- Email: [email protected]
Effect of botulinum toxin injection on length and lengthening velocity of rectus femoris during gait in hemiparetic patients
Clinical Biomechanics 28 (2013) 164–170
Contents lists available at SciVerse ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Effect of botulinum toxin injection on length and lengthening velocity of rectus femoris during gait in hemiparetic patients N. Lampire a, b,⁎, N. Roche c, P. Carne a, L. Cheze b, D. Pradon c a b c
Laboratoire d'analyse du mouvement, CMPR L'ADAPT Loiret, Amilly, France Laboratoire de Biomécanique et Mécanique des Chocs, Université de Lyon, UMR_T 9406, Université Lyon 1, IFSTTAR, Villeurbanne, France EA 4497 GRCTH, CIC-IT, CHU Raymond Poincaré, UVSQ, Garches, France
a r t i c l e
i n f o
Article history: Received 23 August 2012 Accepted 10 December 2012 Keywords: Gait Stroke Spasticity Rectus femoris Botulinum toxin
a b s t r a c t Background: In hemiparetic patients, rectus femoris spasticity is one of the main causes of reduced knee flexion in swing phase, known as stiff knee gait. Botulinum toxin is often used to reduce rectus femoris spasticity and to increase knee flexion during swing phase. However, the mechanisms behind these improvements remain poorly understood. The aim of this study was (1) to quantify maximal rectus femoris length and lengthening velocity during gait in ten adult hemiparetic subjects with rectus femoris spasticity and stiff knee gait and to compare these parameters with those of ten healthy subjects and (2) to study the effect of botulinum toxin injection in the rectus femoris muscle on the same parameters. Methods: 10 patients with stiff knee gait and rectus femoris spasticity underwent 3D gait analysis before and one month after botulinum toxin injection of the rectus femoris (200 U Botox®, Allergan Inc., Markham, Ontario, CANADA). Rectus femoris length and lengthening velocity were quantified using a musculoskeletal model (SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA). Findings: Maximal length and lengthening velocity of the rectus femoris were significantly reduced on the paretic side. There was a significant increase in muscle length as well as lengthening velocity during gait following botulinum toxin injection. Interpretation: This study showed that botulinum toxin injection in the spastic rectus femoris of hemiparetic patients improves muscle kinematics during gait. However maximal rectus femoris length did not reach normal values following injection, suggesting that other mechanisms are likely involved. © 2013 Elsevier Ltd. All rights reserved.
1. Introduction Following stroke, the characteristics of gait are altered in comparison with healthy subjects. Aside from changes in spatiotemporal and postural parameters, there are also alterations in lower limb kinematics such as reduced peak knee flexion during swing phase of the gait cycle, known as stiff knee gait (SKG) (Kerrigan et al., 2001). One of the main causes of SKG is inappropriate electromyographic activity (EMG) of the rectus femoris (RF) muscle, related to spasticity of this muscle (Goldberg et al., 2004). Other causes are reduced peak hip flexion and reduced peak plantar flexion moment (Kerrigan et al., 1999). One of the classic treatments for RF spasticity associated with abnormal EMG activity is botulinum toxin injection (BTI). This neuro-toxin works by blocking neurotransmitter release at the neuromuscular junction. In consequence this treatment induces a paresis
Abbreviations: RF, Rectus Femoris muscle; SKG, Stiff Knee Gait; BTI, Botulinum Toxin Injection; EMG, Electromyographic; SD, Standard Deviation. ⁎ Corresponding author at: Laboratoire d'analyse du mouvement, CMPR L'ADAPT Loiret, 658D rue des Bourgoins, 45200 Amilly, France. E-mail address: [email protected] (N. Lampire). 0268-0033/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinbiomech.2012.12.006
of the injected muscle, which is ‘voluntary, reversible and transient’ (Poulain and Humeau, 2003). Several studies in hemiparetic subjects showed that BTI in the spastic RF muscle significantly increased peak knee flexion during swing phase as well as knee flexion velocity at toe off (Hutin et al., 2010; Robertson et al., 2009; Stoquart et al., 2008). However, the mechanisms behind these kinematic improvements are as yet poorly understood. It is largely supposed that the focal paralysis which occurs following BTI leads to an increase in maximal muscle length and the muscle's capacity to lengthen during gait. The study of these parameters may help to understand the mechanisms associated with spasticity and the effect of botulinum toxin on spastic muscles. This hypothesis is based on the results of a study by Jonkers et al. (2006) in which a musculoskeletal model (SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA) was used to quantify RF resting length as well as its length and variations during gait in 35 children with spastic diplegia. The results showed that maximal length and maximal lengthening velocity of the RF were both reduced during gait compared with those of healthy subjects. The SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA musculoskeletal model has two useful properties: the dynamic module can be used
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
to perform forward and inverse dynamic simulations on musculoskeletal models. The forward simulation is used for the calculation of the motion and the contact forces resulting from the specified muscle excitation. In contrast, the inverse simulation allows calculation of the muscle activations and forces required to generate a specific motion. The forward model was developed for the prediction of the biomechanical consequences of the modifications of model parameters following a surgical procedure or change in muscle strength (Goldberg et al., 2004; McLean et al., 2004; Neptune et al., 2001; Yamaguchi and Zajac, 1990). The inverse simulation is used for the evaluation of changes in gait resulting from pathology or treatment based on data obtained during 3D gait analysis of a patient (Crowninshield and Brand, 1981; Glitsch and Baumann, 1997; Li et al., 1999). This model is the only tool which allows indirect assessment of muscle strength or lengthening during movement such as gait. Nevertheless it is necessary to keep in mind that it is an indirect approach and that the data obtained result from a simulation. In consequence, the results should be interpreted with caution. However when the inverse simulation is carried out on data from the same subject, the inherent bias of this approach is reduced, therefore, allowing comparison of the modifications induced by a treatment (whatever the type) on muscle strength or length. The present study was based on this method and had two aims. The first was to assess maximal RF length and lengthening velocity during gait in adult hemiparetic subjects with RF spasticity and SKG and to compare these parameters with those of the healthy subjects. The second aim was to evaluate the effect of BTI in the RF muscle on the same parameters. 2. Methods 2.1. Subjects 10 hemiparetic subjects constituted the patient group (Table 1): 4 with right hemiparesis, 6 with left, 8 men, 2 women, average age: 39.6, SD: 9.5 years. Inclusion criteria were the following: a single cerebral lesion of vascular origin (ischemic or hemorrhagic) of more than 6 month duration or more than 4 months since last BTI, a decrease in peak knee flexion during swing phase measured using gait
Table 1 Demographic characteristics of the patients.
Subject 1 Subject 2 Subject 3 Subject 4 Subject 5 Subject 6 Subject 7 Subject 8 Subject 9 Subject 10
Age Height Weight Gender Paretic Time (years) (cm) (kg) side since lesion onset (months)
RF spasticity assessed by modified Ashworth scale before and 1 month after BTI
27
180
74
M
L
48
–
40
188
73
M
L
6
0/0
58
160
69
M
R
60
1/1
29
166
72
M
L
60
3/2
33
180
85
M
L
36
–
37
182
82
M
R
60
–
40
181
72
M
R
36
2/2
40
163
51
F
L
84
1/0
52
162
67
F
R
432
2/1
40
173
76
M
L
96
2/1
M = male, F = female, L = left and R = right, RF = rectus femoris, BTI = botulinum toxin injection.
165
analysis, prolonged RF activity during swing phase determined by EMG and ability to walk 20 min without stopping. Patients were excluded if they had any comorbid disability other than stroke, such as any visual impairment or musculoskeletal, cardiovascular, or pharmacological treatment that could interfere with gait or posture. All patients had previously been treated by BTI in the RF muscle before inclusion in this study. 10 healthy subjects (with no central nervous system lesions; 6 men and 4 women; average age: 32.5, SD: 6 years) were also included as a control group. All subjects gave informed consent prior to inclusion. This study was carried out according to “The Ethical Codes of the World Medical Association” (Declaration of Helsinki) and was approved by the local ethical committee (CPP Ile de France, Ambroise Paré). 2.2. Protocol 2.2.1. Hemiparetic patients BT was injected in the spastic RF muscle on the hemiparetic side of all patients included. 200 U of type A botulinum toxin (Botox®, Allergan Inc., Markham, Ontario, CANADA) diluted to 40 U/ml was injected in three anatomical points in the RF muscle using electrostimulation (5 mA) by the same investigator. The technique used was the same to that previously used in Hutin et al. (2010): one point 5 cm under the inguinal crease on a line between the antero-inferior iliac spine and the patella. The second point was ten centimeters above the patella and the third was in the middle of the two others on the same line. The dose used was the same as that in other studies evaluating the impact of BTI in the RF muscle on peak knee flexion in swing phase (Hutin et al., 2010; Robertson et al., 2009; Stoquart et al., 2008). 2.3. Assessments This was an observational study. Two assessments were carried out: before injection (PRE) and 1 month post injection (POST). The second assessment was carried out one month post injection because this is when the toxin is at maximal effectiveness (Juzans et al., 1996). A clinical assessment and gait analysis were carried out at each assessment. • Clinical assessment Sagittal hip, knee and ankle joint ranges of motion were assessed in order to check for muscle contractures. Spasticity of the quadriceps and RF was evaluated using the modified Ashworth scale. All clinical tests were carried out by the same physiotherapist. • Gait analysis Each patient underwent three dimensional (3D) gait analysis using a Motion Analysis System (Motion Analysis Corporation, Santa Rosa, CA, USA, sampling frequency 100 Hz). Markers were positioned according to the Helen Hays protocol (Kadaba et al., 1990). Subjects walked barefoot at their own comfortable pace. Ten gait cycles, corresponding to those during which the patient walked on the force plate, resulting from ten gait trials were recorded and averaged for each patient. Patients were offered a break after five trials. EMG was recorded simultaneously during the kinematic data collection using bipolar surface electrodes (MA 311) placed on the rectus femoris, vastus lateralis, semimembranosus, tibialis anterior, gastrocnemius medialis and soleus muscles. Vastus lateralis activity was assumed to be representative of the three vastii (Nene et al., 1999). Digitalized EMG, Motion Lab System, LA, USA (1000 Hz) was analyzed to determine the periods of activity of each muscle during gait, and to include them in the model. Moreover, the EMG signal of the RF and VL muscles was used to qualitatively determine if their activation patterns were abnormal during swing phase or not.
