We reviewed all 23 patients between the ages of 12 and 21 years with OCD lesions in the patella treated by cell-based therapy from 2001 to 2008. Twenty patients had ACI, whereas three patients had BMSCs implanted. The indications for surgery were: (1) pain; (2) locking; and (3) instability. The contraindications included: (1) early osteoarthritic changes; (2) inflammatory diseases such as rheumatoid arthritis; (3) previous trauma or infection to the growth plate; and (4) underlying tibial femoral malalignment. For this study we included patients only with International Cartilage Repair Society (ICRS) Grade 3 or 4 patella OCD diagnosed clinically and radiographically. We excluded patients older than 21 years, those with concomitant inflammatory arthritis, or syndromic or osteochondral defects as a result of traumatic dislocations. Of the 23 patients, 19 (83%) were male and only four were female patients representing a male-to-female ratio of 5:1. The mean age of our patients was 16.8 years (range, 12–21 years). The majority were Chinese (61%), Malay (26%) followed by Indian (13%). Previous major or insidious trauma was reported in 83% of our patients. The minimum followup was 2 years (mean, 6 years; range, 2–11 years). No patients were lost to followup. Data were retrieved from an actively maintained database of patients obtained through prospective interviews preoperatively and at followup visits. The study protocol was approved by the National Healthcare Group Domain-Specific Review Board (NHG DSRB reference number D/00/814) and the University Hospital Ethic Committee.
Preoperatively all patients had plain radiographs of the knee, including AP, lateral, intercondylar, and skyline views. All patients underwent CT scans of tracking of the patella at 0°, 10°, and 20° of flexion to determine patella subluxation, tilt, and congruence angle. Depending on the degree of patella tilt and/or subluxation, concomitant distal realignment or lateral release surgeries were performed. The criteria for realignment surgery were (1) increased tibial tubercle-trochlear groove distance > 15 mm; and/or (2) increased patellar tilt > 20° guided by a clinical assessment of a patellar glide. Concomitant Elmslie-Trillat and Roux-Goldthwaite were performed in four and two patients, respectively.
All patients were treated by the senior author (JHH). For ACI, a small amount of cartilage tissue (1 cm × 0.5 cm) was taken from nonweightbearing areas that were deemed macroscopically healthy by arthroscopy. The harvested tissue was transferred into a specimen container filled with sterile saline (10 mL) and processed within 60 minutes. The sample was washed twice with phosphate-buffered saline (Gibco BRL, Grand Island, NY, USA) and then minced before being transferred aseptically into a tube with 5 mL collagenase NB6 (Sigma, St Louis, MO, USA) for overnight digestion at 37°C in a water bath. Digested chondrocytes were washed with Dulbecco modified Eagle medium (DMEM)/F12 (Gibco BRL) supplemented with 10% fetal bovine serum (FBS) (Gibco BRL) to stop the enzymatic reaction. These cells were then cultured in T-75-cm2 flasks with DMEM/F12 containing 10% FBS and 50 mg/Ml L-ascorbic acid 2-phosphate sesquimagnesium salt hydrate (Sigma) in a humidified atmosphere of 5% CO2 at 37°C. Cells were seeded at a cell density of 5000 cells/cm2. We changed the initial medium change after 7 days, when adherent cells were recognized. We subsequently changed the medium two to three times a week until the preparation of cell sheets, which were formed in the presence of ascorbic acid (passage 1). For each surgery, at least four cell sheets were prepared and approximately 2 million cells/cm2 were applied.
