100 Section - Reliability of multifidus MRI signal versus ultrasound-assisted Mini-Paraspinal Mapping after lumbar medial branch neurotomy - Preliminary data from a pilot study

Evaluate the validity of multifidus Magnetic Resonance Imaging Short Tau Inversion Recovery (MRI STIR) signal versus ultrasound-assisted Mini-Paraspinal Mapping (Mini-PM) Electromyography (EMG) to document denervation following fluoroscopically-guided lumbar medial branch radiofrequency neurotomy (LMBRN).

Design: Prospective cohort study from patients who were candidates for LMBRN after screening with anesthetic injections. Baseline lumbar spine MRI was documented and ultrasound-assisted Mini-PM documented pre-LMBRN multifidus signal. Unilateral LMBRN was performed at the identified spinal levels using a single anatomically-informed technique. 2-4 weeks after LMBRN Mini-PM and research MRI were collected. Initial analysis planned after data collected for 10 subjects to perform reliability and validity assessment of research MRI sequences. Protocol was approved by VA TVHS IRB.

Results:

9 subjects were recruited for this phase of the study. One subject was removed from data analysis due to presence of confounding MRI findings on pre-LMBRN MRI, but remaining subjects had no confounding findings on MRI or Mini-PM. All subjects were male, average age 45.6 (range 26-57). Research MRI STIR protocols in the axial and coronal planes were developed. Ultrasound assistance during Mini-PM confirmed sampling depth. Considering Mini-PM as the criterion standard for denervation, segmental multifidus MRI STIR signal showed sensitivity 98%, specificity 90.5, positive likelihood ratio 10.3 and negative likelihood ratio 0.02. Both Mini-PM and MRI STIR signal demonstrated 95% accuracy to the level treated by LMBRN.

Conclusion:

Preliminary review of these data indicate that segmental multifidus MRI STIR signal corresponds well to ultrasound-assisted Mini-PM findings of multifidus denervation. These data support the proof of concept that MRI STIR signal offers a non-invasive way to document physiologically successful LMBRN. Additional validation of MRI STIR signal analysis is underway. Additional data will be collected to document whether denervation can predict clinical success after a variety of LMBRN techniques.