166
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
2.4. Data analysis 2.4.1. Joint kinematics A 6 Hz low pass Butterworth filter was used in the signal processing (Winter et al., 1974). Kinematics of the pelvis, hips, knees and ankles were calculated using Orthotrak 6.2.8 (Motion Analysis Corporation, Santa Rosa, CA, USA). This software uses Euler rotations (Grood & Suntay method) to calculate kinematic parameters. Knee flexion/extension angular velocity at toe off was calculated by deriving the displacement data of the tibia relative to the femur. 2.4.2. Muscle kinematics SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA modeling software (Delp et al., 1990) was used to calculate the length of the musculo–tendinous complex of the RF. The standard generic SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA model was used to scale the model from the gait analysis data. This model consists of 13 rigid segments (pelvis, left and right femurs, left and right tibias, left and right patellae, left and right heels, left and right feet and left and right toes) with 17 degrees of freedom. In this model, each lower limb consists of 43 musculo–tendinous complexes. Each complex is defined by its origin, insertion, and if necessary, by defined via points in order to specify the muscle trajectory as much as possible. Particular attention was paid during the collection of anthropometric data for each subject (age, weight, height, foot length, ankle and knee diameters) in order to optimize the model scaling. The RF length was quantified during the gait cycle and when the subject was virtually placed in the anatomical reference position (defined as an angle of 0° for the hip and knee joints). Three muscles parameters were studied: • Normalized maximal RF length (maxNL): defined as maximal RF length during gait minus RF length in the ‘virtual’ anatomical reference position, in millimeters. This normalization reduces bias related to morphological differences between subjects, as well as differences related to marker placement between PRE and POST recordings for each patient. • Maximal RF lengthening velocity (maxLV): calculated for each gait cycle; for example MaxLV=Peak4 in Fig. 1. The average maximal lengthening velocity was then calculated for each subject. • Mean lengthening velocity at lengthening peaks (meanLV): defined as the mean of the positive peak lengthening velocity during one gait cycle. The mean velocity was then calculated for each subject. The velocity peaks were defined by zero acceleration. For example: MeanLV of cycle 1=(Peak1+Peak2+Peak3+Peak4)/4 (Fig. 1). Four peaks are present in this example. When the curve included more peaks, the average of all of the peaks was used.
These parameters were calculated in the PRE and POST conditions for the hemiparetic patients as well as for the healthy subjects. Because joint kinematics have been shown to be influenced by gait velocity (Schwartz et al., 2008), slow gait was recorded for the healthy subjects (cadence of 69 steps/min guided by a metronome) so that the average gait velocity of the healthy group was similar to that of the patient group. 2.5. Statistical analysis A Mann–Whitney test was used to compare age, height, weight and RF length in the anatomical reference position of the two groups. A Wilcoxon test was used to compare differences in maximal normalized RF length, maxLV and meanLV during gait, between the PRE and POST conditions in the patient group and the control group. P b 0.05 was considered significant in all cases. Correlations between (i) the increase in knee flexion during swing phase and the increase in gait velocity, (ii) the increase in knee flexion angular velocity at toe off and the increase in peak knee flexion during swing phase, maxLV and meanLV, (iii) the increase in maxNL and the increase in peak knee flexion during swing phase, and (iv) the increase of maxNL and maxLV and the increase in gait speed after BTI were assessed using a Spearman correlation. Statistical analyses were carried out using Statistica 7.0 software StatSoft, Inc., OK, USA. 3. Results 3.1. Anthropometrical comparison between patients and healthy subjects There was no statistical difference for age, height and weight between hemiplegic patients and healthy subjects (P b 0.05). 3.2. Clinical evaluation Seven of the 10 subjects underwent clinical examinations. The remaining three were outpatients and only able to complete the gait analysis. RF spasticity ranged from 1 to 3 in PRE on the modified Ashworth scale and was decreased one month after BTI (range 1–2). 3.3. Spatiotemporal The spatiotemporal parameters analyzed for the control group and for the patient group in the PRE and POST conditions are shown in Table 2. There was a significant mean increase of 0.13 m/s (24%) in gait velocity (P b 0.05) in the POST condition for the patient group. There was no significant difference in gait velocity between the patient group and the healthy subjects in the PRE or POST conditions (P > 0.05). 3.4. Kinematics The kinematic parameters analyzed for the control group and for the patient group in the PRE and POST conditions are shown in Table 2. There was a significant increase in peak knee flexion of 7.5° (+27%) in the POST condition for the patient group (P b 0.05). There was also a significant increase of 53°/s (+43%) in knee flexion velocity at toe off (P b 0.05). Despite the significant increase in peak knee flexion and knee flexion velocity at toe off, these parameters remained significantly lower than those of the healthy subjects (P b 0.05). 3.5. Biomechanical parameters of the RF
Fig. 1. Lengthening velocity of the rectus femoris during one gait cycle of a hemiparetic patient. The ellipses highlight the lengthening velocity peaks. The square highlights the maxLV peak.
The biomechanical parameters of the RF analyzed for the control group and for the patient group in the PRE and POST conditions are shown in Table 2 and Figs. 2–3. On the paretic side, maximal normalized length of the RF (RF maximal length during gait−RF length in virtual anatomical position) was 17 mm
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
167
Table 2 Mean values of the parameters studied. Paretic PRE Gait velocity (m/s) Peak knee flexion (°) Knee flexion angular velocity at toe off (°/s) MaxNL (mm) MaxLV (m/s) MeanLV (m/s)
0.57 27.3 122.5 17 0.07 0.03
(SD 0.24)⁎ (SD 7.3)⁎,⁎⁎⁎ (SD 68.3)⁎,⁎⁎,⁎⁎⁎ (SD 9)⁎,⁎⁎,⁎⁎⁎ (SD 0.03)⁎,⁎⁎,⁎⁎⁎ (SD 0.02)⁎,⁎⁎,⁎⁎⁎
Paretic POST 0.70 34.8 175.7 22 0.09 0.05
Non-paretic PRE
(SD 0.28)⁎ (SD 7.1)⁎,⁎⁎⁎ (SD 50.1)⁎,⁎⁎,⁎⁎⁎ (SD 8)⁎,⁎⁎,⁎⁎⁎ (SD 0.03)⁎,⁎⁎,⁎⁎⁎ (SD 0.03)⁎,⁎⁎
0.57 61 280.2 38 0.12 0.07
(SD 0.24)⁎ (SD 5.8)⁎,⁎⁎⁎ (SD 54.1)⁎,⁎⁎,⁎⁎⁎ (SD 9)⁎⁎ (SD 0.03)⁎,⁎⁎ (SD 0.02)⁎,⁎⁎
Non-paretic POST 0.70 63.4 305.4 39 0.13 0.09
(SD 0.28)⁎ (SD 4.5)⁎,⁎⁎⁎ (SD 58.9)⁎,⁎⁎,⁎⁎⁎ (SD 7)⁎⁎ (SD 0.04)⁎,⁎⁎ (SD 0.03)⁎,⁎⁎
Healthy subjects 0.56 48 197.2 41 0.13 0.06
(SD 0.05) (SD 5.8) (SD 22.2) (SD 4) (SD 0.02) (SD 0.01)
PRE = before botulinum toxin injection; POST = post botulinum toxin injection; MaxNL = normalized maximal RF length; MaxLV = maximal RF lengthening velocity; MeanLV = mean lengthening velocity at lengthening peaks. ⁎ Denotes a statistically significant difference between PRE and POST conditions. P ≤ 0.05 in each case. ⁎⁎ Denotes a statistically significant difference between the paretic and non-paretic sides. P ≤0.05 in each case. ⁎⁎⁎ Denotes a statistically significant difference between hemiparetic group GH and healthy group. P ≤ 0.05 in each case.
