MRI evaluation of a new scaffold-based allogenic chondrocyte implantation for cartilage repair

European Journal of Radiology 75 (2010) 72–81

Contents lists available at ScienceDirect

European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad

MRI evaluation of a new scaffold-based allogenic chondrocyte implantation for cartilage repair A.A.M. Dhollander a,∗,1 , W.C.J. Huysse b,1,2 , P.C.M. Verdonk a,3 , K.L. Verstraete b,4 , R. Verdonk a,5 , G. Verbruggen c,6 , K.F. Almqvist a,7 a

Department of Orthopaedic Surgery and Traumatology, Ghent University Hospital, De Pintelaan 185, 1P5, B9000 Gent, Belgium Department of Radiology, Ghent University Hospital, De Pintelaan 185, -1K12 IB, B9000 Gent, Belgium c Laboratory of Connective Tissue Biology, Department of Rheumatology, Ghent University Hospital, De Pintelaan 185, Ghent, Belgium b

a r t i c l e

i n f o

Article history: Received 29 September 2008 Received in revised form 2 February 2009 Accepted 4 March 2009 Keywords: Cartilage Knee Allogenic Chondrocyte Alginate Magnetic resonance imaging

a b s t r a c t Aim: The present study was designed to evaluate the implantation of alginate beads containing human mature allogenic chondrocytes for the treatment of symptomatic cartilage defects of the knee. MRI was used for the morphological analysis of cartilage repair. The correlation between MRI findings and clinical outcome was also studied. Methods: A biodegradable, alginate-based biocompatible scaffold containing human mature allogenic chondrocytes was used for the treatment of symptomatic chondral and osteochondral lesions in the knee. Twenty-one patients were prospectively evaluated with use of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and the Visual Analogue Scale (VAS) for pain preoperatively and at 3, 6, 9 and 12 months of follow-up. Of the 21 patients, 12 had consented to follow the postoperative MRI evaluation protocol. MRI data were analyzed based on the original MOCART (Magnetic Resonance Observation of Cartilage Repair Tissue) and modified MOCART scoring system. The correlation between the clinical outcome and MRI findings was evaluated. Results: A statistically significant clinical improvement became apparent after 6 months and patients continued to improve during the 12 months of follow-up. One of the two MRI scoring systems that were used, showed a statistically significant deterioration of the repair tissue at 1 year of follow-up. Twelve months after the operation complete filling or hypertrophy was found in 41.6%. Bone-marrow edema and effusion were seen in 41.7% and 25% of the study patients, respectively. We did not find a consistent correlation between the MRI criteria and the clinical results. Discussion: The present study confirmed the primary role of MRI in the evaluation of cartilage repair. Two MOCART-based scoring systems were used in a longitudinal fashion and allowed a practical and morphological evaluation of the repair tissue. However, the correlation between clinical outcome and MRI findings was poor. Further validation of these scoring systems is mandatory. The promising shortterm clinical outcome of the allogenic chondrocytes/alginate beads implantation was not confirmed by the short-term MRI findings. © 2009 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

∗ Corresponding author. Tel.: +32 0498 50 11 65. E-mail addresses: [email protected] (A.A.M. Dhollander), [email protected] (W.C.J. Huysse), [email protected] (P.C.M. Verdonk), [email protected] (K.L. Verstraete), [email protected] (R. Verdonk), [email protected] (G. Verbruggen), [email protected] (K.F. Almqvist). 1 Both authors have contributed equally to this article and share first authorship. 2 Tel.: +32 09 332 6685; fax: +32 09 332 6145. 3 Tel.: +32 09 332 4708. 4 Tel.: +32 09 332 2912. 5 Tel.: +32 09 332 2227. 6 Tel.: +32 09 332 2250. 7 Tel.: +32 09 332 22 24; fax: +32 09 332 49 75. 0720-048X/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2009.03.056

The interest in cartilage repair is growing worldwide, with particular focus on tissue engineering and cell-based therapies. This has resulted in an increasing need for accurate, reproducible and ideally, noninvasive methods for the assessment of cartilage lesions and cartilage repair. Magnetic resonance imaging (MRI) is the standard imaging method for evaluation of cartilage defects and repair tissue in the pre- and postoperative period [1,2]. The most commonly used techniques are intermediate-weighted, fast spin-echo (FSE) and three-dimensional (3D), fat-suppressed gradient-echo (GRE) sequences [3–7]. MRI provides an adequate assessment of cartilage repair tissue and has led to the definition of pertinent variables for the description of articular cartilage repair tissue

A.A.M. Dhollander et al. / European Journal of Radiology 75 (2010) 72–81

after different biological repair techniques (MOCART: Magnetic Resonance Observation of Cartilage Repair Tissue) as described by Marlovits et al. [8,9]. Microfracture (M), autologous or allograft osteochondral transplantations (OATS) and autologous chondrocyte implantation (ACI) are currently used in the repair of cartilage defects [10–12]. Compared to other cartilage reconstructive techniques such as OATS, ACI presents some theoretical advantages, e.g. the use of autologous engineered material, reduced donor site morbidity and no treatment limitations related to defect size [13]. However, a number of problems have been observed with the standard first and secondgeneration ACI techniques. These include the difficulty in handling a delicate liquid suspension of chondrocytes at implantation surgery, the need to construct a watertight periosteum or collagen synthetic membrane seal using sutures, the need of a second operative procedure, and possible complications such as hypertrophy related to the use of a periosteal flap [14]. More importantly, during in vitro propagation of the chondrocytes, dedifferentiation of the cells can occur and afterwards these fibroblast-like chondrocytes show different biosynthetic properties than the original cartilage cells in the knee joint [15]. In the present study a novel biodegradable, alginate-based biocompatible scaffold containing human allogenic donor chondrocytes was used for the treatment of chondral and osteochondral lesions in the knee. Previous research showed that human chondrocytes keep their phenotype in alginate with neosynthesis of an extracellular cartilage matrix [16,17]. This technique avoids the problem of dedifferentiation of chondrocytes during culture and obviates the need for harvesting surgery. The human allogenic donor chondrocytes are retrieved, in this case, from the Ghent Tissue Retrieval Programme. The two aims of this paper were firstly to follow-up a new allogenic chondrocyte implantation technique by MRI with well defined postsurgical intervals over 1 year and secondly to compare the MRI scoring system MOCART and a modified MOCART score with the clinical scores WOMAC and VAS. 2. Materials and methods 2.1. Study population Patients with focal (osteo)chondral defects involving the femoral condyles, patella and trochlea, and with clinical symptoms (pain,

