PURPOSE

Sciatica is pain radiating from lumbar spine or the buttock downward along the sciatic nerve into one leg or, less frequently, both legs. The simultaneous TOF ¹⁸F-FDG PET/MRI is an innovational diagnostic approach for chronic sciatica [1], especially for lumbar spinal implant patients. The high sensitivity of ¹⁸F-FDG PET is helpful to detect an abnormal increase in metabolism caused by painful inflammation while the MRI can offer the subtle structural abnormalities. However, metallic implants would result in the signal loss in the MRAC images, and it makes the large artifacts of metallic implants on 3D T₁-weighted LAVA-Flex images, which is usually used for MRAC images. The inaccurate MRAC images would lead to inaccuracy SUV data of patients with metallic implants. To eliminate the influence of metallic implants in MRAC images, previous literatures mentioned that the TOF information might reduce the effect of the incorrect AC map on the PET image [2,3]. Yet no literatures demonstrated the improvement of MRAC with TOF in lumbar assessment, and therefore, this study aimed to compare mean standardized uptake value (SUV_mean) of both side paraspinal and psoas muscles in patients with and without implant for TOF reconstructed PET images.

METHODS

A total of 6 patients (3 without and 3 with implant) who underwent an ¹⁸F FDG PET/MR of lumbar spine and lower extremity with chronic sciatica symptoms from June to October 2021 were collected retrospectively. Exclusion criteria were the abnormal uptake of lumbar paraspinal muscle due to lesions or adverse local tissue reactions (Fig. 1). One hour after a standard dose injection of 370.0 MBq ¹⁸F-FDG, patients underwent PET/MRI in a GE scanner (SIGNA TOF PET/MR; GE Healthcare). They were scanned from lower thoracic spine to the feet in 5 or 6 bed positions for around 1 hour, depending on the height of the patient. The rainbow color scale ranging from 0 to 3 for SUV value was used to render the PET images reconstructed with product MR based attenuation correction techniques (Fig. 2). The following MRI sequences were performed simultaneously with PET scans for each bed: 3-dimensional coronal LAVA Flex (repetition time (TR), 9.2 ms; echo time (TE), 1.9 ms; resolution, 1x1x1 mm; flip angle, 12), and 2-dimensional axial T2-weighted Flex (TR, 8600 ms; TE, 90 ms; resolution, 1.25x2x4 mm; echo train length, 15) with an integrated spine coil, and two 32-channel body phase-array coils were used. All images were reviewed by using Advantage Workstation 4.7 (GE Healthcare). ROIs were drawn over six normal both side paraspinal and psoas muscles (at the level of L4) in patients with and without implant to estimate SUVmean. Statistical analyses were performed using MATLAB R2017b. The paired t-tests were performed to compare SUVs obtained from MR attenuation correction.

RESULTS and DISCUSSION

The overall scanning time for PET/MRI was 56.6±4.2 minutes. The mean time interval between FDG injection and PET/MR image acquisition was 62.3±3.5 minutes. The data of both side paraspinal and psoas muscles were divided into two groups. The average SUV of right Multifidus, left Multifidus, right Psoas Major, left Psoas Major were 0.58±0.03, 0.61±0.03, 0.59±0.03, and 0.62±0.03, respectively, for no-implant group; the average SUV were 0.56±0.05, 0.6±0.03, 0.63±0.08, and 0.63±0.09, respectively, for with-implant group (Fig. 3a). The SUV_mean of normal psoas muscles using MR attenuation-corrected is 0.63± 0.16 [4], and our results were agreed with the values illustrated in the previous literatures. Using paired t-test, no significant differences in the average SUV among the no-implant and with-implant groups (p > 0.44). Further, the average SUV ratio (SUVR) of Multifidus normalized to Psoas was used to compare regional differences in lumbar metabolism pattern between no-implant and with-implant patients. No significance was found in the SUVR between no-implant and with-implant groups (p > 0.1) (Fig. 3b).

CONCLUSION

This study showed overall no significance in the SUV measurements obtained from MRAC between no-implant and with-implant patients. SUVs measured from TOF PET/MR of lumbar implant indicated good reliability and robust of MRAC.

Acknowledgements

No acknowledgement found.