Non-invasive longitudinal assessments of large bone repairs with specific implanted scaffolds is of major concern today in the field of bone tissue engineering. Being a non-ionizing technique, Magnetic Resonance Imaging remains highly suited for in vivo evaluation of tissue repair over long time periods. Here, three differently conditioned bone scaffolds (containing either nanocrystalline hydroxyapatite (na-HAp) or fucoidan (F) or both) were implanted in rats condyles. Knowing that the cortical bone exhibits much shorter T₂ relaxation times than implanted scaffolds and that substantial bone marrow fat composition hampers scaffold detection, a (bSSFP) MRI sequence coupled to a proper fat suppression module and banding artifacts correction method was developed to visualize in 3D those scaffolds.

Then, a 3D DCE-MRI (Dynamic Contrast Enhanced-MRI) was performed to analyze the evolution of the perfusion within the different scaffolds. Here, the aim of our work was to determine using MRI whether combined HApF construct shall significantly stimulate osteogenesis and angiogenesis for correct bone tissue repair.

Biomaterials: Three differently conditioned bone constructs, each composed of a pullulan/dextran scaffold associated with either na-HAp or F or both (i.e PuDna-HAp, PuD-F, or PuDna-HApF), were implanted into a large bone defect situated in femoral rat condyles (n=12).

MRI protocol: For each construct, longitudinal MRI at 7T was performed at 1, 3 and 5 weeks after implantation. A 3D water-selective bSSFP sequence was implemented [1] with the following parameters: 5 150μs-hermite subpulses with an intensity ratio of 1:2:3:2:1 (interpulse delay of 200μs); TE/TR=2.6/5.2ms; reception bandwidth (BW)=75kHz; FOV=30x35x25mm³; matrix=192x192x128; 4 phase offsets; FA=27°; 2 averages; acquisition time(TA)= 4min15s. 3D images were reconstructed after a “Sum-Of-Square” of the 4 offsets, eliminating the banding artifacts. DCE T1-weighted images were also acquired using a 3D gradient-echo sequence (FLASH): TE/TR=1.9/6.5 ms; BW=75kHz; FOV=25x30x14mm³; matrix=128x96x32; FA=30°; 1 average (19sec); 144 repetitions and TA=45mins36s. After acquiring 3 baseline images, Gd-DOTA was administered as a rapid bolus (0.2 mmol of Gd/kg).

Image analysis: Volumetric data were obtained through pixel count of the biomaterial hyperintense signal on the bSSFP images using Amira Software. All DCE-MRI analyses were performed on a pixel-wise basis using in-house programs developed with Matlab. To minimize effects from inter-individual physiological variation and contrast agent dose variability, each pixel was normalized to a mean reference signal measured in leg muscle and the noise. The area under curve after the 45 minutes follow up AUC₄₅ was computed for each pixel.

The binomial pulse induced a drop in fat signal located subcutaneously and within the bone marrow in the entire 3D FOV. The biomaterials were easily detected as hyper-intense areas within the bone marrow. Each PuDna-HAp, PuD-F, or PuDna-HApF constructs showed different volume reduction patterns over time (as in fig 1) with a significant drop to null in volume for the HAp and the combined HApF constructs on the 5^th week. However, the pattern for PuD-F constructs revealed a much slower decrease (≃40% volume reduction). DCE-MRI images correlated with volumetric analysis. AUC₄₅ parametric maps (in fig 2) showed that the biomaterial AUC45 values was slowly converted into that of the healthy bone on week 5 for HAp and HApF constructs. This biomaterial degradation has been correlated with cortical bone formation observed with µCT.The 3D water-selective bSSFP sequence allowed precise assessments of implanted biomaterials degradation longitudinally. AUC analysis and histological evaluations are ongoing for further conclusive results on promoting bone regeneration and angiogenesis stimulations in the bone defects. This current research would surely lead to a better tissue-engineered bone construct suitable for large bone defects as opposed to gold standard bone grafts.