73

swelling, locking and “giving away”) were eligible for treatment. Exclusion criteria were an age under 10 and above 60 years, untreatable tibiofemoral or patellofemoral malalignment or instability, diffuse osteoarthritis or bipolar “kissing” lesions, major meniscal deficiency and other general medical conditions such as diabetes or inflammatory joint diseases (e.g. rheumatoid arthritis). This experimental treatment was approved by the Ghent University Hospital Ethics Committee and informed consent to participate in the study and to comply with the postoperative regimen was obtained from all patients. Twenty-one patients (13 males, 8 females) were treated consecutively and followed for 12 months. The right/left (R/L) ratio was 12/9. In all these cases the lesions were focal. Fifteen cartilage defects were located on the medial femoral condyle (MFC), 4 on the lateral femoral condyle (LFC), 1 on the patella and 1 on the trochlea. All lesions were ICRS (International Cartilage Repair Society) grade III–IV [18] and had a mean size of 2.6 cm2 (1–9.25 cm2 ). The etiology was traumatic in 12 cases and focal osteoarthritis lesions in 9 cases. Thus, 9 of 21 patients included in this study suffered from early stages of osteoarthritis or display deformities predisposing to osteoarthritis that are idiopathic or follow trauma. The mean age of the patients was 33 years (12–47 years). Previous surgery in 10 of the patients included 6 partial meniscectomies, 2 anterior cruciate ligament (ACL) reconstructions, 1 meniscal suture and 5 cartilage repair procedures, such as shaving (1), debridement (2) and microfracturing (2) of chondral lesions. In 5 patients associated procedures were performed: 1 ACL reconstruction, 1 Fulkerson osteotomy, 1 high tibial osteotomy, 1 lateral and 1 medial allogenic viable meniscal transplantation [19] (Table 1). 2.2. Chondrocyte harvesting and culture Human articular chondrocytes were isolated as described elsewhere, with a few modifications [20,21]. Briefly, human articular cartilage was obtained from the femoral condyles of different donors within 24 h of death. All donors had died after a short illness and were under 40 years of age. None of them had received corticosteroids or cytostatic drugs. Visually intact cartilage was harvested from the femoral condyles and diced into small fragments. The chondrocytes were isolated by sequential enzymatic digestion (hyaluronidase, pronase, and collagenase) of the extracellular

Table 1 Patient characteristics. Patients marked with asterisks (*,**) were treated with chondrocytes from the same donor. Size is given in cm2 . M: male, F: female, Part menisc: partial meniscectomy, ACL: anterior cruciate ligament reconstruction, menisc sut: meniscal suture, MF: microfracturing, FO: Fulkerson osteotomy, HTO: high tibial osteotomy, LM Tx and MM Tx: lateral and medial allogenic meniscal transplantation. Patients

Age

Sex

Side

Size

Site

Previous surgery

Associated procedure

1 2 3** 4** 5 6** 7 8 9 10 11 12 13* 14* 15 16** 17* 18 19 20 21

16 30 34 42 16 47 38 37 30 35 35 36 25 34 46 47 35 36 12 31 33

M M M F M F M M F F F M M M F F M M M F M

R L R R R L R R L L L R R R R L R R L L L

3.00 1.50 1.50 2.00 2.00 1.50 3.00 1.60 3.00 2.00 3.00 9.25 2.25 5.00 2.25 2.00 1.00 2.00 3.00 1.00 3.00

Trochlea LFC MFC LFC MFC MFC MFC MFC MFC MFC MFC Patella MFC MFC LFC LFC MFC MFC MFC MFC MFC

None MF, part menisc ACL, part menisc None Debridement None MF, part menisc Part menisc Menisc sut, part menisc None None Shaving Debridement None Part menisc None None None None ACL, part menisc None

None LM Tx None None None None None ACL None None HTO FO None None None None None MM Tx None None None

74

A.A.M. Dhollander et al. / European Journal of Radiology 75 (2010) 72–81

Fig. 2. The alginate beads were inserted manually.

Fig. 1. Alginate beads containing human allogenic chondrocytes, before (top) and after (bottom) removal of the nutrient medium.

matrix (ECM) [22]. Isolated cells were then centrifuged for 10 min at 800 rpm, washed three times in Dulbecco’s modified Eagle’s medium (DMEM; Gibco BRL, Grand Island, NY) supplemented with 10% acceptor serum, tested for viability (trypan blue exclusion test) and counted. More than 95% of the cells were viable after isolation. Chondrocyte cultures in alginate beads were prepared as described elsewhere, with some modifications [23]. Donor chondrocytes were suspended in one volume of double-concentrated Hanks’s balanced salt solution without calcium and magnesium (HBSS; Gibco BRL, Grand Island, NY), carefully mixed with an equal volume of 2% alginate (low viscosity, highly purified alginate from Macrocystis pyrifera; Sigma, St. Louis, MO) in HBSS, previously autoclaved for 15 min. The final cell concentration of chondrocytes was 20 × 106 /ml in 1% alginate. The chondrocyte/alginate suspension was then slowly dripped through a 23-gauge needle into a 102 mM calcium chloride solution in phosphate balanced buffer, pH 7.3. The beads were allowed to polymerize for 10 min at room temperature. The calcium chloride was then removed and the beads were washed three times with 0.15 M sodium chloride. The chondrocytes in the alginate beads were cultured for two weeks in a six-well plate in an incubator at 37 ◦ C under 5% CO2 (Fig. 1). Three ml of DMEM supplemented with 10% acceptor serum and 50 ␮g freshly dissolved ascorbate per milliliter were then added and replaced three times weekly. 2.3. Surgical technique A mini-arthrotomy in a tourniquet-controlled bloodless surgical field was performed in order to properly reach the defect. After freshening the bottom of the cartilage defect avoiding point bleeding from the subchondral bone, and trimming the edges of the defect back to stable walls of healthy cartilage, the lesion was measured. Once the defect was cleaned and ready to accept the alginate beads, it was sealed with a periosteal flap with the cambium layer facing the defect. The consistency of the implanted alginate beads was uniform. Before sealing the defect completely, a small opening was left unsutured for implantation of the alginate beads. The alginate beads were inserted manually (Fig. 2). Subsequently, the open edge of the flap was sutured to the remaining borders of the defect (Fig. 3), which was then easily sealed watertight with fibrin glue. The periosteal flap was sutured with single resorbable Vicryl 6/0 stitches. The periosteal flap was harvested from the proximal tibia. To this end, a skin incision was made in the centre of the medial tibial surface, depending on the size of the periosteal flap needed

