2.1

Study design and patients

This retrospective study included children receiving treatment at the Timone Children’s Hospital, Marseille, France, between January 2007 and June 2010. Written informed consent was obtained from parents at the time of treatment. According to French law, approval by an ethics committee was not required for the study.

Children with hemiplegia or diplegia attending the hospital underwent a series of examinations before and after abobotulinumtoxinA injection. Eligibility for inclusion was determined from medical histories. In brief, study participants were children < 6 years of age, with a diagnosis of CP by a pediatric neurologist (BC), a Gross Motor Function Classification System (GMFCS) level 1 or 2 (good or very good functional abilities), and equinovalgus (defined as initial contact with the hallux or head of the first metatarsal [M1]) confirmed independently by a physical medicine and rehabilitation physician (CB) and a physical therapist (GA). Patients were excluded if they had a fixed equinus deformity due to a triceps surae contracture (tested passively with the planovalgus corrected) or a limb length discrepancy > 1 cm that could provoke a false equinus by compensation at initial contact. EMG was used to identify premature activity of the PL.

2.2

Treatment

AbobotulinumtoxinA (Dysport ^® , Ipsen, Paris; 6–7 U/kg) was injected into the PL, localized by using stimulo-detection with an intramuscular electrode (Needle Electrodes, Bioject™ DHN 37). The needle was introduced in the lateral side of the leg with the same landmarks used during EMG electrode placement.

During the study period, no patients wore a serial cast or ankle-foot orthotic, but all wore a foot plantar orthotic, which was custom-made by the same orthotic technician, to correct planovalgus (i.e., to mitigate forefoot pronation and hindfoot valgus). Physiotherapy was not prescribed.

2.3

Clinical evaluation

Clinical and video observations were used to determine ankle, and fore-, mid- and hindfoot positions at initial contact and to confirm an equinovalgus diagnosis. A posterior view at initial contact was used to characterize both the position of the hindfoot and the area of initial contact. Clinical examination of the plantar side of the foot for hyperkeratosis of the M1 head and/or hallux provided objective evidence of the preferred M1 and/or hallux placement at initial contact. Two physicians (C.B. and G.A.) performed a clinical bilateral goniometric assessment to assess pre- and post-treatment angles for dorsiflexion of the talocrural articulation; this assessment involved low-speed measurement of end-range resistance with the child in the supine position with knees flexed, then extended, while ensuring no pronation of the forefoot. These steps prevented contracture and/or spasticity of the triceps surae. Lower-limb length (distance from anterior superior iliac spine to medial malleolus) and foot morphology in a standing position were assessed.

The mean number of gait cycles analyzed was 15 ± 10. Because previous authors have demonstrated that muscle activation depends on speed , patients were barefoot during gait analysis so that their walking velocity was spontaneous (rather than at an imposed pace) and thus more easily reproduced .

2.4

EMG evaluation

Children underwent bilateral EMG examination. EMG data were collected by using a Wifi system (Zero Wire, Aurion, Italy). EMG signals were acquired at a rate of 2000 Hz, amplified 1000 times, filtered (10 Hz high-pass filter, 500 Hz low-pass filter) and rectified to zero. This EMG technique integrates a foot-switch system that identifies the anatomic area at initial contact. A force platform was not used.

Before positioning the electrodes, the skin was degreased, stripped and cleaned. Two electrodes (Ag-AgCl, diameter 10 mm) were positioned parallel to the PL fibres one-quarter of the way down the line running from the fibular head to the edge of the lateral malleolus. As recommended in the Surface EMG for the Non-Invasive Assessment of Muscles protocol , electrode centers were placed 10 mm apart and electrode positioning was validated to minimize crosstalk between muscles (gastrocnemius medialis and PL, and PL and tibialis anterior). Validation was achieved by monitoring EMG activity while the patient performed plantar flexion movements with stretched knees (gastrocnemius medialis testing), then eversion by moving the M1 downwards and pronating the forefoot (PL testing), and finally dorsal inversion–flexion (for tibialis anterior testing). The procedure ensured asynchronicity in the bursts of EMG activity during phases of voluntary movement.

Premature muscle activity was determined by EMG as described previously for the PL . In brief, the onset of EMG activity was identified by using a semi-automatic method: the EMG signal was defined as active with rectified raw signal > 20 μV . This threshold is commonly used for studying motor unit potentials and excluding artifacts. The timing of EMG activity in the PL was analyzed and expressed as a percentage of gait cycle (0–100%). Premature onset of PL activity was defined as EMG activity during the swing phase before initial contact (when PL contraction does not occur in children without pes planovalgus) . As a convention, an activation that occurred before the initial contact (0%) was noted as a negative value.

2.5

Radiology evaluation

A set of weight-bearing radiographs, including standard standing lateral, dorso-plantar and anteroposterior ankle radiographs, was obtained for both feet of children before and 4 to 6 months after the last abobotulinumtoxinA injection to determine a therapeutic direct benefit. The following variables were assessed: forefoot pronation (metatarsal stacking angle), midfoot planus (lateral talo-M1 angle and talocalcaneal angle) and hindfoot equinovalgus (anteroposterior ankle angle for hindfoot valgus and calcaneal pitch angle for plantar-flexed hindfoot). Intra- and interobserver variability was assessed and compared with published data .

2.6

Statistical analyses

Paired t -test was used for pairwise comparisons of data from children without pes planovalgus (reference) and for data for angles before and after treatment, except for the anteroposterior ankle angle, for which no reference data are available for pediatric patients. P < 0.05 was considered statistically significant.

