European Journal of Trauma and Emergency Surgery

, Volume 37, Issue 4, pp 397–403

Comparison of two different matrix-based autologous chondrocyte transplantation systems: 1 year follow-up results


  • S. Flohé
    • Department of Trauma and Hand SurgeryHeinrich Heine University Hospital Düsseldorf
    • Department of Trauma and Hand SurgeryHeinrich Heine University Hospital Düsseldorf
  • K. Ruße
    • Department of Arthroscopic Surgery and Sports TraumatologySankt Josef Orthopaedic and Rheumatology Centre Wuppertal
  • M. Wild
    • Department of Trauma and Hand SurgeryHeinrich Heine University Hospital Düsseldorf
  • J. Windolf
    • Department of Trauma and Hand SurgeryHeinrich Heine University Hospital Düsseldorf
  • M. Schulz
    • Department of SurgerySt. Johannes-Hospital Varel
Original Article

DOI: 10.1007/s00068-010-0057-2

Cite this article as:
Flohé, S., Betsch, M., Ruße, K. et al. Eur J Trauma Emerg Surg (2011) 37: 397. doi:10.1007/s00068-010-0057-2



The treatment of full-thickness cartilage defects of the knee is a common problem in orthopaedic surgery. Autologous chondrocyte transplantation (ACT) is one of the few reliable treatment options of cartilage defects with good long-term outcomes. The improvement of ACT led to the matrix-based ACT (MACT). The purpose of the study was to compare two different commercially available MACT systems.


Eleven patients with a cartilage defect of the knee were treated with the MACI® system and another nine patients with the CaRes® implant. The patients were prospectively followed and re-examined after one year.


One year after surgery all but one patient have significantly improved in their clinical outcome. Both types of MACT revealed similar results in terms of increase in level of activity, pain relief and knee function.


The study showed that MACT is a good therapeutic option for full-size defects of the femoral condyle. The MACI® implant seems to be easier to handle which is reflected by smaller incisions and a shorter operation time.


Cartilage defectAutologous chondrocyte transplantationArthroscopyChondrocytesMatrix-based autologous chondrocyte transplantation


Treatment of full-thickness cartilage defects of the articular surface of the knee represents a common problem in orthopaedic surgery. Different approaches and considerations exist regarding therapeutic possibilities, leading to numerous treatment options. Based on the idea that autologous chondrocytes would be the best solution for treatment of cartilage defects, in 1994 Brittberg et al. [1] introduced the autologous chondrocyte transplantation (ACT) in clinical practice. In the ACT procedure, cultured chondrocytes are injected beneath a tightly-sealed periosteal flap. By now ACT has developed into an accepted therapeutic option for greater cartilage lesions of the femoral condyle of the knee. ACT is the only possibility to resurface cartilage defects with hyaline-like cartilage besides the osteochondral cartilage transplantation (OCT), which revealed good hyaline cartilage survival but is limited in its application for greater defects due to the donor site morbidity. ACT and OCT are both claimed to be successful for the repair of full-thickness cartilage defects of the knee [2, 3]. In spite of the fact that histopathological studies showed only a 50% success rate in the development of hyaline-like cartilage after ACT, there do exist long-term results that demonstrate good clinical results 10 years after ACT [4]. However, despite the encouraging clinical results obtained, ACT carries limitations which are associated with the complexity of the surgical procedure and the biological activity of the periosteum. Besides technical problems and a time consuming procedure, graft hypertrophy of the periosteal flap may also cause problems after the classical ACT [5, 6]. Clinical symptoms that can be related to periosteal hypertrophy have been found in 10–26% of the cases in long-term follow-up studies [7]. The lack of a collagen matrix in a liquid suspension culture may also be a reason for limited chondrocyte differentiation. All these aspects led to the development of a matrix-based ACT (MACT). In MACT, isolated autologous chondrocytes are cultured on three-dimensional scaffolds which consist of different types of collagen. The development of MACT eased the surgical handling of the procedure resulting in a reduced time of surgery [5]. The first long-term studies comparing classical ACT and matrix-based autologous chondrocyte transplantation revealed similar results for both procedures [8]. As a systematic review recently pointed out, there are no studies that proved that ACT is more effective than other conventional techniques in treating chondral lesions of the knee [9]. In spite of this fact, the industry offers a growing number of different matrices with various chondrocyte culture techniques. The differences between ACT and MACT, as well as both procedures in comparison to conventional treatment of cartilage defects, have been analyzed in clinical trials. Matrices for chondrocytes differ in terms of physical properties (coated membranes or gels) as well as chemical composition (different types of collagen, hyaluronic acid). So far the differences of the various types of MACT have not yet been investigated in vivo. In the study, we compared two different types of matrix-based ACT. Therefore, we investigated the transplantation of immobilized cultured chondrocytes on a “Chondro-Gide” membrane (Geistlich Biomaterials, Wollhausen, Switzerland) which consists of highly purified porcine collagen I/III (MACI®, Genzyme, formerly Verigen) with a three-dimensional collagen type-I gel seeded with autologous chondrocytes directly after isolation (CaReS®, Ars Arthro).