For BMSCs, with the patient under local anesthesia, 30 mL of bone marrow was aspirated using a Jamshidi needle from the iliac crests of each patient into heparinized syringes and transferred into sterile containers. Seventy milliliters of each patient’s blood was collected as well. The bone marrow aspirate was processed within 60 minutes. The heparinized bone marrow aspirate was mixed with a one-fifth volume of 6% (w/v) dextran (molecular weight 100,000) (Sigma) and left standing at room temperature for 30 minutes to eliminate erythrocytes. The remaining cells were washed twice with DMEM. These cells were cultured in T-75 cm2 flasks with an initial culture medium consisting of DMEM containing 10% FBS, 50 mg/mL l-ascorbic acid 2-phosphate sesquimagnesium salt hydrate, and 1% antibiotic-antimycotic (penicillin 100 U/mL, streptomycin 0.1 mg/mL, amphotericin B 0.25 mg/mL) (Sigma) in a humidified atmosphere of 5% CO2 at 37°C. The cells were seeded at a density of 10,000 cells/cm2. We initially changed the medium after 5 days when adherent cells were recognized. Subsequently, culture media without antibiotics were used and changed two to three times a week. Sheets of cells were formed in the presence of ascorbic acid (passage 1) and for each surgery. A minimum of four cell sheets with approximately 2 million cells/cm2 was applied. Seventy milliliters of venous blood from each patient was transferred into two 50-mL tubes for overnight incubation at 4°C. After centrifuging the tube with slow acceleration, the serum was carefully aspirated and transferred to a new tube. Repeated centrifugation with slow acceleration for 3 minutes at 3000 rpm at ambient temperature was performed. The serum was aspirated into a syringe and filtered with a sterile 0.2-mm filter. The filtered serum was tested for sterility, antihuman immunodeficiency virus, and hepatitis B antigen, and then stored at a temperature of −20°C. Flow cytometry against CD90+, CD105+, CD14−, and CD34− was used to confirm that cultured cells were mesenchymal stem cells. Saline that was used for transporting the cartilage biopsy specimen to the laboratory, aspirated bone marrow, and culture media (without antibiotic) were tested for sterility and Mycoplasma hominis contamination. For each surgery, at least 10 to 15 million cells (with a viability rate of 96%) were returned for implantation.
Four to 5 weeks after the cells were harvested, surgery was performed following the guidelines outlined by Brittberg et al. [5
]. The cell sheets were transported to the operating room in a sterile container within the patients’ own serum. The débrided chondral defect (without damaging subchondral bone) was measured after arthrotomy. Subsequently, we harvested a periosteal patch from the proximal part of the tibia or distal part of the femur according to the measured size. Next, the harvested periosteum was sutured precisely to the rim of the débrided defect(s). The cultured chondrocytes or BMSCs were implanted beneath the patch and very fine stitches (microsuture 7-0) were used to hold the periosteum to the defected site. To avoid cell leakage, we used fibrin glue to create a watertight seal.
To derive maximum benefit from the surgery, patients were advised to strictly follow the rehabilitation protocol. The rehabilitation protocol began on the day of surgery and included passive ROM and isometric muscle contractions. Patients were able to begin active motion and partial weightbearing at 6 weeks progressing to full weightbearing exercises. The rehabilitation protocol varies according to the patient age, previous activity level, and concomitant procedures performed. Rehabilitation focuses on four areas: walking/weightbearing, ROM, strength, and cardiovascular capacity.
Patients were evaluated preoperatively and at 6, 12, and 24 months postoperatively. Assessments were performed by our trained research staff using the ICRS Cartilage Injury Evaluation Package, which included questions from the IKDC subjective knee evaluation form, the knee scale reported by Lysholm and Gillquist [18
], and the activity level scale reported by Tegner and Lysholm [27
The MIXED effect model (with random intercept) was used to evaluate the effect of visit time on individual outcomes such as IKDC, Tegner-Lysholm, and Lysholm-Gillquist scores, respectively. This method of analysis appropriately accounts for the possible correlation between repeated measurements of an individual. Residual plots examining the effect of visit time on IKDC, Tegner-Lysholm, and Lysholm-Gillquist scores showed that the residuals were approximately normally distributed with a mean of 0. All statistical evaluations were made assuming a two-sided test using STATA Version 11 (StataCorp, College Station, TX, USA).