Purpose Autologous chondrocyte membrane transplantation (ACMT), by combining autologous chondrocyte transplantation technique1 and cell membrane technology2, is a novel technique that forms an autologous cartilage cell sheet for the treatment of cartilage defects. Magnetization transfer MRI can effectively quantify in vivo alterations to the amount of free water and bound water (water bound to collagen) to generate image contrast within cartilage tissue3. However, the relationship between MT MRI measurements and ACMT treatment of cartilage defects has not been established. The objective of the present study was to demonstrate the feasibility of using MT-MRI for evaluation of knee cartilage recovery following the treatment of cartilage defects using ACMT.

Methods The study was approved by the clinical research ethics committee. Eleven patients underwent arthroscopic implantation of autologous chondrocyte membrane. The patients were evaluated clinically with a visual analog scale (VAS) for pain and Lysholm scores before surgery, as well as 6 days, 1 month, 3 months, 6 months, 12 months, 24 months and 36 months respectively after surgery. MRI evaluations were carried out using a 1.5 T MR scanner (MAGNETOM Aera, Siemens Healthcare, Erlangen, Germany) with an 8-channel knee coil. MT-MRI was performed at 24 and 36 months after the ACMT. T2 weighted images were acquired with the following parameters: TR = 1000 ms, TE = 69 ms, slice thickness = 3 mm, number of average (NA) = 1, FOV = 160 x 160 mm2, matrix = 256 x 256, pixel bandwidth = 225 Hz. MT MRI sequence included a Gaussian pre-saturation pulse followed by a three-dimensional (3D) spoiled gradient-echo pulse sequence. The Gaussian RF pulse was applied with pulse length of 10 ms, flip angle (FA) of 600°, and off-resonance frequency of 1.5 kHz. The scanning parameters included: TR = 35 ms, TE = 3.55 ms, FA = 15 degree, bandwidth = 300Hz/pixel, matrix = 320 × 320, slice thickness = 1 mm, FOV =160 × 160 mm2, number of slices = 96. MTR maps were generated voxel-wise using MATLAB software (MathWorks, Natick, MA using the following equation MTRs = 100 × Msat / M0, where Msat and M0 represents the signal intensity of images with and without MT saturation. Region-of-interests were identified and drawn by a radiologist on the MT weighted images and copied to the corresponding MTR maps.

Results A representative lesion area in the knee cartilage is shown in Fig. 1a. Eleven patients were scanned at 24 months after ACMT implantation with five patients experienced 36-month MRI follow-up to date. The average VAS for pain of the eleven patents decreased from 3.91 pre-surgery to 0.88 at 36 months after implantation. The average Lysholm scores increased from 50.45 to 90.67 at 36 months after surgery. T2 weighted images at 24 and 36 months after the ACMT implantation are shown in Fig. 1b and c. The MT-weighted images and MTR maps overlaid on the MT-weighted images at 24 and 36 months are shown in Fig.2. A visible increase of MTR values at the lesion area (circles in Fig.2d) at 36 months after implantation comparing to 24 months (Fig. 2b). As shown in Fig. 3, an increase of the average MTRs at the area of ACMT implantation was found from 31.16 ± 3.54 at 24 months from the eleven patients to 35.43 ± 5.64 at 36 months after ACMT implantation, while normal articular showed consistence of MTRs of 37.84 ± 3.68 at 24 months and 38.53 ± 4.09.

Discussion MRI techniques, such as T1 rho, dGEMRIC, gagCEST, and sodium imaging, require either ultra-high magnetic field systems or other dedicated hardware or software, thus having unclear clinical relevance. Preliminary studies showed controversy results between damaged and repaired cartilage MTR in repair tissue 4,5. Though further investigations are necessary, our results indicated MTR to be capable of detecting differences between normal cartilage and areas of cartilage repair using the ACMT technique and possibly a useful tool in imaging biochemical changes in cartilage after repair.

Conclusion Our study demonstrated the feasibility of MT MRI for evaluation of knee cartilage recovery following the treatment of ACMT in the clinical MRI settings.