Materials and methods

Study protocol

Twenty patients were prospectively followed after matrix-based ACT for full-thickness cartilage defects of the femoral condyle. Between March 2007 and September 2008, nine patients were treated with the CaRes® implant, and between March 2008 and May 2009, 11 patients were treated with the MACI® membrane in a consecutive series. The 20 participants were prospectively followed and the below mentioned scores were assessed pre-operatively and at a 1 year follow-up visit. All patients fulfilled the following criteria for ACT of the femoral condyle according to the joined advisory board of the German Trauma Society (DGU) and the Society of Orthopaedic Surgery (DGOOC) [10]:
  1. 1.

    Defects of the femoral condyle

  2. 2.

    Maximally two independent defects not located on corresponding articular regions (no “kissing lesions”)

  3. 3.

    Biomechanical stable knee

  4. 4.

    Intact meniscus (maximal resection 30%)

  5. 5.

    Absence of malalignment (varus or valgus abnormalities below 5°)

  6. 6.

    Body mass index below 30


International Knee Documentation Committee (IKDC) evaluation [11], Tegner–Lysholm activity score [12], degree of pain (visual analogue score, VAS), ICRS score [13] and SF-36 were obtained pre-operatively and 12 months after surgery. All patients gave their written consent for the evaluation of their data in this comparative study.

Surgical procedure

All surgical procedures were performed by a single surgeon (M.S) in a single centre. The MACT procedure was carried out in two stages. MACT began with the arthroscopic harvesting of cartilage tissue for chondrocyte isolation. Cartilage tissue was obtained from the lateral rim of the supracondylar region of the femoral condyle in a non-weight bearing region. A 5 mm × 5 mm full-thickness cartilage fragment was removed for chondrocyte isolation and culture. Cells were cultured/isolated according to the respective companies procedures in serum collected from the patient at the time of surgery. Principal differences between CaRes® and MACI® exist in the fact that CaRes® does not establish a suspension culture of isolated chondrocytes but rather seeds and cultures the primarily isolated donor cells in a three-dimensional rat collagen I gel matrix for 2 weeks. In contrast, for MACI® a suspension culture of the isolated chondrocytes is established for 4 weeks in order to expand the cell number to 15–20 × 106 cells. At the end of this period, the cells are transferred on a membrane, which consists of bovine collagen I/III. The varying cell culture approaches result in different cell numbers present in the transplant (2.15 × 105 cells/cm2 in CaRes® vs. 1 × 106/cells/cm2 in MACI®).

For chondrocyte matrix implantation the defect site was accessed via a medial or lateral parapatellar incision and arthrotomy in a tourniquet controlled field. The defects were thoroughly curetted to remove reactive fibrous tissue and to define stable defect borders. The chondrocyte scaffold was then shaped to match the defect size. After the scaffold was correctly shaped, the defect was filled with Tisseel fibrin sealant (Baxter, Vienna, Austria) and the scaffold was press-fitted into the defect. For the MACI® implant in some cases additional fixation was achieved by 4/0 PDS sutures to the intact cartilage rim.

Rehabilitation program

Post-operatively the knee was immobilized for 48 h. Beginning on day 3 after surgery continuous passive motion (CPM) with a range of movement of 0/0/60° was performed for 8 h a day during the first 3 weeks. Partial weight bearing with 15 kg was allowed. After 3 weeks, the range of motion was increased to 0/0/90°. After 6 weeks, 30 kg weight bearing was allowed. After 12 weeks, patients returned to full weight bearing, normal daily and sport activities without extraordinary burden for the knee joint (e.g. cycling, swimming).


All values are expressed as mean ± standard deviation. Statistical evaluation was done between both treatment groups and in relation to pre-operative scores using paired and unpaired Student’s t test. The level of significance was set at p < 0.05.


A total of 20 patients were treated with MACT in the study period between March 2007 and May 2009. Underlying diseases consisted of idiopathic osteochondrosis dissecans (n = 8) and cartilage defects on the basis of a trauma (n = 12). All treated defects were cartilage lesions grade III or IV according to the ICRS classification [13]. Predominantly the defects were located on the medial femoral condyle. Both treatment groups did not differ in terms of age, BMI, sex distribution, defect size and grade, type of injury or location. Patients’ characteristics of both groups are summarized in Table 1.
Table 1

Patients’ data




Number of patients



Age (years) (mean ± SD)

37.3 ± 6.3

33.1 ± 9.1




BMI (mean ± SD)

25 ± 2.6

24 ± 5.4

Defect size (cm2) (mean ± SD)

7.1 ± 2.1

7.1 ± 1.7

Grade of lesion (mean ± SD)

3.3 ± 0.5

3.7 ± 0.5

Medial/lateral femoral condyle



Traumatic/osteochondrosis dissecans



During the study only one significant complication occurred (infection with subsequent graft loss in the CaRes® group). Despite the graft loss the patient was included in the study and his outcome was considered in the results.