to cover the defect. A flap typically has a tendency to shrink and should be slightly oversized by 2 mm in all directions. The postoperative regimen consisted of nonweightbearing for three weeks and the immediate institution of continuous passive motion (CPM). 2.4. MRI technique All MRI examinations before the operation and at 2, 4, 6 and 12 months of follow-up were performed on a 1.5 T MR unit (either a Magnetom Symphony or a Magnetom Avanto; Siemens Medical Solutions, Erlangen, Germany). Of the 21 patients, 12 had consented to follow the postoperative MRI evaluation protocol. We first used a standard knee MRI protocol incorporating proton density and FSE acquisitions for cartilage evaluation. Using a dedicated send-receive extremity coil, the following sequences were performed: - Sagittal proton density and T2-weighted turbo spin echo (TSE) images (TE: 24/96 ms; TR: 4000 ms; turbofactor: 5; slice thickness: 3 mm with a 0.3 mm intersection gap; field of view (FOV): 220 mm; matrix size: 307 × 512; number of signals acquired: 1) (Fig. 4). - Sagittal 3D fast low-angle shot images with water excitation (Flash 3Dwe): spoiled GRE. (TE: 27.0 ms; TR: 13.7 ms; flip angle: 30◦ ; slice thickness: 1 mm consecutively; FOV: 180 mm; matrix size: 208 × 256; number of signals acquired: 1) (Fig. 5). - Coronal and transverse 3D dual echo steady state (DESS 3D) images: GRE (TE: 5.5 ms; TR: 19 ms; flip angle: 40◦ for the images in the coronal plane and 25◦ for images in the transverse plane;

Fig. 3. Sutured periosteal flap sealing the alginate beads.

A.A.M. Dhollander et al. / European Journal of Radiology 75 (2010) 72–81

75

Table 2 Different variables of the MOCART (Magnetic Resonance Observation of Cartilage Repair Tissue) scoring system and their corresponding values allowing statistical analysis. Cartilage repair tissue grading scale (MOCART). Variables 1. Degree of defect repair and filling of the defect Complete (on a level with adjacent cartilage) = 3 Hypertrophy (over the level of the adjacent cartilage) = 2 Incomplete (under the level of the adjacent carilage; underfilling) >50% of the adjacent cartilage = 2 <50% of the adjacent cartilage = 1 Subchondral bone exposed (complete delamination or dislocation and/or loose body) = 0 2. Integration to border zone Complete (complete integration with adjacent cartilage) = 3 Incomplete (incomplete integration with adjacent cartilage) Demarcating border visible (split-like) = 2 Defect visible < 50% of the length of the repair tissue = 1 > 50% of the length of the repair tissue = 0

Fig. 4. Sagittal proton density fast spin echo MR image. One year after surgery, the cartilage defect is completely filled to the level of the adjacent cartilage, without hypertrophy. A small split is visible at the anterior margin, between the native cartilage and the repair tissue.

slice thickness: 2 mm in the coronal plane, 3 mm in the transverse plane, consecutively; FOV: 180 mm in the coronal plane, 160 mm in the transverse plane; matrix size: 192 × 256; number of acquisitions: 1). 2.5. Original MOCART system For the description of the repair tissue, we used the MOCART system previously published by Marlovits et al. [8,9]. Nine variables were used to describe the morphology and signal intensity of the repair tissue compared to the adjacent native cartilage (Table 2). The repair was considered complete when the repair tissue appeared as thick as the adjacent native cartilage with complete integration of the margins, and had a smooth articular surface that reproduced the original articular contour with no adhesions and an intact subchondral bone plate and marrow. The signal intensity of the repair tissue was evaluated separately for FSE (dual T2-FSE) and fat-suppressed

3. Surface of the repair tissue Surface intact (lamina splendens intact) = 2 Surface damaged (fibrillations, fissures and ulcerations) <50% of repair tissue depth = 1 >50% of repair tisue depth or total degeneration = 0 4. Structure of the repair tissue Homogenous = 1 Inhomogenous or cleft formation = 0 5. Signal intensity of the repair tissue Dual T2-FSE Isointense = 2 Moderately hyperintense = 1 Markedly hyperintense = 0 3D-GE-FS Isointense = 2 Moderately hypointense = 1 Markedly hypointense = 0 6. Subchondral lamina Intact = 1 Not intact = 0 7. Subchondral bone Intact = 1 Non-intact (edema, granulation tissue, cysts, sclerosis) = 0 8. Adhesions No = 1 Yes = 0 9. Effusion No = 1 Yes = 0

GRE (3D-GE-FS) sequences. A complete repair was graded as isointense when the repair tissue had the same signal intensity as the adjacent native cartilage. 2.6. Modified MOCART system

Fig. 5. Sagittal FLASH-3D spoiled gradient echo MR image with water excitation. One year after surgery, the repair tissue is clearly demarcated on this image by a line of metallic artefacts, both at the anterior and posterior border.

Besides the original MOCART system, we also used a modification of this system (Table 3) in order to increase the sensitivity. In our opinion, some variables were more important than others. We gave 3 times more weight to “degree of defect repair and filling of the defect”, “surface of the repair tissue”, “subchondral lamina” and “subchondral bone”. The weight of “integration to border zone” (Fig. 6) was doubled. Both morphological MRI classification systems were applied to the MRI images taken at 2, 4, 6 and 12 months of follow-up. All MR images were evaluated by a senior musculoskeletal radiologist. Both the original and modified MOCART scores were expressed as a percentage of the maximum score.

76

A.A.M. Dhollander et al. / European Journal of Radiology 75 (2010) 72–81

Table 3 Variables of the original MOCART system compared to the modified MOCART system (weighting factors). Original MOCART system

Modified MOCART system

Degree of defect repair and filling of the defect Integration to border zone Surface of the repair tissue Structure of the repair tissue Signal intensity of the repair tissue Subchondral lamina Subchondral bone Adhesions Effusion

Degree of defect repair and filling of the defect (×3) Integration to border zone (×2) Surface of the repair tissue (×3) Structure of the repair tissue Signal intensity of the repair tissue Subchondral lamina (×3) Subchondral bone (×3) Adhesions Effusion

status following a variety of interventions (pharmacologic, surgical, physiotherapeutic, etc.). It has already been used for the clinical evaluation of cartilage repair techniques in different studies [24,25]. In order to perform a statistical analysis of the WOMAC scores, each possible answer (none, slight, moderate, severe and extreme) corresponded with a score 0, 1, 2, 3 or 4. The VAS for pain is a simple, reliable [26] and valid [27] measurement instrument that measures the amount of pain a patient feels, ranging across a continuum from none to extremely severe. 2.8. Statistical methods All data are expressed in terms of means and standard deviations. The Wilcoxon test was used to analyze statistical differences between the preoperative and follow-up WOMAC and VAS pain scores, between the postoperative values of the original MOCART system and between the postoperative values of the modified MOCART system. To determine the correlation of clinical outcome to MRI, the VAS and WOMAC scores were correlated to the nine variables of the MRI scoring system. For the statistical analysis, the Spearman correlation coefficient (rs ) and the Student’s t-test were calculated. For all tests p < 0.05 was considered significant. Statistical analysis was performed using SPSS 11. 3. Results 3.1. Clinical outcome

Fig. 6. Sagittal proton density fast spin echo MR image. Four months after surgery there is no integration yet between the repair tissue and the surrounding cartilage, with gaps both at the anterior and posterior interface.