(SD 9) in the PRE condition and 22 mm (SD 8) in the POST condition. This increase of 5 mm (30%) following BTI was significant (Pb 0.05). This increase in maximal length in the POST condition was significantly correlated (R=0.85) with a significant increase (Pb 0.05) in peak knee flexion at the moment of the maximal peak normalized RF length (PRE: 23.7°, SD 8.4°; POST: 30.2°, SD 9.2°). There was no related change in flexion/extension angle of the hip at the same instant of the gait cycle (PRE=6.3°, SD 11°, POST=7.9°, SD 10°, P>0.05). Despite the significant increase in normalized RF length on the paretic side in the POST condition, it remained significantly lower than the non-paretic length and healthy subjects' values (Pb 0.05). On the non-paretic side, maxNL was 38 mm (SD 9) in the PRE condition. It was increased by 1 mm (4%) in the POST condition to 39 mm (SD 7). This difference was not significant (P>0.05). There was no significant difference in hip or knee flexion/extension angles between the PRE and POST conditions at maximal peak normalized RF length of the same side (hip PRE=7.6°, SD 10.2°, hip POST 6.9°, SD 7.7°, P>0.05; knee PRE= 51.6°, SD 6.4°, knee POST=54.4°, SD 4.6°, P>0.05). There were no significant differences for these parameters between the patient group and healthy group in either the PRE or POST condition (P>0.05). On the paretic side, in the PRE condition, maxLV of the RF was 0.07 m/s (SD 0.03) and meanLV was 0.03 m/s (SD 0.02). In the POST
condition, the maxLV was 0.09 m/s (SD 0.03) and was significantly increased (+28%, Pb 0.05). MeanLV was 0.05 m/s (SD 0.03) and was also significantly increased (+66%, Pb 0.05). Comparison with the healthy group showed that maximal velocity and mean velocity at lengthening peaks in the PRE condition were significantly lower in the patient group (P b 0.05). In the POST condition, only the maxLV remained significantly different from that of the healthy subjects (P b 0.05). On the non-paretic side, maxLV of the non-injected RF in the PRE condition was 0.12 m/s (SD 0.03) and mean velocity of the lengthening peaks was 0.07 m/s (SD 0.02). In the POST condition, there was a significant increase in maxLV of 8% to 0.13 m/s (SD 0.04) (P b 0.05) and a significant increase in meanLV of 28% to 0.09 m/s (SD 0.03) (P b 0.05). There were no significant differences between the patient and healthy groups for these parameters in the PRE or POST conditions. 3.6. Correlations There was a significant correlation between the increase in maxNL and the increase in peak knee flexion during swing phase (R = 0.86, P b 0.05) in both the PRE and POST conditions. There was a significant correlation between the increase in knee flexion during swing phase and the increase in gait velocity (R = 0.79, P b 0.05). There was no
Fig. 2. Normalized rectus femoris length, and hip and knee kinematics for patient No. 9. GS = healthy subjects, PRE = before botulinum toxin injection; POST = post botulinum toxin injection, RF = rectus femoris.
168
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
Fig. 3. Mean lengthening velocity curve of rectus femoris. GS = healthy subjects, GH PRE = hemiparetic patients before botulinum toxin injection; POST = hemiparetic patients post botulinum toxin injection.
correlation between the increase in max NL and maxLV and the increase in gait speed (P > 0.05). There was also a significant correlation between the increase in knee flexion angular velocity at toe off and the increase in peak knee flexion during swing phase (R = 0.67, P b 0.05). Moreover, knee flexion angular velocity at toe off was correlated with maxLV in the PRE (R = 0.95, P b 0.05) and POST (R = 0.87, P b 0.05) conditions, and with meanLV in the PRE (R = 0.80, P b 0.05) and POST (R = 0.85, P b 0.05) conditions. 4. Discussion The aim of this study was to quantify maximal RF length and lengthening velocity during gait in hemiparetic patients with SKG and RF spasticity, before and after BTI in this muscle. The results showed that maxNL of RF during gait was reduced on the hemiparetic side in comparison with the non-paretic side and with healthy subjects walking at a similar gait velocity. Furthermore the results: i) confirm that BTI in the spastic RF significantly increases peak knee flexion in swing phase as well as gait velocity in hemiparetic patients (Hutin et al., 2010; Robertson et al., 2009; Stoquart et al., 2008); and ii) show that BTI increases maxNL of the injected muscle during movement although the values remain lower than those of the non-paretic side and healthy subjects. Before discussing the results of this study, several points should be discussed: Concerning the gender of the patients included, the patient group was predominantly composed of men (8 men, 2 women) whereas in the control group there were 6 men and 4 women. To our knowledge, there is no data in the literature suggesting that the gait pattern of men differs from those of women either in healthy subjects or in hemiplegic patients. It seems therefore unlikely that this difference in gender could constitute a limitation in the interpretation of the result. Concerning the “heterogeneity” of the patients included and more particularly the time since the lesion onset, we do not feel that these differences constitute a limitation in the interpretation of the results. Indeed, all the patients had moderate spasticity of the rectus femoris muscle, which was independent of the time since the lesion onset.
Moreover all had the same proportion of improvement in the different kinematic muscle parameters studied after BTI injection. It therefore seems that the heterogeneity of the time since the lesion onset does not constitute a bias but on the contrary means that the results of this study are applicable to a large population of hemiplegic patients. Concerning the anthropometric models: there was no anthropometric difference between the two groups assessed. Moreover, as far as we are aware, there is no specific model for men or women or for patients and healthy subjects. Thus similar to Jonkers et al. (2006), we used the same anthropometric model for the two groups. We cannot exclude that this methodological point might constitute a limitation in the interpretation of the results but no other solution was, to our knowledge, available. The same investigator performed all the anthropometric measures in order to limit the bias inherent to this approach. For the EMG signal analysis, similar to Jonkers et al. (2006) and Robertson et al. (2009), we used a qualitative approach to determine if the RF muscle activation pattern during swing phase was abnormal. An abnormal RF pattern is characterized by inappropriate EMG activity in the middle of the swing phase. In contrast, the quantitative EMG of all muscles recorded was integrated in the musculoskeletal model to optimize the results. Concerning the results obtained with the SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA model. There were no significant differences between RF lengths in the anatomical reference position in the PRE and POST conditions in the patient group. Moreover the maxNL of the non-paretic side was not modified following BTI. This suggests that the results obtained for the paretic limb can be interpreted with confidence. Concerning the gait velocity, the improvement following BTI was significant. However, although the gait velocity of the patient group was increased after BTI, it was not significantly different from that of the healthy subjects. This lack of difference is likely to be due to the difference in the SD between the two groups (paretic and healthy groups). The standard deviation is higher for the group of the patients. The results of this study showed firstly that maxNL of the paretic side was significantly lower than that of the non-paretic side as well as that of the healthy subjects walking at a similar velocity. These results conform to those of Jonkers et al. (2006) who found a decrease in maximal RF length in 35 diplegic children with reduced knee flexion during swing phase.
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
The first question that can be addressed is: What are the mechanisms involved in the decrease in spastic RF length in hemiparetic patients? Spasticity has been defined as a reflex muscle contraction in response to stretch of the same muscle (Lance, 1980). All the patients included in this study had spasticity of the RF muscle (Table 1). Spasticity can lead to abnormal muscle stiffness, which in turn could reduce the maximal normalized length of the muscle and its general capacity to lengthen during movement. Our results support this hypothesis. Indeed, the results show that 1 month after BTI in the paretic RF, its maximal normalized length was significantly increased by about 30%. These results are similar to those of Eames et al. (1997) who found that increased gastrocnemius length following BTI was significantly inversely correlated with clinically evaluated spasticity of the same muscle in 39 children with cerebral palsy. They showed that the lower the degree of spasticity before injection, the greater the increase in muscle length after injection. It is interesting to note that in the population of hemiparetic patients included, there was i) no significant difference between the maxNL of the non-paretic RF and that of the healthy subjects in the PRE or POST conditions; and ii) no change in non-paretic maxNL in the POST condition. This further confirms the hypothesis that the increased maxNL is principally due to the decrease in spasticity following BTI and only occurs in the injected muscle. The second question that can be addressed is: What are the main causes of decreased peak knee flexion in swing phase? Jonkers et al. (2006) highlighted the fact that in order to interpret the impact of a change in muscle length, it is necessary to take into account the kinematics of the joints on which the muscle acts. The RF is a bi-articular muscle. It originates in the antero-inferior iliac spine and inserts into the upper border of the patella. In order to interpret changes in RF length, kinematics of the hip and knee joints in the sagittal plane must therefore be taken into account. Our results suggest that the hip flexion angle at the time of maximal length of the injected RF was not significantly different between PRE and POST conditions. There was, however, a significant increase in knee flexion of 7.5° at this instant. This therefore suggests that the increase in maxNL during gait following BTI results in an increase in peak knee flexion. This result is confirmed by the significant correlation between the increase in maxNL and peak knee flexion in swing phase (R= 0.86) and is in accordance with the results of previous studies (Robertson et al., 2009; Stoquart et al., 2008) which showed that injection of equivalent doses of botulinum toxin led to an increase in peak knee flexion during swing phase of respectively +5° and +8°. Although the injection of botulinum toxin in the RF muscle led to a significant increase in maxNL and an increase in peak knee flexion, these parameters remain significantly reduced in comparison with the non-paretic side and the healthy group. An insufficient dose of botulinum toxin may explain the lack of knee flexion that persists; however the injected dose is in agreement with the French Agency for Safety of Health Products and is similar to that used in other studies which assessed the effects of BTI in RF on peak knee flexion in hemiplegic patients (Hutin et al., 2010; Robertson et al., 2009; Stoquart et al., 2008). This therefore suggests that other factors could underlie the lack of knee flexion during swing. Kerrigan et al. (2001) showed that the lack of knee flexion in hemiparetic patients could be explained by several mechanisms: inappropriate activity of another head of the quadriceps, reduced hip flexion or reduced peak plantar flexion moment. In our study, only 1 patient also had abnormal activity of another head of the quadriceps determined by EMG during swing phase of the gait cycle. However, mean peak hip flexion was significantly reduced on the paretic side compared with the non-paretic side and also with the healthy group. This suggests that in the group of patients included, who all had spasticity of the RF and reduced peak knee flexion during swing phase, the lack of knee flexion is multi-factorial. It is at least related to a combination of abnormal RF EMG activity in swing phase of the gait cycle and a lack
169
of hip flexion, which may be caused by weak hip flexors. The clinical implications of these results are that the management of patients with reduced knee flexion and inappropriate RF activity should include strengthening of the hip flexor muscles as well as BTI in the RF muscle. In a literature review on spasticity and muscle hypertonia, Dietz and Sinkjaer (2007) cite many studies which demonstrate changes in intrinsic muscle fiber length, in the conjunctive tissue of muscles (collagen, fatty tissue) and tendons, and of the fiber type (slow and fast) of spastic muscles. According to the authors, these changes in muscle properties lead to muscle stiffness which in turn causes i) decreased maximal length of the spastic muscle, ii) a large decrease in lengthening capacity and iii) decreased range of movement of the associated joint. These results can be related to the definition of spasticity proposed by Crenna (1998) in which spastic hypertonia, classically described as an increase in the stretch reflex, is associated with increased mechanical muscle resistance. The third question that can be discussed is: Are the results influenced by the improvement of the gait speed? In healthy subjects, it has been shown that gait speed influences kinematic gait parameters. Lelas et al. (2003) showed that gait speed was correlated with an increase in peak knee flexion during loading response, peak knee flexion during swing phase and peak hip and knee extension. Recently Hutin et al. (2012) studied the impact of gait speed on kinematic gait parameters in stroke patients. They showed that increasing gait velocity is associated with an increase in knee flexion velocity at toe off and an increase in peak knee flexion during swing phase of the gait cycle. It could be thus hypothesized, in stroke patients with stiff knee gait and spasticity of the RF, that the increase in length and lengthening capacity found after BTI is not directly due to the paresis of the muscle injected but might also be due to the increase in gait speed. To answer the question we analyzed the correlation between the percentage of improvement of RF length or RF maxLV induced by BTI and the percentage of improvement of the gait speed. Our results indicate that there was no correlation between these two parameters suggesting that the modifications of these two kinematic muscle parameters were mainly due to the focal botulinum toxin action. The last question that needs to be discussed is: Is the action of botulinum toxin limited to the neuromuscular junction? It is interesting to note that in the PRE condition, the maxLV of the patient group was significantly lower on the paretic side than on the non-paretic side and in the healthy group. Similarly, the meanLV of the RF was significantly lower on the paretic side than on the non-paretic side and in the healthy group. These results indicate that maxLV and meanLV (which induces a spastic response) are reduced in spastic RF muscles compared with non-spastic muscles. MaxLV and meanLV were also significantly correlated with the reduced knee flexion angular velocity at toe off (R = 0.95). These results could be explained by the increased excitability of the monosynaptic reflex arc (muscle spindle – Ia fibre – α motoneurone) following a central nervous system lesion or by a rectus femoris contracture or by a weakness of the triceps surae. Nevertheless, after BTI, in the spastic RF muscle, there were i) a significant increase in maxLV (+ 28%) and ii) a significant increase in meanLV (+ 66%) of the injected RF. The increase in maxLV and meanLV of rectus femoris after BTI indicates that BTI, which decreases the excitability of the monosynaptic reflex arc, is responsible for the improvement of these parameters. In addition, any rehabilitation which the patient followed was not changed during the study period, both in terms of frequency and intensity. The changes which occurred in the muscles do not appear to be related to changes which occur following training. It therefore appears that the reduction of knee flexion is related to a decrease in maxLV and meanLV, and therefore with hyperexcitability of the monosynaptic reflex arc. These results suggest that as well as blocking the neuromuscular junction, botulinum toxin could also modify the threshold of the stretch reflex.