One year after the MACT all but one patient have improved significantly in terms of pain score (pre-operative 7.1 ± 1.7 to 1 year post-operative 3.0 ± 2.4, p > 0.001 Student’s t test), level of activity measured by Tegner–Lysholm activity score (pre-operative 2.5 ± 0.8 to 1 year post-operative 5.0 ± 1.9, p > 0.001 Student’s t test) and ICRS score (pre-operative 41.7 ± 9.9 to 1 year post-operative 64.6 ± 11.6, p > 0.001 Student’s t test). In addition some aspects of general quality of life improved after MACT as measured by the SF-36 (categories “body function”, “body role” and “pain”).

Patients were treated with two different types of MACT that were applied during different time periods. All patients were prospectively followed. A comparative evaluation of the 1 year follow-up results can be performed as a cohortive study. The comparison between patients treated with ACT in the three-dimensional collagen I matrix (CaReS, group 1) and the chondrocytes seeded on a membrane, which consists of bovine collagen I/IIII (MACI, group 2) revealed the following results.

Both types of MACT revealed similar results in terms of increase in level of activity, relief from pain and improvement of knee function assessed by the IKDC. There were no statistically significant differences between group 1 (CaReS®) and group 2 (MACI®) in pre-operative or post-operative values of the VAS, Tegner–Lysholm score, subjective ICRS score or objective results of the IKDC score (Fig. 1a–d). The general evaluation of quality of life by the SF-36-score also did not reveal any significant differences between the two groups (Fig. 2). Return to work was the same in both groups with a mean of 3 months after the operation (14 weeks in group 1; 15 weeks in group 2). As a possible marker for the surgical handling of the implant, we analysed the duration of the surgery and the length of the skin incision as a marker for the exposure necessary to perform the operation. Time of surgery and length of skin incision were smaller in the MACI® (group 2), suggesting a more comfortable surgical handling of the membrane matrix at least in the hand of the treating surgeon of this study (Table 2).
Fig. 1

a The mean ± standard deviation of the pain score (VAS visual analogue scale) pre-operatively and 1 year after surgical procedure. In both treatment groups, the degree of pain significantly decreased 1 year after surgery (p < 0.001) without any differences between both treatment groups. b The mean ± standard deviation of the Tegner–Lysholm activity score pre-operatively and 1 year after surgical procedure. In both treatment groups the activity score significantly increased 1 year after surgery (p < 0.001) without any differences between both treatment groups. c The mean ± standard deviation of the ICRS scores pre-operatively and 1 year after surgical procedure. In both treatment groups, the ICRS scores significantly increased 1 year after surgery (p < 0.001) without any differences between both treatment groups. d Distribution of objective results of the IKDC evaluation score. The number of patients with normal knee function (A), nearly normal (B), abnormal (C) and severely abnormal (D) were not significantly different 1 year after surgery in both treatment groups
Fig. 2

The mean ± standard deviation of all seven categories of SF-36 scores pre-operatively and 1 year after surgical procedure. In both treatment groups, the SF-36 scores significantly increased in the categories body function, body role and pain 1 year after surgery (p < 0.001) without any differences between both treatment groups

Table 2

Length of incision and duration of surgery for MACI® and CaRes®




p value

Duration of surgery (min)

77.7 ± 16.7

65.1 ± 9.7


Length of incision (cm)

7.3 ± 0.6

6.0 ± 0.8


Summarizing, all but one patient with an infection, showed a benefit from the treatment. There may exist some technical advantages of a more rigid membrane scaffold in comparison to the gel matrix, but overall comparative trial did not reveal any clinically relevant differences between the two scaffolds.