2.7. Clinical evaluation All 21 patients were clinically prospectively evaluated with use of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) and the Visual Analogue Scale (VAS) for pain preoperatively and at 3, 6, 9 and 12 months of follow-up. The WOMAC is a disease-specific, self-administered health status measure assessing clinically important symptoms in the areas of pain, stiffness and physical function in patients with osteoarthritis of the hip and/or knee. The index consists of 24 questions (5 dealing with pain, 2 with stiffness and 17 with physical function) and can be completed in less than 5 min. The WOMAC is a valid, reliable and sensitive instrument for the detection of clinically important changes in health

During the follow-up period the VAS scores for pain indicated by the patients improved significantly (Fig. 7). The differences between the preoperative and postoperative (3, 6, 9 and 12 months) values were statistically significant (p < 0.05). All WOMAC subscale scores (physical function, pain and stiffness) improved progressively with time when pre- and postoperative values were compared (Fig. 8). The WOMAC pain score was statistically significantly improved at all follow-up periods (p < 0.05). The WOMAC scores for physical function and joint stiffness only became different after 6, 9 and 12 months of follow-up (p < 0.05). One of the patients (4.76%) presented loosening of the periosteal flap, which was attributed to a failure of the surgical procedure. 3.2. Twelve-month longitudinal follow-up of the repair tissue with the original MOCART system The original MOCART scores remained stable over time (Fig. 9). There were no statistically significant differences between 2 months and 4, 6, 12 months of follow-up (p ≥ 0.05).

Fig. 7. Mean values and standard deviations of the Visual Analogue Scale (VAS) for pain: preoperative (pre) (89.40 ± 35.51) and postoperative: 3 months (68.25 ± 30.28; 0 → 3 months: p: 0.004), 6 months (41.50 ± 25.09; 0 → 6 months: p: 0.001), 9 months (20.85 ± 16.89; 0 → 9 months: p: 0.001) and 12 months (8.65 ± 13.05; 0 → 12 months: p: 0.001). Values are expressed in mm. The black squares indicate statistically significant differences (p < 0.05) between the preoperative and postoperative (3, 6, 9 and 12 months) values.

A.A.M. Dhollander et al. / European Journal of Radiology 75 (2010) 72–81

77

Fig. 8. Mean values and standard deviations of the WOMAC subscales. The black squares indicate statistically significant differences (p < 0.05) between the preoperative and postoperative (3, 6, 9 and 12 months) values. (a) Physical function: preoperative (pre) (32.90 ± 19.20) and postoperative: 3 months (29.25 ± 13.64; 0 → 3 months: p: 0.286), 6 months (15.85 ± 12.19; 0 → 6 months: p: 0.002), 9 months (8.85 ± 9.13; 0 → 9 months: p: 0.001), 12 months (4.20 ± 6.20; 0 → 12 months: p: 0.001). (b) Pain: preoperative (pre) (9.90 ± 5.21) and postoperative: 3 months (8.10 ± 3.89; 0 → 3 months: p: 0.045), 6 months (4.10 ± 2.73; 0 → 6 months: p: 0.001), 9 months (2.35 ± 2.16; 0 → 9 months: p: 0.001), 12 months (1.50 ± 1.79; 0 → 12 months: p: 0.001). (c) Stiffness: preoperative (pre) (3.60 ± 2.50) and postoperative: 3 months (3.40 ± 1.70; 0 → 3 months: p: 0.700), 6 months (2.05 ± 1.50; 0 → 6 months: p: 0.007), 9 months (1.10 ± 1.52; 0 → 9 months: p: 0.001), 12 months (0.55 ± 0.94; 0 → 12 months: p: 0.001).

Fig. 9. Mean values and standard deviations of the original MOCART scores expressed in percentages: postoperative: 2 months (53.13 ± 13.48), 4 months (55.56 ± 11.32; 2 → 4 months: p: 0.226), 6 months (51.89 ± 12.27; 2 → 6 months: p: 0.483) and 12 months (53.82 ± 10.87; 2 → 12 months: p: 0.276).

3.3. Twelve-month longitudinal follow-up of the repair tissue with the modified MOCART system

3.4. MRI data evaluated with the original MOCART system at 12 months of follow-up

As with the original MOCART system, a higher modified MOCART system score signified a more cartilage-like aspect of the repair tissue on MRI and a more complete filling of the defect, without it being overfilled. The percentages slightly decreased over time (Fig. 10). The differences between 2 months, 4 months and 6 months of follow-up were not statistically significant (p ≥ 0.05). With the modified MOCART system, the difference between 2 months and 12 months of follow-up was statistically significant (p < 0.05) indicating that the condition of the repair tissue deteriorated significantly.

One year after allogenic chondrocyte implantation the MRI data were analyzed according to original MOCART system (Table 4) [9]. Complete filling of the defect was found in one case (8.3%), and incomplete filling in seven cases, including one graft failure due to periosteal flap loosening. Hypertrophy of the repair tissue was seen in four cases (25.0%). Complete integration of the border zone with the adjacent cartilage was observed in 5 patients (41.7%). The surface of the repair tissue was intact in two patients (16.7%) and had a homogeneous structure in also two (16.7%) patients. An intact subchondral lamina was found in none of the patients (0%) and an intact subchondral bone in seven patients (58.3%). Adhesions were

Fig. 10. Mean values and standard deviations of the modified MOCART scores: postoperative: 2 months (49.29 ± 17.24), 4 months (51.97 ± 15.29; 2 → 4 months: p: 0.325), 6 months (48.35 ± 13.94; 2 → 6 months: p: 0.482) and 12 months (46.78 ± 15.24; 2 → 12 months: p: 0.027). The black square indicates a statistically significant difference (p < 0.05) between 2 months and 12 months of follow-up.

78

A.A.M. Dhollander et al. / European Journal of Radiology 75 (2010) 72–81

Table 4 MRI evaluation of repair tissue 1 year after allogenic chondrocyte implantation in terms of number and percentage.

Table 6 T-test determined when comparing the VAS and the three subscales form the WOMAC with MRI variables at 12 months of follow-up.

