Pattern-specific effects of botulinum neurotoxin type A injections and selective dorsal rhizotomy on gait in children with spastic cerebral palsy

Abstract

Botulinum neurotoxin type A (BoNT) injections and selective dorsal rhizotomy (SDR) are often applied tone reduction procedures in children with spastic cerebral palsy (CP).[1,2] BoNT is focal and temporary, whereas SDR is generalized and non-reversible. Previous studies have investigated the changes inflicted by these treatments in generic CP-groups.[3,4] It is not yet clear whether specific gait patterns would respond differently to each treatment. What are the short-term, gait pattern-specific changes inflicted by BoNT injections or SDR in children with CP? Retrospective samples that had been treated either BoNT injections (NBoNT=117; baseline ageBoNT= 6y4mo±2y4mo; GMFCS I/II/III: 70/31/16) or SDR (NSDR=89; baseline ageSDR=9y5mo±2y3mo; GMFCS I/II/III: 18/54/17) were selected. All patients underwent three-dimensional gait analysis (3DGA) sessions at baseline and post-treatment (on average 1 y post-SDR and 2mo post-BoNT). The baseline 3DGA was used to classify the gait patterns of the patients, using the gait pattern classification system for children with spastic CP (GaP-CP).[5] For children with bilateral CP, both lower limbs were considered in case of asymmetric patterns between the two lower limbs, Their most affected side was selected when they displayed symmetric gait patterns, similar to the affected lower limb for children with unilateral CP. Gait-related changes focused on sagittal plane kinematics, which were compared with statistical non-parametric mapping (vector of four components, paired Hotellings T2 test, α=0.05 and post-hoc component-level comparisons, paired t-tests, α=0.0125). The comparisons were conducted in the total cohorts, as well as in gait pattern-specific subgroups. Thereafter, statistical clusters were deemed clinically relevant if their duration exceeded 3% of the gait cycle and the respective standard errors of measurement (SEM).[6,7] Changes in neuromuscular impairments were evaluated using the composite spasticity, weakness and selectivity scores of the muscles acting in the sagittal plane,[8] based on the clinical examination. Apparent equinus and jump gait were the best BoNT-responders, followed by dropfoot, where improvements were only observed in the ankle joint. In these three gait patterns, spasticity was improved, but not at the expense of additional weakness or selectivity. For SDR, the best responders were children with jump gait, crouch gait and apparent equinus. Spasticity was improved, while weakness and selectivity either improved or remained stable, in all gait patterns and for the total cohort. Fig. 1 shows the pre- vs post-treatment kinematics and statistically identified clusters of the three best responders to each treatment. "Fig. 1. Pre- vs post-treatment kinematics and statistically identified clusters of the three best responders to each treatment."Download : Download high-res image (251KB)Download : Download full-size image These results highlight the need to inspect the short-term effects of SDR or BoNT injections based on subgroups defined according to the baseline gait patterns and not only in generic groups. Such comprehensive analyses might facilitate optimal patient selection for these treatments.