The presented data show that MACT is a therapeutic option for full-size cartilage defects of the femoral condyle as is clearly illustrated by reduced pain levels and increased activity scores. These data are in accordance with numerous previous studies, which demonstrate the effects of ACT in terms of clinical results [1417]. The degree of subjective improvement in the present study is very similar to previous reports using the same scaffolds. Maus et al. [16] found an improvement of the subjective ICRS score of 25 points in knee defects treated with CaRes®, while we noticed a very similar improvement of 22 score points using the same scaffold. Correspondingly, Behrens et al. [17] reported an increase of the ICRS score of 22 points after MACI® treatment 5 years after surgery. The same improvement was observed in our study after 1 year. The clinical improvements with both scaffolds in our study are very similar compared to previously reported data in the literature. However, the study has some limitations that need to be discussed. First, the study design was not randomized but represents a comparison of two consecutive case series. However, the reason for changing the applied type of MACT in this cohortive study merely depended on financing aspects by the health insurance companies, so that no bias to either group can be assumed. The fact that all operative procedures were performed in a single institution by a single surgeon supports the comparability of the two groups in spite of a rather small sample size. We did not perform any second look arthroscopy and histological evaluation of the MACT. Therefore, the presented study cannot give any substantial information about the quality of the transplanted cartilage tissue. The good clinical results may indicate a high degree of hyaline or hyaline-like cartilage; however, this remains speculative. Finally, the 12 months follow-up may be too short to reveal any differences between the two procedures. Most controlled clinical studies have a 2 or 3 year follow-up [1820]. However, studies with longer follow-up (5 years) comparing microfracture and second-generation autologous chondrocyte implantation (Hyalograft C) revealed the same results after 2 years, but showed better results in terms of sport activity in the group treated with autologous chondrocyte transplantation in the 5 year follow-up examination [21]. This underlines the importance of longer follow-up, since regenerative cartilage tissue may deteriorate over time becoming only clinically apparent in longer follow-up examinations.

In terms of the clinical results we were not able to show any differences between the different types of MACT. However, concerning biological attributes, there are substantial differences between the two types of scaffolds. First, there exists a significant difference in the number of chondrocytes present in the transplant with five times more cells present in the MACI® membrane. In vitro data suggest that a certain amount of chondrocytes/cm2 (12–25 million cells/cm2) is necessary to increase matrix production and to create mechanical stability [22]. Both applied products use cell numbers lower than 12–25 million/cm2. Especially, the cell number in the CaRes® scaffold is about a 100 times lower. However, the cell number present in the MACI® group is, with 1 × 106 cells/cm2 , about ten times lower than the recommendation of the in vitro studies. There is a certain conflict between the available amount of cells under the aspect of limiting donor site morbidity as well as the danger of a dedifferentiation of chondrocytes during culture. Chondrocyte dedifferentiation is common in monolayer cultures, while three-dimensional matrices support phenotype preservation [23]. It has been demonstrated that chondrocytes preserve good cartilaginous quality and proliferation capacity after suspension culture [24]. In addition, histological studies of chondrocytes seeded to a MACI® membrane revealed histologically differentiated hyaline chondrocytes [25]. It could be speculated, that the direct cultivation in a three-dimensional scaffold of the CaRes® implant supports an even better differentiation of the chondrocytes and thereby compensates the lower cell number. On the other hand, the lower cell number could simply be sufficient. Clinically, we were not able to confirm any differences between the two scaffolds. Since quality of the transplanted cartilage may become relevant over time, the follow-up of 1 year may be too short to reveal problems with inadequate cartilage regeneration.

Both scaffolds were fixed with fibrin glue. The impact of fibrin glue on transplanted chondrocytes, however, remains controversial. Brittberg et al. [26] reported that chondrocytes do not migrate into the fibrin glue and he therefore concluded that fibrin glue is not a suitable scaffold for the treatment of cartilage defects. At least for some in vitro experiments, fibrin glue has been demonstrated to induce apoptotic cell death of cultured chondrocytes [27]. On the other hand, Malovits et al. [28] demonstrated a reliable attachment of the MACI membrane with fibrin glue. Furthermore, there are reports about fibrin as an addition to bioresorbable scaffolds. In vitro experiments with polylactic-co-glycolic acid (PLGA) as a chondrocyte carrier have shown that chondrocytes cultured in PLGA in combination with fibrin showed the same cell proliferation as PGLA scaffolds alone, and the fibrin/PGLA-scaffold showed macroscopically and histologically even more cartilage-like morphology [29].

The only differences observed between the two scaffolds refer to the physical property of the matrix. MACI® as a membrane appears to be easier to handle, which is reflected by smaller incisions and shorter operation time. The physical properties of the MACI®-membrane or other more rigid scaffolds even allow minimal invasive or arthroscopically-assisted implantation techniques, which have been reported for both the tibial [30] and also the femoral condyle [31].


In conclusion, the study demonstrates the therapeutic effect of MACT for cartilage defects of the femoral condyle without revealing any significant differences between the CaRes® and the MACI® implant in spite of the different philosophy of culture and scaffold. Both techniques revealed a significant improvement in knee function. However, one has to keep in mind that according to the most recent systemic study analysis, there is still no proven advantage of an autologous cartilage implantation over simple procedures such as microfracture [32].


The manuscript has been read and approved by all identified authors. The authors did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity.

Conflict of interest


Copyright information

© Urban & Vogel 2010