170
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
During movement, the threshold of the stretch reflex is constantly regulated by the γ motoneuron which controls the sensitivity of the muscle spindle. Filippi et al. (1993) and Rosales et al. (1996) respectively showed in animal models that synaptic transmission between the γ motoneuron and the muscle spindle is reduced after BTI. The fact that mean LV of the RF (which induces the spastic response) is significantly increased after BTI is in accordance with this hypothesis and suggests that in humans, the toxin could also increase the stretch reflex threshold. This hypothesis is also indirectly in agreement with the results of Hutin et al. (2011) which showed in 15 hemiparetic patients that BTI of the RF increases the fluidity of movement during gait demonstrated by a reduction in the number of continuous relative phase reversals during swing phase. To conclude, the results of this study suggest that, in hemiparetic patients with reduced knee flexion in swing phase and RF spasticity, the length of the RF muscle is decreased. The study also demonstrated that BTI increased maximal RF length as well as its lengthening capacity, although these parameters did not reach normal values. Further studies will be necessary to determine the neuro-physiological mechanisms responsible for these biomechanical changes in patients with central nervous system lesions. Conflict of interest We have no conflict of interest with this work. References Crenna, P., 1998. Spasticity and ‘spastic’ gait in children with cerebral palsy. Neurosci. Biobehav. Rev. 22 (4), 571–578. Crowninshield, R., Brand, R., 1981. A physiological based criterion of muscle force prediction in locomotion. J. Biomech. 14, 793–801. Delp, S.L., Loan, P., Hoy, M.G., Zajac, F.E., Topp, E.L., Rosen, J.M., 1990. An interactive graphics-based model of the lower extremity to study orthopaedic surgical procedures. IEEE Trans. Biomed. Eng. 37 (8), 757–767. Dietz, V., Sinkjaer, T., 2007. Spastic movement disorder: impaired reflex function and altered muscle mechanics. Lancet Neurol. 6 (8), 725–733. Eames, N., Baker, R., Cosgrove, A., 1997. The effect of botulinum injections on gastrocnemius muscle length. Gait & Posture 5 (2), 138. Filippi, G.M., Errico, P., Santarelli, R., Bagolini, B., Manni, E., 1993. Botulinum A toxin effects on rat jaw muscle spindles. Acta Otolaryngol. 113, 400–404. Glitsch, U., Baumann, W., 1997. The three-dimensional determination of internal loads in the lower extremity. J. Biomech. 30 (11/12), 1123–1131. Goldberg, S.R., Anderson, F.C., Pandy, M.G., Delp, S.L., 2004. Muscles that influence knee flexion velocity in double support: implications for stiff-knee gait. J. Biomech. 37, 1189–1196.
Hutin, E., Pradon, D., Barbier, F., Gracies, J.M., Bussel, B., Roche, N., 2010. Lower limb coordination in hemiparetic subjects: impact of botulinum toxin injections into rectus femoris. Neurorehabil. Neural Repair 24 (5), 442–449. Hutin, E., Pradon, D., Barbier, F., Gracies, J.M., Bussel, B., Roche, N., 2011. Lower limb coordination patterns in hemiparetic gait: factors of knee flexion impairment. Clin. Biomech. 26, 304–311. Hutin, E., Pradon, D., Barbier, F., Bussel, B., Gracies, J.M., Roche, N., 2012. Walking velocity and lower limb coordination in hemiparesis. Gait & Posture 36 (2), 205–211. Jonkers, I., Stewart, C., Desloovere, K., Molenaers, G., Spaepen, A., 2006. Musculo-tendon length and lengthening velocity of rectus femoris in stiff knee gait. Gait & Posture 23, 222–229. Juzans, P., Comella, J.X., Molgo, J., Faille, L., Angaut-Petit, D., 1996. Nerve terminal sprouting in botulinum type-A treated mouse levator auris longus muscle. Neuromuscul. Disord. 6 (3), 177–185. Kadaba, M.P., Ramakrishnan, H.K., Wootten, M.E., 1990. Measurement of lower extremity kinematics during level gait. J. Orthop. Res. 8, 383–392. Kerrigan, D.C., Bang, M.S., Burke, D.T., 1999. An algorithm to assess stiff-legged gait in traumatic brain injury J. Head Trauma Rehabil. 14, 136–145. Kerrigan, D., Burke, D., Nieto, T., Riley, P., 2001. Can toe-walking contribute to stifflegged gait? Am. J. Phys. Med. Rehabil. 80, 33–37. Lance, J.W., 1980. Pathophysiology of spasticity and clinical experience with baclofen. Spasticity: Disordered Motor Control, pp. 185–204. Lelas, J.L., Merriman, G.J., Riley, P.O., Kerrigan, D.C., 2003. Predicting peak kinematic and kinetic parameters from gait speed. Gait & Posture 17 (2), 106–112. Li, G., Kaufman, K.R., Chao, E.Y., Rubash, H.E., 1999. Prediction of antagonistic muscle forces using inverse dynamic optimization during flexion/extension of the knee. J. Biomech. Eng. 121, 316–322. McLean, S., Huang, X., Su, A., van den Bogert, A., 2004. Sagittal plane biomechanics cannot injure the ACL during sidestep cutting. Clin. Biomech. 19, 828–838. Nene, A., Mayagoitia, R., Veltink, P., 1999. Assessment of rectus femoris function during initial swing phase. Gait & Posture 9, 1–9. Neptune, R.R., Kautz, S.A., Zajac, F.E., 2001. Contributions of the individual ankle plantar flexors to support, forward progression and swing initiation during walking. J. Biomech. 34, 1387–1398. Poulain, B., Humeau, Y., 2003. Le mode d'action des neurotoxines botuliques: aspects pathologiques, cellulaires et moléculaires. Ann. Readapt. Med. Phys. 46 (6), 265– 275. Robertson, J.V., Pradon, D., Ben Smail, D., Fermanian, C., Bussel, B., Roche, N., 2009. Relevance of botulinum toxin injection and nerve block of Rectus femoris to Kinematic and functional parameters of low peak knee flexion in hemiplegic adult. Gait & Posture 29 (1), 108–112. Rosales, R.L., Arimura, K., Takenaga, S., Osame, M., 1996. Extrafusal and intrafusal muscle effects in experimental botulinum toxin-A injection. Muscle Nerve 19, 488–496. Schwartz, M.H., Rozumalskia, A., Trosta, J.P., 2008. The effect of walking speed on the gait of typically developing children. J. Biomech. 41 (8), 1639–1650. Stoquart, G.G., Detrembleur, C., Palumbo, S., Deltombe, T., Lejeune, T.M., 2008. Effect of botulinum toxin injection in rectus femoris on stiff-knee gait in people with stroke: a prospective observational study. Arch. Phys. Med. Rehabil. 89, 56–61. Winter, D.A., Sidwall, H.G., Hobson, D.A., 1974. Measurement and reduction of noise in kinematics of locomotion. J. Biomech. 7, 157–159. Yamaguchi, G., Zajac, F., 1990. Restoring unassisted natural gait to paraplegics via functional neuromuscular stimulation: a computer simulation study. IEEE Trans. Biomed. Eng. 37 (9), 886–902.