Cartilage repair tissue grading scale (MOCART) Variables

Structure

Subcondral lamina

Subchondral bone

Adhesions

Effusion

0.068 0.127 0.358 0.096

0.052 0.023* 0.054 0.061

0.05 0.040* 0.152 0.064

0.081 0.328 0.809 0.123

0.059 0.062 0.166 0.079

(N = 12)

1. Degree of defect repair and filling of the defect Complete (on a level with adjacent cartilage) Hypertrophy (over the level of the adjacent cartilage) Incomplete (under the level of the adjacent carilage; underfilling) >50% of the adjacent cartilage <50% of the adjacent cartilage Subchondral bone exposed (complete delamination or dislocation and/or loose body) 2. Integration to border zone Complete (complete integration with adjacent cartilage) Incomplete (incomplete integration with adjacent cartilage) Demarcating border visible (split-like) Defect visible < 50% of the length of the repair tissue > 50% of the length of the repair tissue 3. Surface of the repair tissue Surface intact (lamina splendens intact) Surface damaged (fibrillations, fissures and ulcerations) <50% of repair tissue depth >50% of repair tissue depth or total degeneration 4. Structure of the repair tissue Homogenous Inhomogenous or cleft formation 5. Signal intensity of the repair tissue Dual T2-FSE Isointense Moderately hyperintense Markedly hyperintense 3D-GE-FS Isointense Moderately hypointense Markedly hypointense 6. Subchondral lamina Intact Not intact

1 (8.3) 4 (33.3) 3 (25.0) 3 (25.0) 1 (8.3)

5 (41.7) 4 (33.3) 1 (8.3) 2 (16.7) 2 (16.7) 7 (58.3) 3 (25.0) 2 (16.7) 10 (83.3)

3 (25.0) 9 (75.0) 0 (0.0) 3 (25.0) 8 (66.7) 1 (8.3) 0 (0.0) 12 (100.0)

7. Subchondral bone Intact Not intact (edema, granulation tissue, cysts, sclerosis) 8. Adhesions No Yes

7 (58.3) 5 (41.7) 10 (83.3) 2 (16.7)

9. Effusion No Yes

8 (75.0) 4 (25.0)

present in two patients (16.7%). Eight patients (75%) showed no effusions. The signal intensities were described as isointense in 3 of the 12 patients (25%). 3.5. Correlation of clinical outcome with MRI The calculated correlation coefficient and the t-test for the subjective outcome and the different variables of the MRI classification system at 12 months of follow-up are presented in Tables 5 and 6. For the variable “3D-GE-FSE signal intensity of the repair tissue”,

VAS WOMAC pain WOMAC stiffness WOMAC ADL *

Statistically significant at p < 0.05.

a statistically significant correlation with the WOMAC stiffness subscale was observed. For the VAS, a negative correlation was calculated with the variable “surface of the repair tissue”, which indicated that the value decreased as the surface of the defect appeared more intact. For the variables “filling of the defect”, “integration to border zone” and “dual T2-FSE signal intensity of the repair tissue”, no statistically significant correlations were found. The variables “subchondral lamina” and “subchondral bone” showed a statistically significant correlation with the WOMAC pain scores. For the variables “structure”, “adhesions”, and “effusion”, no statistically significant correlation with the clinical scores was observed. 4. Discussion In this pilot feasibility study human mature allogenic chondrocytes in alginate beads were used to the repair of symptomatic cartilage defects. This procedure is a novel technique and requires only one surgical session contrary to the current two-staged surgical cartilage cell implantation techniques who need two interventions [25]. Classic ACI necessitates the use of chondrocytes isolated from biopsy samples taken from the patient [12]. These cells are propagated in monolayer cell culture which inevitably results in their dedifferentiation to fibroblast-like cells having lost their original phenotypical traits. The use of allogenic cells obviates the need for a biopsy and monolayer cell expansion. We chose alginate as a scaffold for cell seeding because of its ability to maintain the chondrocytic phenotype [28]. The patients included in this study showed a strong gradual clinical improvement after surgery, as demonstrated by the significantly improved WOMAC and VAS pain scores 3 months after allogenic chondrocyte implantation. The rate of improvement was slightly lower for the WOMAC stiffness and physical function scores, which significantly improved after 6 months and, together with the other clinical variables, continued to improve during the 12 months of follow-up. Other cartilage repair techniques including microfracture (MF), mosaicplasty and ACI also achieved good clinical results after 1 year of follow-up [29]. In a randomized trial Knutsen et al. [30] compared MF with ACI. Their study population consisted of 80 patients who were followed for 2 years. The mean VAS pain scores were ±55 (ACI), ±55 (MF) preoperatively, and ±40 (ACI), ±35 (MF) at 1 year, indicating relative 27% and 36% decreases in VAS pain scores for the ACI and the MF procedure, respectively. In the present study the VAS pain scores decreased by 90.3% 1 year after surgery. Minas [24] treated

Table 5 Spearman’s correlation coefficient (rs ) determined when comparing the VAS and the three subscales form the WOMAC with MRI variables at 12 months of follow-up. Filling of the defect VAS WOMAC pain WOMAC stiffness WOMAC ADL *

(−)0.081 (−)0.030 0.368 0.064

Statistically significant at p < 0.05.

Integration with border zone (−)0.286 (−)0.076 (−)0.056 0.074

Surface *

(−)0.633 0.112 0.008 0.299

Dual T2-FSE

3D-GE-FSE

0.232 (−)0.022 0.07 0.326

0.465 (−)0.348 0.602* 0.286

A.A.M. Dhollander et al. / European Journal of Radiology 75 (2010) 72–81

26 patients with ACI. The mean total WOMAC values were 34.1 before surgery and 19.6 at 1 year, resulting in a 42.5% decrease of the WOMAC score. Twelve months after implantation of alginate beads containing human allogenic chondrocytes, the WOMAC scores decreased by 86.5% in the present study. Compared with the results achieved by Knutsen et al. [30] and by Minas [24], our results suggest that the patients who participated in this study experienced more clinical symptoms before surgery and less 1 year after the procedure. In the present study both the original and the modified MOCART system were used in a longitudinal fashion to evaluate the repair tissue. The scores were expressed as a percentage of the maximum scores. In this way a comparison could be made between both MOCART systems. These percentages give a general idea of the aspect of the repair tissue: a higher percentage signified a better condition of the repair tissue on MRI in that the repair tissue exhibited a more cartilage-like appearance on MRI. This however does not mean that the tissue is effectively hyaline cartilage tissue, as already mentioned by Tins et al. [31], who have shown that the morphological appearance of cartilage implants did not correlate with histological findings. The original MOCART scores were moderate and remained stable over time. No statistically significant differences were observed indicating that the MRI findings of the repair tissue showed no significant signs of deterioration or improvement during the 12 months of follow-up. We also used a modified MOCART system. More weight was given to the variables “degree of defect repair and filling of the defect”, “surface of the repair tissue”, “subchondral lamina”, “subchondral bone” and “integration to border zone” because we felt that these were more important and had a better prognostic value. For instance, the condition of the subchondral bone plays an important role in cartilage metabolism [32] and may influence the final outcome of the repair tissue. Contrary to the original MOCART scores, the modified MOCART scores were statistically significantly different between 2 months and 12 months of follow-up, indicating that the aspect of the repair tissue on MRI significantly worsened over time. In other words, these MRI data showed signs of graft failure. Although different classification systems exist for the description of articular cartilage repair tissue, each has its limitations and deficiencies that can lead to confusion [1]. We opted for the MOCART system because it was found to be a valid, accurate and reproducible MRI grading and scoring method [9]. From our point of view, the original and the modified MOCART system are practical tools that allow us to morphologically evaluate cartilage repair techniques. Furthermore, the use of both systems in daily practice supports the conclusion that MRI is the standard imaging method of cartilage defects and repair. The main advantages of MRI are its noninvasiveness, reproducibility and accuracy [4,33–35]. The role of MRI in cartilage imaging should not be restricted to an evaluation tool alone, although more research is needed to explore the prognostic value of MRI findings in cartilage repair. The authors are aware that the 220 mm field of view is relatively large to evaluate cartilage implants. However this large FOV was only used in the sagittal proton density and T2-weighted turbo spin echo (TSE) images and not in the specific cartilage sequences (FLASH-3D with water excitation) which has isotropic voxels of less than 0.5 mm3 (0.7 × 0.7 × 1). Based on the original MOCART system, we analyzed our patient group 12 months after allogenic chondrocyte/alginate bead transplantation. The subchondral lamina was not intact in all cases of this study. Irregularities of the subchondral bone were also interpreted as a non-intact subchondral lamina. Before the alginate beads were implanted into the cartilage lesion, the bottom of the defect was

