Objective: Selective dorsal rhizotomy via a single-level approach (SL-SDR) to treat spasticity 100% relies on the interpretation of results from the intra-operative neuroelectrophysiological monitoring. The current study is to investigate the role EMG interpretation plays during SL-SDR procedure with regard to the selection of nerve rootlets for partially sectioning in pediatric cases with spastic cerebral palsy (CP).
Methods: A retrospective study was conducted in pediatric patients with spastic CP undergone our modified rhizotomy protocol-guided SL-SDR from May 2016 to Mar. 2019 in our hospital. Our study focused on intra-operative EMG interpretation and its correlation with pre-op evaluation results, and dorsal rootlet selection difference when data of our intra-operative EMG recordings interpreted using different rhizotomy protocols.
Results: Clinical and intra-operative neuroelectrophysiological monitoring data of a total of 318 consecutive cases were reviewed, which include 231 boys and 87 girls with 32 hemiplegias, 161 diplegias, and 125 quadriplegias. Age at the time of SL-SDR in those cases was between 3.0-14.0 (5.9 ± 1.9) years. The number of targeted muscle ranged from 2 to 8 over these cases (the muscle in lower limbs with its pre-op muscle tone ≥ 2 grade, Modified Ashworth scale). Among 21,728 nerve rootlets tested (68.3 ± 8.2/case), 6272 (28.9%) were identified sphincter related by our intra-operative neuromonitoring. In the rest of 15,456 (48.6 ± 7.6/case) nerve rootlets which neuromonitoring suggested associated with lower limbs, 11,009 were taken as the dorsal ones (34.6 ± 7.4/case). A total of 3370 (10.6 ± 4.7/case) rootlets matched our rhizotomy criteria with 3061 (9.6 ± 4.1/case) sectioned 50% and 309 (1.0 ± 1.0/case) cut 75%. The rhizotomy ratio (partially transected nerve rootlets/all dorsal rootlets associated with lower limbs in a particular case) was 15.8%, 22.3%, 33.4%, 41.8%, and 45.7% across cases with their pro-op GMFCS level from I to V, respectively. Rootlets required 75% cut had a tendency to increase as well in our cases with their pro-op GMFCS level from I to V, which comprising 1.5%, 4.8%, 8.5%, 14.1%, and 15.2% of all rootlets transected, respectively. The muscle tone of 2068 targeted muscles in these cases at the time of 3 weeks after the SL-SDR was revealed a significant decrease when compared to pre-op (1.7 ± 0.5 vs. 2.7 ± 0.6). Further investigation to compare our rootlet selection with the one guided by the traditional rhizotomy criteria using our intro-operative EMG recordings in 318 cases, revealed that the overlap ratio had a tendency to increase in cases when their pre-op GMFCS level increased (39.5%, 41.3%, 52.2%, 54.1%, and 62.8% in cases with levels I-V, respectively). While our modified rhizotomy protocol successfully identified 2-23 rootlets for sectioning in all of our 318 cases, the traditional rhizotomy protocol failed to distinguish any for cutting in about 20% of cases with their pre-op GMFCS levels I and II.
Conclusions: The rhizotomy criteria fully rely on the EMG interpretation making intra-operative neuroelectrophysiological monitoring crucial when SDR is performed via a single-level approach. Our modified rhizotomy protocol is feasible, safe, and effective to guide SL-SDR to treat all types of spastic CP cases by decreasing muscle tone in particular spastic muscle groups in their lower limbs. Data of EMG responses during SL-SDR procedure and as well as the clinical outcomes based on their interpretation could help clinicians to further understand how neuronal circuits work in the spinal cord of these patients.
Keywords: Dorsal rhizotomy; EMG interpretation; Intra-operative neurophysiology; Spastic cerebral palsy.