Contents lists available at SciVerse ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Effect of botulinum toxin injection on length and lengthening velocity of rectus femoris during gait in hemiparetic patients N. Lampire a, b,⁎, N. Roche c, P. Carne a, L. Cheze b, D. Pradon c a b c
Laboratoire d'analyse du mouvement, CMPR L'ADAPT Loiret, Amilly, France Laboratoire de Biomécanique et Mécanique des Chocs, Université de Lyon, UMR_T 9406, Université Lyon 1, IFSTTAR, Villeurbanne, France EA 4497 GRCTH, CIC-IT, CHU Raymond Poincaré, UVSQ, Garches, France
a r t i c l e
i n f o
Article history: Received 23 August 2012 Accepted 10 December 2012 Keywords: Gait Stroke Spasticity Rectus femoris Botulinum toxin
a b s t r a c t Background: In hemiparetic patients, rectus femoris spasticity is one of the main causes of reduced knee flexion in swing phase, known as stiff knee gait. Botulinum toxin is often used to reduce rectus femoris spasticity and to increase knee flexion during swing phase. However, the mechanisms behind these improvements remain poorly understood. The aim of this study was (1) to quantify maximal rectus femoris length and lengthening velocity during gait in ten adult hemiparetic subjects with rectus femoris spasticity and stiff knee gait and to compare these parameters with those of ten healthy subjects and (2) to study the effect of botulinum toxin injection in the rectus femoris muscle on the same parameters. Methods: 10 patients with stiff knee gait and rectus femoris spasticity underwent 3D gait analysis before and one month after botulinum toxin injection of the rectus femoris (200 U Botox®, Allergan Inc., Markham, Ontario, CANADA). Rectus femoris length and lengthening velocity were quantified using a musculoskeletal model (SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA). Findings: Maximal length and lengthening velocity of the rectus femoris were significantly reduced on the paretic side. There was a significant increase in muscle length as well as lengthening velocity during gait following botulinum toxin injection. Interpretation: This study showed that botulinum toxin injection in the spastic rectus femoris of hemiparetic patients improves muscle kinematics during gait. However maximal rectus femoris length did not reach normal values following injection, suggesting that other mechanisms are likely involved. © 2013 Elsevier Ltd. All rights reserved.
1. Introduction Following stroke, the characteristics of gait are altered in comparison with healthy subjects. Aside from changes in spatiotemporal and postural parameters, there are also alterations in lower limb kinematics such as reduced peak knee flexion during swing phase of the gait cycle, known as stiff knee gait (SKG) (Kerrigan et al., 2001). One of the main causes of SKG is inappropriate electromyographic activity (EMG) of the rectus femoris (RF) muscle, related to spasticity of this muscle (Goldberg et al., 2004). Other causes are reduced peak hip flexion and reduced peak plantar flexion moment (Kerrigan et al., 1999). One of the classic treatments for RF spasticity associated with abnormal EMG activity is botulinum toxin injection (BTI). This neuro-toxin works by blocking neurotransmitter release at the neuromuscular junction. In consequence this treatment induces a paresis
Abbreviations: RF, Rectus Femoris muscle; SKG, Stiff Knee Gait; BTI, Botulinum Toxin Injection; EMG, Electromyographic; SD, Standard Deviation. ⁎ Corresponding author at: Laboratoire d'analyse du mouvement, CMPR L'ADAPT Loiret, 658D rue des Bourgoins, 45200 Amilly, France. E-mail address: [email protected] (N. Lampire). 0268-0033/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinbiomech.2012.12.006
of the injected muscle, which is ‘voluntary, reversible and transient’ (Poulain and Humeau, 2003). Several studies in hemiparetic subjects showed that BTI in the spastic RF muscle significantly increased peak knee flexion during swing phase as well as knee flexion velocity at toe off (Hutin et al., 2010; Robertson et al., 2009; Stoquart et al., 2008). However, the mechanisms behind these kinematic improvements are as yet poorly understood. It is largely supposed that the focal paralysis which occurs following BTI leads to an increase in maximal muscle length and the muscle's capacity to lengthen during gait. The study of these parameters may help to understand the mechanisms associated with spasticity and the effect of botulinum toxin on spastic muscles. This hypothesis is based on the results of a study by Jonkers et al. (2006) in which a musculoskeletal model (SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA) was used to quantify RF resting length as well as its length and variations during gait in 35 children with spastic diplegia. The results showed that maximal length and maximal lengthening velocity of the RF were both reduced during gait compared with those of healthy subjects. The SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA musculoskeletal model has two useful properties: the dynamic module can be used
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
to perform forward and inverse dynamic simulations on musculoskeletal models. The forward simulation is used for the calculation of the motion and the contact forces resulting from the specified muscle excitation. In contrast, the inverse simulation allows calculation of the muscle activations and forces required to generate a specific motion. The forward model was developed for the prediction of the biomechanical consequences of the modifications of model parameters following a surgical procedure or change in muscle strength (Goldberg et al., 2004; McLean et al., 2004; Neptune et al., 2001; Yamaguchi and Zajac, 1990). The inverse simulation is used for the evaluation of changes in gait resulting from pathology or treatment based on data obtained during 3D gait analysis of a patient (Crowninshield and Brand, 1981; Glitsch and Baumann, 1997; Li et al., 1999). This model is the only tool which allows indirect assessment of muscle strength or lengthening during movement such as gait. Nevertheless it is necessary to keep in mind that it is an indirect approach and that the data obtained result from a simulation. In consequence, the results should be interpreted with caution. However when the inverse simulation is carried out on data from the same subject, the inherent bias of this approach is reduced, therefore, allowing comparison of the modifications induced by a treatment (whatever the type) on muscle strength or length. The present study was based on this method and had two aims. The first was to assess maximal RF length and lengthening velocity during gait in adult hemiparetic subjects with RF spasticity and SKG and to compare these parameters with those of the healthy subjects. The second aim was to evaluate the effect of BTI in the RF muscle on the same parameters. 2. Methods 2.1. Subjects 10 hemiparetic subjects constituted the patient group (Table 1): 4 with right hemiparesis, 6 with left, 8 men, 2 women, average age: 39.6, SD: 9.5 years. Inclusion criteria were the following: a single cerebral lesion of vascular origin (ischemic or hemorrhagic) of more than 6 month duration or more than 4 months since last BTI, a decrease in peak knee flexion during swing phase measured using gait
Table 1 Demographic characteristics of the patients.
Subject 1 Subject 2 Subject 3 Subject 4 Subject 5 Subject 6 Subject 7 Subject 8 Subject 9 Subject 10
Age Height Weight Gender Paretic Time (years) (cm) (kg) side since lesion onset (months)
RF spasticity assessed by modified Ashworth scale before and 1 month after BTI
27
180
74
M
L
48
–
40
188
73
M
L
6
0/0
58
160
69
M
R
60
1/1
29
166
72
M
L
60
3/2
33
180
85
M
L
36
–
37
182
82
M
R
60
–
40
181
72
M
R
36
2/2
40
163
51
F
L
84
1/0
52
162
67
F
R
432
2/1
40
173
76
M
L
96
2/1
M = male, F = female, L = left and R = right, RF = rectus femoris, BTI = botulinum toxin injection.
165
analysis, prolonged RF activity during swing phase determined by EMG and ability to walk 20 min without stopping. Patients were excluded if they had any comorbid disability other than stroke, such as any visual impairment or musculoskeletal, cardiovascular, or pharmacological treatment that could interfere with gait or posture. All patients had previously been treated by BTI in the RF muscle before inclusion in this study. 10 healthy subjects (with no central nervous system lesions; 6 men and 4 women; average age: 32.5, SD: 6 years) were also included as a control group. All subjects gave informed consent prior to inclusion. This study was carried out according to “The Ethical Codes of the World Medical Association” (Declaration of Helsinki) and was approved by the local ethical committee (CPP Ile de France, Ambroise Paré). 2.2. Protocol 2.2.1. Hemiparetic patients BT was injected in the spastic RF muscle on the hemiparetic side of all patients included. 200 U of type A botulinum toxin (Botox®, Allergan Inc., Markham, Ontario, CANADA) diluted to 40 U/ml was injected in three anatomical points in the RF muscle using electrostimulation (5 mA) by the same investigator. The technique used was the same to that previously used in Hutin et al. (2010): one point 5 cm under the inguinal crease on a line between the antero-inferior iliac spine and the patella. The second point was ten centimeters above the patella and the third was in the middle of the two others on the same line. The dose used was the same as that in other studies evaluating the impact of BTI in the RF muscle on peak knee flexion in swing phase (Hutin et al., 2010; Robertson et al., 2009; Stoquart et al., 2008). 2.3. Assessments This was an observational study. Two assessments were carried out: before injection (PRE) and 1 month post injection (POST). The second assessment was carried out one month post injection because this is when the toxin is at maximal effectiveness (Juzans et al., 1996). A clinical assessment and gait analysis were carried out at each assessment. • Clinical assessment Sagittal hip, knee and ankle joint ranges of motion were assessed in order to check for muscle contractures. Spasticity of the quadriceps and RF was evaluated using the modified Ashworth scale. All clinical tests were carried out by the same physiotherapist. • Gait analysis Each patient underwent three dimensional (3D) gait analysis using a Motion Analysis System (Motion Analysis Corporation, Santa Rosa, CA, USA, sampling frequency 100 Hz). Markers were positioned according to the Helen Hays protocol (Kadaba et al., 1990). Subjects walked barefoot at their own comfortable pace. Ten gait cycles, corresponding to those during which the patient walked on the force plate, resulting from ten gait trials were recorded and averaged for each patient. Patients were offered a break after five trials. EMG was recorded simultaneously during the kinematic data collection using bipolar surface electrodes (MA 311) placed on the rectus femoris, vastus lateralis, semimembranosus, tibialis anterior, gastrocnemius medialis and soleus muscles. Vastus lateralis activity was assumed to be representative of the three vastii (Nene et al., 1999). Digitalized EMG, Motion Lab System, LA, USA (1000 Hz) was analyzed to determine the periods of activity of each muscle during gait, and to include them in the model. Moreover, the EMG signal of the RF and VL muscles was used to qualitatively determine if their activation patterns were abnormal during swing phase or not.