79

freshened. Point bleeding of the subchondral bone was thereby avoided. The edges of the defect were trimmed back to stable walls of healthy cartilage. It is possible that during this debriding of the defect, the subchondral lamina was touched. This can explain why the subchondral lamina was not intact. Henderson et al. recommended MRI at 12 months as a reasonable noninvasive method of assessing graft maturation after ACI [36]. Their MRI analysis of 81 lesions in 58 knees showed that 81.6% of the lesions had normal or nearly normal cartilage at the site of repair 12 months after implantation. Four variables were analyzed, including “filling of the repair site”, “signal intensity of the repair site”, “presence and severity of bone-marrow edema”, and “effusion”. A complete fill of the repair site was found in 79%, while 13.6% had 50–100% filling and 2.5% < 50% filling. In our series, we observed a complete filling or hypertrophy in 41.6% and underfilling in 50%. Henderson et al. reported signal intensities identical to those of the adjacent articular cartilage in 63% and nearly normal signals in 29.6%. We found normal signals in 25% of our patients. Henderson et al. also reported the presence of bone-marrow edema and effusion in 40.8% of their study patients, compared to 41.7% and 25%, respectively, in our patients. Overall, there was a marked difference in MRI findings between Henderson’s results after classic ACI and our results after allogenic chondrocyte/alginate bead implantation, suggesting that the repair tissue with the latter technique was of lower quality. On the other hand, the presence of bone-marrow edema was comparable between both studies and the presence of effusion scored slightly better in our patient group. Further follow-up is necessary to confirm whether these differences are due to a slower healing progress of the repair tissue after the allogenic chondrocytes/alginate beads or to a failure of the technique. However, care must be taken in the interpretation of MRI images, because even normal well-functioning grafts demonstrate signal abnormalities [37]. Graft failure through delamination has been reported to occur in approximately 5% of patients and usually presents within the first six postoperative months [38]. Henderson et al. found graft failure with a full-thickness defect was found in 4.9% of their patients [36]. Marlovits et al. found a graft failure in one patient (7.7%), with complete dissemination of the graft due to the emergence of borrelial arthritis [9,39]. In our study, one of the 21 patients (4.76%) presented loosening of the periosteal flap, which was attributed to a failure of the surgical procedure. The authors are aware that it is currently common practice to use a synthetic collagen membrane in this type of surgery, but in order to have a homogenous patient group a periosteal flap was used in all cases in the present study. In the future we intend to use a synthetic collagen membrane. We correlated the MRI variables with the subjective patient evaluation using the VAS and WOMAC criteria at 12 months of follow-up and compared our results with those published by Marlovits et al. [9] (Tables 7 and 8). At 24 months of follow-up Marlovits et al. found a statistically significant correlation between the clinical outcome and some of the radiological variables, including filling of the defect, structure of the repair tissue, changes in the subchondral bone and the signal intensities. At 12 months of follow-up Henderson et al. found no significant linear correlation between the overall MRI score and the four criteria (fill and signal at the repair site, bone-marrow edema and effusion), and the subjective and objective knee scores [36]. Our analysis could not confirm the findings of Marlovits et al. but rather supported the results published by Henderson et al. We did not find a good consistent correlation between the MRI criteria and the clinical outcome. However, the WOMAC pain scores were significantly correlated with subchondral bone/lamina abnormalities. The correlation between the VAS and subchondral bone/lamina abnormalities was nearly significant (0.050 and 0.052, respectively). Remarkably, neither Henderson et

80

A.A.M. Dhollander et al. / European Journal of Radiology 75 (2010) 72–81

Table 7 Spearman’s correlation coefficient (rs ) determined when comparing the subjective patient evaluation with MRI variables described by Marlovits et al. [9] and Dhollander et al. Filling

Integration

Surface

Dual T2-FSE

3D-GE-FSE

(1) Marlovits et al. VAS KOOS pain KOOS symptoms KOOS ADL KOOS sport KOOS QL

(−)0.860* 0.808* 0.633* 0.770* 0.698* 0.834*

(−)0.299 0.381 0.175 0.398 0.295 0.371

(−)0.414 0.345 0.474 0.415 0.35 0.31

(−)0.439 0.407 0.484* 0.482* 0.598* 0.349

(−)0.439 0.407 0.484* 0.482* 0.598* 0.349

(2) Dhollander et al. VAS WOMAC pain WOMAC stiffness WOMAC ADL

(−)0.081 (−)0.030 0.368 0.064

(−)0.286 (−)0.076 (−)0.056 0.074

(−)0.633* 0.112 0.008 0.299

0.232 (−)0.022 0.07 0.326

0.465 (−)0.348 0.602* 0.286

KOOS: knee injury and osteoarthritis outcome score, ADL: activities of daily living, QL: quality of life. * Statistically significant at p < 0.05. Table 8 T-test determined when comparing the subjective patient evaluation with MRI variables described by Marlovits et al. [9] and Dhollander et al. Structure

Subcondral lamina

Subchondral bone

Adhesions

Effusion

(1) Marlovits et al. VAS KOOS pain KOOS symptoms KOOS ADL KOOS sport KOOS QL

0.019* 0.032* 0.063 0.020* 0.037* 0.007*

0.92 0.937 0.968 0.749 0.893 0.619

0.001* 0.001* 0.067 0.001* 0.017* 0.002*

0.141 0.365 0.181 0.215 0.436 0.235

0.821 0.32 0.883 0.215 0.436 0.737

(2) Dhollander et al. VAS WOMAC pain WOMAC stiffness WOMAC ADL

0.068 0.127 0.358 0.096

0.052 0.023* 0.054 0.061

0.050 0.040* 0.152 0.064

0.081 0.328 0.809 0.123

0.059 0.062 0.166 0.079

KOOS: knee injury and osteoarthritis outcome score, ADL: activities of daily living, QL: quality of life. * Statistically significant at p < 0.05.