2.1

Study design and patients

This retrospective study included children receiving treatment at the Timone Children’s Hospital, Marseille, France, between January 2007 and June 2010. Written informed consent was obtained from parents at the time of treatment. According to French law, approval by an ethics committee was not required for the study.

Children with hemiplegia or diplegia attending the hospital underwent a series of examinations before and after abobotulinumtoxinA injection. Eligibility for inclusion was determined from medical histories. In brief, study participants were children < 6 years of age, with a diagnosis of CP by a pediatric neurologist (BC), a Gross Motor Function Classification System (GMFCS) level 1 or 2 (good or very good functional abilities), and equinovalgus (defined as initial contact with the hallux or head of the first metatarsal [M1]) confirmed independently by a physical medicine and rehabilitation physician (CB) and a physical therapist (GA). Patients were excluded if they had a fixed equinus deformity due to a triceps surae contracture (tested passively with the planovalgus corrected) or a limb length discrepancy > 1 cm that could provoke a false equinus by compensation at initial contact. EMG was used to identify premature activity of the PL.

2.2

Treatment

AbobotulinumtoxinA (Dysport ^® , Ipsen, Paris; 6–7 U/kg) was injected into the PL, localized by using stimulo-detection with an intramuscular electrode (Needle Electrodes, Bioject™ DHN 37). The needle was introduced in the lateral side of the leg with the same landmarks used during EMG electrode placement.

During the study period, no patients wore a serial cast or ankle-foot orthotic, but all wore a foot plantar orthotic, which was custom-made by the same orthotic technician, to correct planovalgus (i.e., to mitigate forefoot pronation and hindfoot valgus). Physiotherapy was not prescribed.

2.3

Clinical evaluation

Clinical and video observations were used to determine ankle, and fore-, mid- and hindfoot positions at initial contact and to confirm an equinovalgus diagnosis. A posterior view at initial contact was used to characterize both the position of the hindfoot and the area of initial contact. Clinical examination of the plantar side of the foot for hyperkeratosis of the M1 head and/or hallux provided objective evidence of the preferred M1 and/or hallux placement at initial contact. Two physicians (C.B. and G.A.) performed a clinical bilateral goniometric assessment to assess pre- and post-treatment angles for dorsiflexion of the talocrural articulation; this assessment involved low-speed measurement of end-range resistance with the child in the supine position with knees flexed, then extended, while ensuring no pronation of the forefoot. These steps prevented contracture and/or spasticity of the triceps surae. Lower-limb length (distance from anterior superior iliac spine to medial malleolus) and foot morphology in a standing position were assessed.

The mean number of gait cycles analyzed was 15 ± 10. Because previous authors have demonstrated that muscle activation depends on speed , patients were barefoot during gait analysis so that their walking velocity was spontaneous (rather than at an imposed pace) and thus more easily reproduced .

2.4

EMG evaluation

Children underwent bilateral EMG examination. EMG data were collected by using a Wifi system (Zero Wire, Aurion, Italy). EMG signals were acquired at a rate of 2000 Hz, amplified 1000 times, filtered (10 Hz high-pass filter, 500 Hz low-pass filter) and rectified to zero. This EMG technique integrates a foot-switch system that identifies the anatomic area at initial contact. A force platform was not used.

Before positioning the electrodes, the skin was degreased, stripped and cleaned. Two electrodes (Ag-AgCl, diameter 10 mm) were positioned parallel to the PL fibres one-quarter of the way down the line running from the fibular head to the edge of the lateral malleolus. As recommended in the Surface EMG for the Non-Invasive Assessment of Muscles protocol , electrode centers were placed 10 mm apart and electrode positioning was validated to minimize crosstalk between muscles (gastrocnemius medialis and PL, and PL and tibialis anterior). Validation was achieved by monitoring EMG activity while the patient performed plantar flexion movements with stretched knees (gastrocnemius medialis testing), then eversion by moving the M1 downwards and pronating the forefoot (PL testing), and finally dorsal inversion–flexion (for tibialis anterior testing). The procedure ensured asynchronicity in the bursts of EMG activity during phases of voluntary movement.

Premature muscle activity was determined by EMG as described previously for the PL . In brief, the onset of EMG activity was identified by using a semi-automatic method: the EMG signal was defined as active with rectified raw signal > 20 μV . This threshold is commonly used for studying motor unit potentials and excluding artifacts. The timing of EMG activity in the PL was analyzed and expressed as a percentage of gait cycle (0–100%). Premature onset of PL activity was defined as EMG activity during the swing phase before initial contact (when PL contraction does not occur in children without pes planovalgus) . As a convention, an activation that occurred before the initial contact (0%) was noted as a negative value.

2.5

Radiology evaluation

A set of weight-bearing radiographs, including standard standing lateral, dorso-plantar and anteroposterior ankle radiographs, was obtained for both feet of children before and 4 to 6 months after the last abobotulinumtoxinA injection to determine a therapeutic direct benefit. The following variables were assessed: forefoot pronation (metatarsal stacking angle), midfoot planus (lateral talo-M1 angle and talocalcaneal angle) and hindfoot equinovalgus (anteroposterior ankle angle for hindfoot valgus and calcaneal pitch angle for plantar-flexed hindfoot). Intra- and interobserver variability was assessed and compared with published data .

2.6

Statistical analyses

Paired t -test was used for pairwise comparisons of data from children without pes planovalgus (reference) and for data for angles before and after treatment, except for the anteroposterior ankle angle, for which no reference data are available for pediatric patients. P < 0.05 was considered statistically significant.