166
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
2.4. Data analysis 2.4.1. Joint kinematics A 6 Hz low pass Butterworth filter was used in the signal processing (Winter et al., 1974). Kinematics of the pelvis, hips, knees and ankles were calculated using Orthotrak 6.2.8 (Motion Analysis Corporation, Santa Rosa, CA, USA). This software uses Euler rotations (Grood & Suntay method) to calculate kinematic parameters. Knee flexion/extension angular velocity at toe off was calculated by deriving the displacement data of the tibia relative to the femur. 2.4.2. Muscle kinematics SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA modeling software (Delp et al., 1990) was used to calculate the length of the musculo–tendinous complex of the RF. The standard generic SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA model was used to scale the model from the gait analysis data. This model consists of 13 rigid segments (pelvis, left and right femurs, left and right tibias, left and right patellae, left and right heels, left and right feet and left and right toes) with 17 degrees of freedom. In this model, each lower limb consists of 43 musculo–tendinous complexes. Each complex is defined by its origin, insertion, and if necessary, by defined via points in order to specify the muscle trajectory as much as possible. Particular attention was paid during the collection of anthropometric data for each subject (age, weight, height, foot length, ankle and knee diameters) in order to optimize the model scaling. The RF length was quantified during the gait cycle and when the subject was virtually placed in the anatomical reference position (defined as an angle of 0° for the hip and knee joints). Three muscles parameters were studied: • Normalized maximal RF length (maxNL): defined as maximal RF length during gait minus RF length in the ‘virtual’ anatomical reference position, in millimeters. This normalization reduces bias related to morphological differences between subjects, as well as differences related to marker placement between PRE and POST recordings for each patient. • Maximal RF lengthening velocity (maxLV): calculated for each gait cycle; for example MaxLV=Peak4 in Fig. 1. The average maximal lengthening velocity was then calculated for each subject. • Mean lengthening velocity at lengthening peaks (meanLV): defined as the mean of the positive peak lengthening velocity during one gait cycle. The mean velocity was then calculated for each subject. The velocity peaks were defined by zero acceleration. For example: MeanLV of cycle 1=(Peak1+Peak2+Peak3+Peak4)/4 (Fig. 1). Four peaks are present in this example. When the curve included more peaks, the average of all of the peaks was used.
These parameters were calculated in the PRE and POST conditions for the hemiparetic patients as well as for the healthy subjects. Because joint kinematics have been shown to be influenced by gait velocity (Schwartz et al., 2008), slow gait was recorded for the healthy subjects (cadence of 69 steps/min guided by a metronome) so that the average gait velocity of the healthy group was similar to that of the patient group. 2.5. Statistical analysis A Mann–Whitney test was used to compare age, height, weight and RF length in the anatomical reference position of the two groups. A Wilcoxon test was used to compare differences in maximal normalized RF length, maxLV and meanLV during gait, between the PRE and POST conditions in the patient group and the control group. P b 0.05 was considered significant in all cases. Correlations between (i) the increase in knee flexion during swing phase and the increase in gait velocity, (ii) the increase in knee flexion angular velocity at toe off and the increase in peak knee flexion during swing phase, maxLV and meanLV, (iii) the increase in maxNL and the increase in peak knee flexion during swing phase, and (iv) the increase of maxNL and maxLV and the increase in gait speed after BTI were assessed using a Spearman correlation. Statistical analyses were carried out using Statistica 7.0 software StatSoft, Inc., OK, USA. 3. Results 3.1. Anthropometrical comparison between patients and healthy subjects There was no statistical difference for age, height and weight between hemiplegic patients and healthy subjects (P b 0.05). 3.2. Clinical evaluation Seven of the 10 subjects underwent clinical examinations. The remaining three were outpatients and only able to complete the gait analysis. RF spasticity ranged from 1 to 3 in PRE on the modified Ashworth scale and was decreased one month after BTI (range 1–2). 3.3. Spatiotemporal The spatiotemporal parameters analyzed for the control group and for the patient group in the PRE and POST conditions are shown in Table 2. There was a significant mean increase of 0.13 m/s (24%) in gait velocity (P b 0.05) in the POST condition for the patient group. There was no significant difference in gait velocity between the patient group and the healthy subjects in the PRE or POST conditions (P > 0.05). 3.4. Kinematics The kinematic parameters analyzed for the control group and for the patient group in the PRE and POST conditions are shown in Table 2. There was a significant increase in peak knee flexion of 7.5° (+27%) in the POST condition for the patient group (P b 0.05). There was also a significant increase of 53°/s (+43%) in knee flexion velocity at toe off (P b 0.05). Despite the significant increase in peak knee flexion and knee flexion velocity at toe off, these parameters remained significantly lower than those of the healthy subjects (P b 0.05). 3.5. Biomechanical parameters of the RF
Fig. 1. Lengthening velocity of the rectus femoris during one gait cycle of a hemiparetic patient. The ellipses highlight the lengthening velocity peaks. The square highlights the maxLV peak.
The biomechanical parameters of the RF analyzed for the control group and for the patient group in the PRE and POST conditions are shown in Table 2 and Figs. 2–3. On the paretic side, maximal normalized length of the RF (RF maximal length during gait−RF length in virtual anatomical position) was 17 mm
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
167
Table 2 Mean values of the parameters studied. Paretic PRE Gait velocity (m/s) Peak knee flexion (°) Knee flexion angular velocity at toe off (°/s) MaxNL (mm) MaxLV (m/s) MeanLV (m/s)
0.57 27.3 122.5 17 0.07 0.03
(SD 0.24)⁎ (SD 7.3)⁎,⁎⁎⁎ (SD 68.3)⁎,⁎⁎,⁎⁎⁎ (SD 9)⁎,⁎⁎,⁎⁎⁎ (SD 0.03)⁎,⁎⁎,⁎⁎⁎ (SD 0.02)⁎,⁎⁎,⁎⁎⁎
Paretic POST 0.70 34.8 175.7 22 0.09 0.05
Non-paretic PRE
(SD 0.28)⁎ (SD 7.1)⁎,⁎⁎⁎ (SD 50.1)⁎,⁎⁎,⁎⁎⁎ (SD 8)⁎,⁎⁎,⁎⁎⁎ (SD 0.03)⁎,⁎⁎,⁎⁎⁎ (SD 0.03)⁎,⁎⁎
0.57 61 280.2 38 0.12 0.07
(SD 0.24)⁎ (SD 5.8)⁎,⁎⁎⁎ (SD 54.1)⁎,⁎⁎,⁎⁎⁎ (SD 9)⁎⁎ (SD 0.03)⁎,⁎⁎ (SD 0.02)⁎,⁎⁎
Non-paretic POST 0.70 63.4 305.4 39 0.13 0.09
(SD 0.28)⁎ (SD 4.5)⁎,⁎⁎⁎ (SD 58.9)⁎,⁎⁎,⁎⁎⁎ (SD 7)⁎⁎ (SD 0.04)⁎,⁎⁎ (SD 0.03)⁎,⁎⁎
Healthy subjects 0.56 48 197.2 41 0.13 0.06
(SD 0.05) (SD 5.8) (SD 22.2) (SD 4) (SD 0.02) (SD 0.01)
PRE = before botulinum toxin injection; POST = post botulinum toxin injection; MaxNL = normalized maximal RF length; MaxLV = maximal RF lengthening velocity; MeanLV = mean lengthening velocity at lengthening peaks. ⁎ Denotes a statistically significant difference between PRE and POST conditions. P ≤ 0.05 in each case. ⁎⁎ Denotes a statistically significant difference between the paretic and non-paretic sides. P ≤0.05 in each case. ⁎⁎⁎ Denotes a statistically significant difference between hemiparetic group GH and healthy group. P ≤ 0.05 in each case.
(SD 9) in the PRE condition and 22 mm (SD 8) in the POST condition. This increase of 5 mm (30%) following BTI was significant (Pb 0.05). This increase in maximal length in the POST condition was significantly correlated (R=0.85) with a significant increase (Pb 0.05) in peak knee flexion at the moment of the maximal peak normalized RF length (PRE: 23.7°, SD 8.4°; POST: 30.2°, SD 9.2°). There was no related change in flexion/extension angle of the hip at the same instant of the gait cycle (PRE=6.3°, SD 11°, POST=7.9°, SD 10°, P>0.05). Despite the significant increase in normalized RF length on the paretic side in the POST condition, it remained significantly lower than the non-paretic length and healthy subjects' values (Pb 0.05). On the non-paretic side, maxNL was 38 mm (SD 9) in the PRE condition. It was increased by 1 mm (4%) in the POST condition to 39 mm (SD 7). This difference was not significant (P>0.05). There was no significant difference in hip or knee flexion/extension angles between the PRE and POST conditions at maximal peak normalized RF length of the same side (hip PRE=7.6°, SD 10.2°, hip POST 6.9°, SD 7.7°, P>0.05; knee PRE= 51.6°, SD 6.4°, knee POST=54.4°, SD 4.6°, P>0.05). There were no significant differences for these parameters between the patient group and healthy group in either the PRE or POST condition (P>0.05). On the paretic side, in the PRE condition, maxLV of the RF was 0.07 m/s (SD 0.03) and meanLV was 0.03 m/s (SD 0.02). In the POST
condition, the maxLV was 0.09 m/s (SD 0.03) and was significantly increased (+28%, Pb 0.05). MeanLV was 0.05 m/s (SD 0.03) and was also significantly increased (+66%, Pb 0.05). Comparison with the healthy group showed that maximal velocity and mean velocity at lengthening peaks in the PRE condition were significantly lower in the patient group (P b 0.05). In the POST condition, only the maxLV remained significantly different from that of the healthy subjects (P b 0.05). On the non-paretic side, maxLV of the non-injected RF in the PRE condition was 0.12 m/s (SD 0.03) and mean velocity of the lengthening peaks was 0.07 m/s (SD 0.02). In the POST condition, there was a significant increase in maxLV of 8% to 0.13 m/s (SD 0.04) (P b 0.05) and a significant increase in meanLV of 28% to 0.09 m/s (SD 0.03) (P b 0.05). There were no significant differences between the patient and healthy groups for these parameters in the PRE or POST conditions. 3.6. Correlations There was a significant correlation between the increase in maxNL and the increase in peak knee flexion during swing phase (R = 0.86, P b 0.05) in both the PRE and POST conditions. There was a significant correlation between the increase in knee flexion during swing phase and the increase in gait velocity (R = 0.79, P b 0.05). There was no
Fig. 2. Normalized rectus femoris length, and hip and knee kinematics for patient No. 9. GS = healthy subjects, PRE = before botulinum toxin injection; POST = post botulinum toxin injection, RF = rectus femoris.