al., nor we found a good correlation between MRI data and clinical outcome at 12 months of follow-up. Marlovits et al. however, reported a significant correlation at 24 months of follow-up. This finding could perhaps be accounted for by the additional 12-month follow-up period in their study. The present study confirms the primary role of MRI in the evaluation of cartilage repair. Both the original and the modified MOCART system were used in a longitudinal fashion and allowed a practical and morphological evaluation of the repair tissue. However, a poor correlation between clinical outcome and MRI findings was observed. Further validation of these scoring systems is mandatory. The promising short-term clinical outcome of allogenic chondrocyte/alginate bead implantation was not confirmed by the short-term MRI data. Possible explanations are: firstly, in all cases repair tissue was seen, suggesting that covering of the pathological area with any kind of repair tissue already reduces the clinical symptoms; secondly, the natural evolution of the graft takes longer to heal than the clinical symptoms of the patients; thirdly, a drawback of this study, concomitant procedures could have influenced the clinical outcome. It must be emphasized that the small sample size lacked the necessary statistical power, that the follow-up period was limited to 12 months and that all MR images were evaluated by only one musculoskeletal radiologist. Drawbacks of the study population are the heterogeneity of the lesion location, the majority of the patients had previous knee surgery and the focal osteoarthritis lesions in 9 cases, however, there is a tendency to broaden inclusion criteria in recent studies concerning cartilage repair techniques, especially those discussing scaffold based techniques [40–45]. In general, more consensus concerning inclusion and exclusion criteria and how to deal with problems such as heterogeneity of the

lesion location and previous surgery is needed. Moreover, if we consider the modified Coleman methodology score relating to the quality of experimental clinical cartilage repair studies, as mentioned in the paper of Jakobsen et al. [46], few studies have a score above 60%. This study – with a matrix (allogenic chondrocytes/alginate beads) used in cartilage repair, which is the first to be described in the literature – allows us to present the clinical and MRI outcome at 1 year postsurgery and now serves as a control for future studies with modified protocols. These short-term results illustrate the feasibility of the application of the described technique. This concept essentially involves a one-step surgical transplantation. The results of this pilot study inspire us to search for further improvements. Future modifications have already been planned, including the selection of donors based on some ideal characteristics such as donor site [47], age [48] and physical condition, a quality control procedure on the implanted cells as advanced by Saris et al. [49], and the enhancement of the chondrogenic capacity of the alginate matrix by changing its chemical structure and characteristics or by adding growth factors [50,51]. Conflict of interest None. References [1] Imhof H, Nöbauer-Huhmann IM, Krestan C, et al. MRI of the cartilage. Eur Radiol 2002;12:2781–93. [2] Trattnig S, Huber M, Breitenseher MJ, et al. Imaging articular cartilage defects with 3D fat-suppressed echo planar imaging: comparison with conventional 3D

A.A.M. Dhollander et al. / European Journal of Radiology 75 (2010) 72–81

[3]

[4] [5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15] [16]

[17]

[18] [19]

[20] [21]

[22]

[23] [24] [25] [26]

fat-suppressed gradient echo sequence and correlation with histology. J Comput Assist Tomogr 1998;22:8–14. Potter HG, Linklater JM, Allen AA, Hannafin JA, Haas SB. Magnetic resonance imaging of articular cartilage in the knee. An evaluation with use of fast-spinecho imaging. J Bone Joint Surg Am 1998;80:1276–84. Recht M, White LM, Winalski CS, Miniaci A, Minas T, Parker RD. MR imaging of cartilage repair procedures. Skeletal Radiol 2003;32:185–200. Recht MP, Piraino DW, Paletta GA, Schils JP, Belhobek GH. Accuracy of fatsuppressed three-dimensional spoiled gradient-echo FLASH MR imaging in the detection of patellofemoral articular cartilage abnormalities. Radiology 1996;198:209–12. Disler DG, McCauley TR, Kelman CG, et al. Fat-suppressed three dimensional spoiled gradient-echo MR imaging of hyaline cartilage defects in the knee: comparison with standard MR imaging and arthroscopy. Am J Roentgenol 1996;167:127–32. Trattnig S, Mlynarik V, Huber M, Ba-Ssalamah A, Puig S, Imhof H. Magnetic resonance imaging of articular cartilage and evaluation of cartilage disease. Invest Radiol 2000;35:595–601. Marlovits S, Striessnig G, Resinger CT, et al. Definition of pertinent parameters for the evaluation of articular cartilage repair tissue with high-resolution magnetic resonance imaging. Eur J Radiol 2004;52:310–9. Marlovits S, Singer P, Zeller P, Mandl I, Haller J, Trattnig S. Magnetic resonance observation of cartilage repair tissue (MOCART) for the evaluation of autologous chondrocyte transplantation: determination of interobserver variability and correlation to clinical outcome after 2 years. Eur J Radiol 2006;57:16–23. Steadman JR, Briggs KK, Rodrigo JJ, Kocher MS, Gill TJ, Rodkey WG. Outcomes of microfracture for traumatic chondral defects of the knee: average 11-year follow-up. Arthroscopy 2003;19:477–84. Hangody L, Füles P. Autologous osteochondral mosaicplasty for the treatment of full-thickness defects of weight-bearing joints: ten years of experimental and clinical experience. J Bone Joint Surg Am 2003;85:25–32. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 1994;331:889–95. Peterson L, Brittberg M, Kiviranta I, Åkerlund EL, Lindahl A. Autologous chondrocyte transplantation. Biomechanics and long-term durability. Am J Sports Med 2002;30:2–12. Sgaglione N, Miniaci A, Gillogly S, Carter TR. Update on advanced surgical techniques in the treatment of traumatic focal articular cartilage lesions in the knee. Arthroscopy 2002;18:9–32. Benya PD, Schaffer JD. Dedifferentiated chondrocytes reexpress the differentiated collagen phenotype when cultured in agarose gels. Cell 1982;30:215–24. Almqvist KF. Human differentiated articular cartilage cells in biodegradable matrices. Preparative studies for their use in the repair of cartilage defects. A thesis submitted for fulfilment of the requirements for the degree: ‘Doctor in the Medical Sciences’. Ghent, Belgium; 2001. Almqvist KF, Wang L, Wang J, et al. Culture of chondrocytes in alginate surrounded by fibrin gel: characteristics of the cells over a period of eight weeks. Ann Rheum Dis 2001;60:781–90. Hjelle K, Solheim E, Strand T, Muri R, Brittber GM. Articular cartilage defects in 1,000 knee arthroscopies. Arthroscopy 2002;18:730–4. Verdonk PCM, Demurie A, Almqvist KF, Veys EM, Verbruggen G, Verdonk R. Transplantation of viable meniscal allograft. Survivorship analysis and clinical outcome of one hundred cases. J Bone Joint Surg Am 2005;87:715–24. Green Jr WT. Behaviour of articular chondrocytes in cell culture. Clin Orthop 1971;75:248–60. Kuettner KE, Pauli BU, Gall G, McMemoli VA, Schenk RA. Synthesis of cartilage matrix by mammalian chondrocytes in vivo. I. Isolation, culture characteristics and morphology. J Cell Biol 1982;93:743–50. Cornelissen M, Verbruggen G, Malfait AM, Veys EM, Broddelez C, De Ridder L. The study of representative populations of native aggrecan aggregates synthesized by human chondrocytes in vitro. J Tiss Cult Meth 1993;15: 139–46. Guo J, Jourdian GW, MacCallum DK. Culture and growth characteristics of chondrocytes encapsulated in alginate beads. Connect Tissue Res 1989;9:277–97. Minas T. Autologous chondrocyten implantation for focal chondral defects of the knee. Clin Orthop Relat Res 2001;391:349–61. Marlovits S, Zeller P, Singer P, Resinger C, Vécsei V. Cartilage repair: generations of autologous chondrocyte transplantation. Eur J Radiol 2006;57:24–31. Bijur PE, Silver W, Gallagher EJ. Reliability of the visual analog scale for measurement of acute pain. Acad Emerg Med 2001;8:1153–7.

