Vascular Alterations and Recruitment in Spinal Cord Injury Revealed by Multislice Arterial Spin Labeling (ASL) Perfusion Imaging

Introduction: In spinal cord injury (SCI) investigation, the combination of diffusion tensor imaging (DTI) and perfusion imaging has the potential to be a useful tool in the detection of functional impairments, white matter tract disruption and deficient tissue blood supply, but also in the evaluation of functional recovery and tissue repair. Whereas multislice DTI is widely used for SCI models investigation, allowing then a large volume coverage per imaging session, assessment of mouse SC blood flow (SCBF) by MRI, which has recently been demonstrated to be feasible, currently relies on the use of a single slice arterial spin labeling (ASL) technique (presaturated FAIR). To better characterize the lesion (regional extension), and to be able to detect potential secondary injury, it would be important to match perfusion with multislice DTI. Multislice ASL in a single imaging session was achieved by the modification of the original presat-FAIR sequence to a presat-FAIRQUIPSSII sequence, optimized to mouse SC. Multislice DTI and ASL were then applied in a follow-up study performed over time on mice having received SCI (compression) at the cervical level. Resulting DTI metrics and SCBF values were additionally correlated to functional assessment tests. Methods: SCI model and functional assessment: Experiments were performed on C57Bl/6J mice (age 10 weeks, 20g). The spinal cord compression was induced by inflation (10 mm, 2.5 bar, 10s duration) of a balloon connected to a catheter and inserted at the C4 epidural space of the SC. Following the compression, mice suffered from left fore-limb paralysis. Grasping test, performed with a Bioseb apparatus (incline grid connected to a strength gauge), measured the developed fore-limb force. MR Imaging: Experiments were performed on an 11.75T vertical MR system (Bruker, AV 500WB) with a transmitter/receiver volume coil (∅ 2cm, length 3cm). A 4shot SE-EPI sequence was used for high resolution imaging (100x100μm, slice thickness 0.75mm). DTI was obtained using a standard Stejskal-Tanner sequence with parameters described in [6]. Multislice (4 slices) perfusion imaging was obtained with a presat-FAIR-QUIPSSII sequence, for which timing parameters were previously optimized on healthy mice. Unlike in human studies, the global inversion pulse labeled almost all the blood, making then possible the use of long inversion time (TI1=1.0s), which is beneficial for SNR considerations. The delay ΔTI after the saturation was optimized to 0.2s. This value ensured that the entire tagged bolus released from the tagging region was delivered to the imaging slices at time TI2 (TI2=TI1+ΔTI). Under these conditions the following equation applied for quantitative SCBF values measurements: ΔM=2Mb/λ.SCBF.α.e/(R1app-R1a).(e-1), with λ =0.9 ml/g (water blood/tissue partition coefficient) and R1a=1/2.1 s (blood longitudinal relaxation rate). Mb (equilibrium magnetization), α (inversion efficiency) and R1app (SC tissue apparent longitudinal relaxation rate) were determined with a slice-selective inversion recovery prescan. Magnetization difference ΔM was averaged during 45 minutes leading to a maximum total experimental time (EPI-adjustment, DTI and ASL) of 2 hours. Force tests, DTI metrics (FA, λ//, λ⊥) and absolute SCBF values were evaluated 1, 8, 21 and 49 days after the SCI. Results: Figure 1 shows the evolution with time of FA, λ// and SCBF obtained on 3 slices (centered on lesion, S1, +0.75mm rostral, S2, and +1.5mm rostral, S3).