168
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
Fig. 3. Mean lengthening velocity curve of rectus femoris. GS = healthy subjects, GH PRE = hemiparetic patients before botulinum toxin injection; POST = hemiparetic patients post botulinum toxin injection.
correlation between the increase in max NL and maxLV and the increase in gait speed (P > 0.05). There was also a significant correlation between the increase in knee flexion angular velocity at toe off and the increase in peak knee flexion during swing phase (R = 0.67, P b 0.05). Moreover, knee flexion angular velocity at toe off was correlated with maxLV in the PRE (R = 0.95, P b 0.05) and POST (R = 0.87, P b 0.05) conditions, and with meanLV in the PRE (R = 0.80, P b 0.05) and POST (R = 0.85, P b 0.05) conditions. 4. Discussion The aim of this study was to quantify maximal RF length and lengthening velocity during gait in hemiparetic patients with SKG and RF spasticity, before and after BTI in this muscle. The results showed that maxNL of RF during gait was reduced on the hemiparetic side in comparison with the non-paretic side and with healthy subjects walking at a similar gait velocity. Furthermore the results: i) confirm that BTI in the spastic RF significantly increases peak knee flexion in swing phase as well as gait velocity in hemiparetic patients (Hutin et al., 2010; Robertson et al., 2009; Stoquart et al., 2008); and ii) show that BTI increases maxNL of the injected muscle during movement although the values remain lower than those of the non-paretic side and healthy subjects. Before discussing the results of this study, several points should be discussed: Concerning the gender of the patients included, the patient group was predominantly composed of men (8 men, 2 women) whereas in the control group there were 6 men and 4 women. To our knowledge, there is no data in the literature suggesting that the gait pattern of men differs from those of women either in healthy subjects or in hemiplegic patients. It seems therefore unlikely that this difference in gender could constitute a limitation in the interpretation of the result. Concerning the “heterogeneity” of the patients included and more particularly the time since the lesion onset, we do not feel that these differences constitute a limitation in the interpretation of the results. Indeed, all the patients had moderate spasticity of the rectus femoris muscle, which was independent of the time since the lesion onset.
Moreover all had the same proportion of improvement in the different kinematic muscle parameters studied after BTI injection. It therefore seems that the heterogeneity of the time since the lesion onset does not constitute a bias but on the contrary means that the results of this study are applicable to a large population of hemiplegic patients. Concerning the anthropometric models: there was no anthropometric difference between the two groups assessed. Moreover, as far as we are aware, there is no specific model for men or women or for patients and healthy subjects. Thus similar to Jonkers et al. (2006), we used the same anthropometric model for the two groups. We cannot exclude that this methodological point might constitute a limitation in the interpretation of the results but no other solution was, to our knowledge, available. The same investigator performed all the anthropometric measures in order to limit the bias inherent to this approach. For the EMG signal analysis, similar to Jonkers et al. (2006) and Robertson et al. (2009), we used a qualitative approach to determine if the RF muscle activation pattern during swing phase was abnormal. An abnormal RF pattern is characterized by inappropriate EMG activity in the middle of the swing phase. In contrast, the quantitative EMG of all muscles recorded was integrated in the musculoskeletal model to optimize the results. Concerning the results obtained with the SIMM®, MusculoGraphics, Inc., Santa Rosa, California, USA model. There were no significant differences between RF lengths in the anatomical reference position in the PRE and POST conditions in the patient group. Moreover the maxNL of the non-paretic side was not modified following BTI. This suggests that the results obtained for the paretic limb can be interpreted with confidence. Concerning the gait velocity, the improvement following BTI was significant. However, although the gait velocity of the patient group was increased after BTI, it was not significantly different from that of the healthy subjects. This lack of difference is likely to be due to the difference in the SD between the two groups (paretic and healthy groups). The standard deviation is higher for the group of the patients. The results of this study showed firstly that maxNL of the paretic side was significantly lower than that of the non-paretic side as well as that of the healthy subjects walking at a similar velocity. These results conform to those of Jonkers et al. (2006) who found a decrease in maximal RF length in 35 diplegic children with reduced knee flexion during swing phase.
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
The first question that can be addressed is: What are the mechanisms involved in the decrease in spastic RF length in hemiparetic patients? Spasticity has been defined as a reflex muscle contraction in response to stretch of the same muscle (Lance, 1980). All the patients included in this study had spasticity of the RF muscle (Table 1). Spasticity can lead to abnormal muscle stiffness, which in turn could reduce the maximal normalized length of the muscle and its general capacity to lengthen during movement. Our results support this hypothesis. Indeed, the results show that 1 month after BTI in the paretic RF, its maximal normalized length was significantly increased by about 30%. These results are similar to those of Eames et al. (1997) who found that increased gastrocnemius length following BTI was significantly inversely correlated with clinically evaluated spasticity of the same muscle in 39 children with cerebral palsy. They showed that the lower the degree of spasticity before injection, the greater the increase in muscle length after injection. It is interesting to note that in the population of hemiparetic patients included, there was i) no significant difference between the maxNL of the non-paretic RF and that of the healthy subjects in the PRE or POST conditions; and ii) no change in non-paretic maxNL in the POST condition. This further confirms the hypothesis that the increased maxNL is principally due to the decrease in spasticity following BTI and only occurs in the injected muscle. The second question that can be addressed is: What are the main causes of decreased peak knee flexion in swing phase? Jonkers et al. (2006) highlighted the fact that in order to interpret the impact of a change in muscle length, it is necessary to take into account the kinematics of the joints on which the muscle acts. The RF is a bi-articular muscle. It originates in the antero-inferior iliac spine and inserts into the upper border of the patella. In order to interpret changes in RF length, kinematics of the hip and knee joints in the sagittal plane must therefore be taken into account. Our results suggest that the hip flexion angle at the time of maximal length of the injected RF was not significantly different between PRE and POST conditions. There was, however, a significant increase in knee flexion of 7.5° at this instant. This therefore suggests that the increase in maxNL during gait following BTI results in an increase in peak knee flexion. This result is confirmed by the significant correlation between the increase in maxNL and peak knee flexion in swing phase (R= 0.86) and is in accordance with the results of previous studies (Robertson et al., 2009; Stoquart et al., 2008) which showed that injection of equivalent doses of botulinum toxin led to an increase in peak knee flexion during swing phase of respectively +5° and +8°. Although the injection of botulinum toxin in the RF muscle led to a significant increase in maxNL and an increase in peak knee flexion, these parameters remain significantly reduced in comparison with the non-paretic side and the healthy group. An insufficient dose of botulinum toxin may explain the lack of knee flexion that persists; however the injected dose is in agreement with the French Agency for Safety of Health Products and is similar to that used in other studies which assessed the effects of BTI in RF on peak knee flexion in hemiplegic patients (Hutin et al., 2010; Robertson et al., 2009; Stoquart et al., 2008). This therefore suggests that other factors could underlie the lack of knee flexion during swing. Kerrigan et al. (2001) showed that the lack of knee flexion in hemiparetic patients could be explained by several mechanisms: inappropriate activity of another head of the quadriceps, reduced hip flexion or reduced peak plantar flexion moment. In our study, only 1 patient also had abnormal activity of another head of the quadriceps determined by EMG during swing phase of the gait cycle. However, mean peak hip flexion was significantly reduced on the paretic side compared with the non-paretic side and also with the healthy group. This suggests that in the group of patients included, who all had spasticity of the RF and reduced peak knee flexion during swing phase, the lack of knee flexion is multi-factorial. It is at least related to a combination of abnormal RF EMG activity in swing phase of the gait cycle and a lack
169
of hip flexion, which may be caused by weak hip flexors. The clinical implications of these results are that the management of patients with reduced knee flexion and inappropriate RF activity should include strengthening of the hip flexor muscles as well as BTI in the RF muscle. In a literature review on spasticity and muscle hypertonia, Dietz and Sinkjaer (2007) cite many studies which demonstrate changes in intrinsic muscle fiber length, in the conjunctive tissue of muscles (collagen, fatty tissue) and tendons, and of the fiber type (slow and fast) of spastic muscles. According to the authors, these changes in muscle properties lead to muscle stiffness which in turn causes i) decreased maximal length of the spastic muscle, ii) a large decrease in lengthening capacity and iii) decreased range of movement of the associated joint. These results can be related to the definition of spasticity proposed by Crenna (1998) in which spastic hypertonia, classically described as an increase in the stretch reflex, is associated with increased mechanical muscle resistance. The third question that can be discussed is: Are the results influenced by the improvement of the gait speed? In healthy subjects, it has been shown that gait speed influences kinematic gait parameters. Lelas et al. (2003) showed that gait speed was correlated with an increase in peak knee flexion during loading response, peak knee flexion during swing phase and peak hip and knee extension. Recently Hutin et al. (2012) studied the impact of gait speed on kinematic gait parameters in stroke patients. They showed that increasing gait velocity is associated with an increase in knee flexion velocity at toe off and an increase in peak knee flexion during swing phase of the gait cycle. It could be thus hypothesized, in stroke patients with stiff knee gait and spasticity of the RF, that the increase in length and lengthening capacity found after BTI is not directly due to the paresis of the muscle injected but might also be due to the increase in gait speed. To answer the question we analyzed the correlation between the percentage of improvement of RF length or RF maxLV induced by BTI and the percentage of improvement of the gait speed. Our results indicate that there was no correlation between these two parameters suggesting that the modifications of these two kinematic muscle parameters were mainly due to the focal botulinum toxin action. The last question that needs to be discussed is: Is the action of botulinum toxin limited to the neuromuscular junction? It is interesting to note that in the PRE condition, the maxLV of the patient group was significantly lower on the paretic side than on the non-paretic side and in the healthy group. Similarly, the meanLV of the RF was significantly lower on the paretic side than on the non-paretic side and in the healthy group. These results indicate that maxLV and meanLV (which induces a spastic response) are reduced in spastic RF muscles compared with non-spastic muscles. MaxLV and meanLV were also significantly correlated with the reduced knee flexion angular velocity at toe off (R = 0.95). These results could be explained by the increased excitability of the monosynaptic reflex arc (muscle spindle – Ia fibre – α motoneurone) following a central nervous system lesion or by a rectus femoris contracture or by a weakness of the triceps surae. Nevertheless, after BTI, in the spastic RF muscle, there were i) a significant increase in maxLV (+ 28%) and ii) a significant increase in meanLV (+ 66%) of the injected RF. The increase in maxLV and meanLV of rectus femoris after BTI indicates that BTI, which decreases the excitability of the monosynaptic reflex arc, is responsible for the improvement of these parameters. In addition, any rehabilitation which the patient followed was not changed during the study period, both in terms of frequency and intensity. The changes which occurred in the muscles do not appear to be related to changes which occur following training. It therefore appears that the reduction of knee flexion is related to a decrease in maxLV and meanLV, and therefore with hyperexcitability of the monosynaptic reflex arc. These results suggest that as well as blocking the neuromuscular junction, botulinum toxin could also modify the threshold of the stretch reflex.