81

[27] Gallaher EJ, Bijur PE, Laimer C, Silver W. Reliability and validity of a visual analog scale for acute abdominal pain in the ED. Am J Emerg Med 2002;20: 287–90. [28] Atala A, Kim W, Paige KT, Vacanti CA, Retik AB. Endoscopic treatment of vesicoureteral reflux with a chondrocyte-alginate suspension. J Urol 1994;152:641–3. [29] Hunziker EB. Articular cartilage repair: basic science and clinical progress. A review of the current status and prospects. Osteoarthritis Cartilage 2002;10:432–63. [30] Knutsen G, Engebretsen L, Ludvigsen TC, et al. Autologous chondrocyte implantation compared with microfracture in the knee. A randomized trial. J Bone Joint Surg Am 2004;86:455–64. [31] Tins BJ, McCall IW, Takashi T, et al. Autologous chondrocyte implantation in knee joint: MR imaging and histologic features at 1-year follow-up. Radiology 2005;234:501–8. [32] Radin EL, Rose RM. Role of subchondral bone in the initiation and progression of cartilage damage. Clin Orthop Relat Res 1986;213:34–40. [33] Bobic V, Noble J. Articular cartilage—to repair or not to repair. J Bone Joint Surg Br 2000;82:165–6. [34] Gold GE, McCauley TR, Gray ML, Disler DG. What’s new in cartilage? Radiographics 2003;23:1227–42. [35] McCauley TR, Disler DG. Magnetic resonance imaging of articular cartilage of the knee. J Am Acad Orthop Surg 2001;9:2–8. [36] Henderson IJ, Tuy B, Connell D, Oakes B, Hettwer WH. Prospective clinical study of autologous chondrocyte implantation and correlation with MRI at three and 12 months. J Bone Joint Surg Br 2003;85:1060–6. [37] Görtz S, Bugbee WD. Allografts in articular cartilage repair. J Bone Joint Surg 2006;88:1374–84. [38] Minas T, Peterson L. Advanced techniques in autologous chondrocyte transplantation. Clin Sports Med 1999;18:13–44. [39] Marlovits S, Khanakah G, Striessnig G, Vévsei V, Stanek G. Emergence of lyme arthritis after autologous chondrocyte transplantation. Arthritis Rheum 2004;50:259–64. [40] Marcacci M, Kon E, Zaffagnini S, et al. Arthroscopic second generation autologous chondrocyte implantation. Knee Surg Sports Traumatol Arthrosc 2007;15:610–9. [41] Ossendorf C, Kaps C, Kreuz PC, Burmester GR, Sittinger M, Erggelet C. Treatment of posttraumatic and focal osteoarthritic cartilage defects of the knee with autologous polymer-based three-dimensional chondrocyte grafts: 2-year clinical results. Arthritis Res Therapy 2007;9:R41. [42] Nehrer S, Domayer S, Dorotka R, Schatz K, Bindreiter U, Kotz R. Three-year clinical outcome after chondrocyte transplantation using a hyaluronan matrix for cartilage repair. Eur J Radiol 2006;57:3–8. [43] Manfredini M, Zerbinati F, Gildone A, Faccini R. Autologous chondrocyte implantation: a comparison between an open periosteal-covered and an arthroscopic matrix-guided technique. Acta Orthop Belg 2007;73:207–18. [44] Gobbi A, Kon E, Berruto M, Francisco R, Filardo G, Marcacci M. Patellofemoral full-thickness chondral defects treated with Hyalograft-C: a clinical, arthroscopic, and histologic review. Am J Sports Med 2006;34:1763–73. [45] Behrens P, Bitter T, Kurz B, Russlies M. Matrix-associated autologous chondrocyte transplantation/implantation (MATC/MACI)-5-year follow-up. Knee 2006;13:194–202. [46] Jakobsen RB, Engebretsen L, Slauterbeck JR. An analysis of the quality of cartilage repair studies. J Bone Joint Surg 2005;87:2232–9. [47] Treppo S, Koepp H, Quan EC, Cole AA, Kuettner KE, Grodzinsky AJ. Comparison of biomechanical and biochemical properties of cartilage from human knee and ankle pairs. J Orthop Res 2000;18:739–48. [48] Verbruggen G, Cornelissen M, Almqvist KF, et al. Influence of aging on the synthesis and morphology of the aggrecans synthesized by differentiated human articular chondrocytes. Osteoarthritis Cartilage 2000;8:170–9. [49] Saris DBF, Vanlauwe J, Victor J, et al. Characterized chondrocyte implantation results in better structural repair when treating symptomatic cartilage defects of the knee in a randomized controlled trial versus microfracture. Am J Sports Med 2008;36:235–46. [50] Genes NG, Rowley JA, Mooney DJ, Bonassar LJ. Effect of substrate mechanics on chondrocyte adhesion to modified alginate surfaces. Arch Biochem Biophys 2004;422:161–7. [51] Rowley JA, Madlambayan G, Mooney DJ. Alginate hydrogels as synthetic extracellular matrix materials. Biomaterials 1999;20:45–53.