AMIC for patellofemoral cartilage lesions. They concluded that procedure is safe and clinically effective for patellofemoral defects at 2-year follow-up. However, in a study published by Gille et al., ³⁵ scores declined significantly in a group of patients with patella defects at 36 months postoperatively. Our preliminary data revealed that patients with patellar lesions reported a lower Tegner score at follow-up. Inferior clinical outcome and the return to sports activities in Multiple and Partly treated subgroups might have also been influenced by predominant patellar region lesions. In addition, according to studies by Dhollander et al. ³⁴ and Kusano et al., ³ MRI data could not fully confirm an improved clinical outcome after AMIC treatment of lesions in the patellofemoral groove. True outcome of treating patellofemoral groove osteochondral defects with AMIC should be further elucidated in future studies with a longer follow-up period.

Sensitivity analysis revealed a significant clinical improvement among respective subgroups. However, when single and multiple subgroups were analyzed, it was clearly evident that clinical improvement was inferior in the latter subgroup, as represented by IKDC and Tegner scores. This may be partially explained by a significantly larger total defect area diagnosed in the multiple subgroup compared to the single subgroup, 6.9 ± 2.1 cm² and 4.3 ± 2.4 cm², respectively (p = 0.02) and untreated grade I–II defects, which comprised 66% of all diagnosed lesion area in the multiple subgroup. Patients from the multiple subgroup might have deliberately taken action to sustain lower physical activity to improve full recovery. In addition, substantially larger lesion areas were left untreated in multiple and partly treated subgroups compared to its respective single and treated subgroups.

A similar area size of grade III–IV defects, which were subsequently treated with AMIC were diagnosed in single and multiple subgroups (4.0 ± 1.3 and 3.3 ± 0.82, respectively), thus a significantly larger untreated defect area of grade I–II lesions in the latter group might have been responsible for inferior clinical outcome at follow-up. It is known that injured cartilage itself does not cause any pain due to lack of pain receptors in the tissue, instead a deteriorating cartilage is one of the causes of discomfort in the knee. Partial-thickness defect tend to become a full-thickness over time, which might have been the case for patients with initially low grade, asymptomatic lesions.

The latter was further supported by an even smaller total area of highly debilitating grade III–IV lesions in the partly treated subgroup compared to the treated subgroup (p = 0.025), thus causing untreated grade I–II defects (62.5% of diagnosed lesions) to negatively influence IKDC and Tegner scores at final follow-up. Therefore, grade III–IV lesions might not be the only factor influencing clinical outcome and activity levels at follow-up; instead, a worse initial condition of developing defects are left untreated and have more potential for progression and deterioration over time.

There is paucity of data related to AMIC treatment comparisons of groups with different defect characteristics. Several studies have analyzed and published varying clinical outcomes after various treatments for single and multiple AC defects. ^6,13,34

Gobbi et al. ³⁶ treated 25 patients using bone marrow aspirate concentrate covered with a collagen-based membrane scaffold membrane and followed up for a minimum of 3 years. In addition to the analysis of significantly improved clinical score data, they studied subgroups based on the defect area, number, and localization. Medium-sized lesions had a better clinical outcome than larger lesions. However, there was no difference between patients with single or multiple chondral lesions. Additional information on defect characteristics of separate subgroups would shed more insight into the impact of lesion multitude and size.

Brix et al. ³⁷ reported the results of 53 patients after hyaluronan-based MACI. Studied cases included a simple lesions group and a complex group, consisting of patients with defects larger than 4 cm² or multiple defects. Despite a significant increase of all scores, they did not find a difference between simple and complex groups, after a mean follow up of 9.07 years (5–12 years).

Solheim et al. ⁶ reported clinical improvement for patients with single and multiple AC defects of the knee after microfracture treatment. Clinical outcome was superior in a group of patients with a single lesion to the group with multiple lesions, after 5-year median follow-up.

With respect to implementing new cartilage repair techniques, a demand for an early and convenient evaluation of cartilage quality is constantly increasing. Intraoperative assessment of cartilage regeneration level has been recently shown in an in vivo sheep model. ³⁸ Electromechanical measurements of cartilage obtained by a handheld arthroscopic device showed a correlation to traditional methods for cartilage repair quality determination. Similar technologies can help with early noninvasive diagnosis of cartilage quality before clear visual deterioration is visible. This can improve postoperative rehabilitation protocols and clinical outcomes for patients.

Limitations of the study were a small number of patients. It did not reveal more significant changes among different treatment groups. We did not use more clinical outcome scores for in-depth analysis of what specific knee motions were limited due to the cartilage defect. An important limitation of our study is a lack of interim results that would enable the analysis of longer follow-up. Another limitation is lack of reimaging, histological sampling of the fibrous repair cartilage. Lack of the repair tissue integration, as a result of micromotion, macromotion, and subsequently loosening, might have impaired clinical outcome.

Conclusions

AMIC can be an effective restorative technique at median follow-up of 4.5 years. Even though patients from single and multiple subgroups exhibit significantly better results at follow-up, clinical outcome was superior in patients with a single defect diagnosed. It was additionally supported by the negative correlation between total area of the diagnosed defect and subjective IKDC score incremental value. More effective diagnostics of early cartilage degeneration might improve expected clinical outcome after the treatment.