170
N. Lampire et al. / Clinical Biomechanics 28 (2013) 164–170
During movement, the threshold of the stretch reflex is constantly regulated by the γ motoneuron which controls the sensitivity of the muscle spindle. Filippi et al. (1993) and Rosales et al. (1996) respectively showed in animal models that synaptic transmission between the γ motoneuron and the muscle spindle is reduced after BTI. The fact that mean LV of the RF (which induces the spastic response) is significantly increased after BTI is in accordance with this hypothesis and suggests that in humans, the toxin could also increase the stretch reflex threshold. This hypothesis is also indirectly in agreement with the results of Hutin et al. (2011) which showed in 15 hemiparetic patients that BTI of the RF increases the fluidity of movement during gait demonstrated by a reduction in the number of continuous relative phase reversals during swing phase. To conclude, the results of this study suggest that, in hemiparetic patients with reduced knee flexion in swing phase and RF spasticity, the length of the RF muscle is decreased. The study also demonstrated that BTI increased maximal RF length as well as its lengthening capacity, although these parameters did not reach normal values. Further studies will be necessary to determine the neuro-physiological mechanisms responsible for these biomechanical changes in patients with central nervous system lesions. Conflict of interest We have no conflict of interest with this work. References Crenna, P., 1998. Spasticity and ‘spastic’ gait in children with cerebral palsy. Neurosci. Biobehav. Rev. 22 (4), 571–578. Crowninshield, R., Brand, R., 1981. A physiological based criterion of muscle force prediction in locomotion. J. Biomech. 14, 793–801. Delp, S.L., Loan, P., Hoy, M.G., Zajac, F.E., Topp, E.L., Rosen, J.M., 1990. An interactive graphics-based model of the lower extremity to study orthopaedic surgical procedures. IEEE Trans. Biomed. Eng. 37 (8), 757–767. Dietz, V., Sinkjaer, T., 2007. Spastic movement disorder: impaired reflex function and altered muscle mechanics. Lancet Neurol. 6 (8), 725–733. Eames, N., Baker, R., Cosgrove, A., 1997. The effect of botulinum injections on gastrocnemius muscle length. Gait & Posture 5 (2), 138. Filippi, G.M., Errico, P., Santarelli, R., Bagolini, B., Manni, E., 1993. Botulinum A toxin effects on rat jaw muscle spindles. Acta Otolaryngol. 113, 400–404. Glitsch, U., Baumann, W., 1997. The three-dimensional determination of internal loads in the lower extremity. J. Biomech. 30 (11/12), 1123–1131. Goldberg, S.R., Anderson, F.C., Pandy, M.G., Delp, S.L., 2004. Muscles that influence knee flexion velocity in double support: implications for stiff-knee gait. J. Biomech. 37, 1189–1196.
Hutin, E., Pradon, D., Barbier, F., Gracies, J.M., Bussel, B., Roche, N., 2010. Lower limb coordination in hemiparetic subjects: impact of botulinum toxin injections into rectus femoris. Neurorehabil. Neural Repair 24 (5), 442–449. Hutin, E., Pradon, D., Barbier, F., Gracies, J.M., Bussel, B., Roche, N., 2011. Lower limb coordination patterns in hemiparetic gait: factors of knee flexion impairment. Clin. Biomech. 26, 304–311. Hutin, E., Pradon, D., Barbier, F., Bussel, B., Gracies, J.M., Roche, N., 2012. Walking velocity and lower limb coordination in hemiparesis. Gait & Posture 36 (2), 205–211. Jonkers, I., Stewart, C., Desloovere, K., Molenaers, G., Spaepen, A., 2006. Musculo-tendon length and lengthening velocity of rectus femoris in stiff knee gait. Gait & Posture 23, 222–229. Juzans, P., Comella, J.X., Molgo, J., Faille, L., Angaut-Petit, D., 1996. Nerve terminal sprouting in botulinum type-A treated mouse levator auris longus muscle. Neuromuscul. Disord. 6 (3), 177–185. Kadaba, M.P., Ramakrishnan, H.K., Wootten, M.E., 1990. Measurement of lower extremity kinematics during level gait. J. Orthop. Res. 8, 383–392. Kerrigan, D.C., Bang, M.S., Burke, D.T., 1999. An algorithm to assess stiff-legged gait in traumatic brain injury J. Head Trauma Rehabil. 14, 136–145. Kerrigan, D., Burke, D., Nieto, T., Riley, P., 2001. Can toe-walking contribute to stifflegged gait? Am. J. Phys. Med. Rehabil. 80, 33–37. Lance, J.W., 1980. Pathophysiology of spasticity and clinical experience with baclofen. Spasticity: Disordered Motor Control, pp. 185–204. Lelas, J.L., Merriman, G.J., Riley, P.O., Kerrigan, D.C., 2003. Predicting peak kinematic and kinetic parameters from gait speed. Gait & Posture 17 (2), 106–112. Li, G., Kaufman, K.R., Chao, E.Y., Rubash, H.E., 1999. Prediction of antagonistic muscle forces using inverse dynamic optimization during flexion/extension of the knee. J. Biomech. Eng. 121, 316–322. McLean, S., Huang, X., Su, A., van den Bogert, A., 2004. Sagittal plane biomechanics cannot injure the ACL during sidestep cutting. Clin. Biomech. 19, 828–838. Nene, A., Mayagoitia, R., Veltink, P., 1999. Assessment of rectus femoris function during initial swing phase. Gait & Posture 9, 1–9. Neptune, R.R., Kautz, S.A., Zajac, F.E., 2001. Contributions of the individual ankle plantar flexors to support, forward progression and swing initiation during walking. J. Biomech. 34, 1387–1398. Poulain, B., Humeau, Y., 2003. Le mode d'action des neurotoxines botuliques: aspects pathologiques, cellulaires et moléculaires. Ann. Readapt. Med. Phys. 46 (6), 265– 275. Robertson, J.V., Pradon, D., Ben Smail, D., Fermanian, C., Bussel, B., Roche, N., 2009. Relevance of botulinum toxin injection and nerve block of Rectus femoris to Kinematic and functional parameters of low peak knee flexion in hemiplegic adult. Gait & Posture 29 (1), 108–112. Rosales, R.L., Arimura, K., Takenaga, S., Osame, M., 1996. Extrafusal and intrafusal muscle effects in experimental botulinum toxin-A injection. Muscle Nerve 19, 488–496. Schwartz, M.H., Rozumalskia, A., Trosta, J.P., 2008. The effect of walking speed on the gait of typically developing children. J. Biomech. 41 (8), 1639–1650. Stoquart, G.G., Detrembleur, C., Palumbo, S., Deltombe, T., Lejeune, T.M., 2008. Effect of botulinum toxin injection in rectus femoris on stiff-knee gait in people with stroke: a prospective observational study. Arch. Phys. Med. Rehabil. 89, 56–61. Winter, D.A., Sidwall, H.G., Hobson, D.A., 1974. Measurement and reduction of noise in kinematics of locomotion. J. Biomech. 7, 157–159. Yamaguchi, G., Zajac, F., 1990. Restoring unassisted natural gait to paraplegics via functional neuromuscular stimulation: a computer simulation study. IEEE Trans. Biomed. Eng. 37 (9), 886–902.