Clinical Neurophysiology xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis Neil G. Simon a,b, Cindy S.-Y. Lin b, Michael Lee a, James Howells c, Steve Vucic a,d, David Burke c, Matthew C. Kiernan a,e,⇑ a
Neuroscience Research Australia, Barker St, Randwick, NSW 2031, Australia Prince of Wales Clinical School, University of New South Wales, Randwick, NSW 2031, Australia Sydney Medical School, K25 – Medical Foundation Building, The University of Sydney, NSW 2006, Australia d Westmead Clinical School, C24 Westmead Hospital, The University of Sydney, NSW 2006, Australia e Brain and Mind Research Institute, The University of Sydney, Mallett St, Camperdown, Australia b c
a r t i c l e
i n f o
Article history: Accepted 29 July 2014 Available online xxxx Keywords: Amyotrophic lateral sclerosis Segmental motoneurone Spinal cord circuitry Pathophysiology H-reflex
h i g h l i g h t s H-reflex parameters correlated strongly with clinical upper motor neurone (UMN) dysfunction in ALS
patients. The Hmax/Mmax ratio may be attenuated by increased collision in ALS limbs, reflecting altered dynam-
ics of motor unit recruitment. Alterations of motor unit recruitment related to UMN signs, raising the possibility of transsynaptic
lower motor neurone modulation after UMN damage.
a b s t r a c t Objectives: There is accumulating evidence of dysfunction of spinal circuits in the pathogenesis of amyotrophic lateral sclerosis (ALS). Methods: The present study was undertaken to characterise the pathophysiological changes in segmental motoneuronal excitability in 28 ALS patients, using recruitment curves of the soleus H-reflex and M-wave, compared with clinical assessments of upper motor neuron (UMN) and lower motor neuron dysfunction. Results: H-reflex recruitment curves established that Hmax/Mmax and slope (Hh/Mh) ratios predicted clinical UMN dysfunction (p < 0.001). Changes in Hh/Mh were driven by reduced Mh. Assessment of Hmax/Mmax was similar in the ALS and control groups, and was affected by overlap of the H and M recruitment curves in ALS patients. Conclusion: Changes in the slope ratio (Hh/Mh) in ALS suggested that alterations in peripheral motor nerve excitability following UMN damage may affect the recorded H-reflex. Increased collision of reflex discharges with antidromically-conducted motor impulses may be exacerbated in ALS due to preferential loss of large-caliber a-motoneurones, which may explain the similarities in Hmax/Mmax between groups. Significance: Findings from the present study provide further insight into the pathophysiology of ALS, specifically the relative contributions of premotoneuronal and segmental motoneuronal dysfunction. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction ⇑ Corresponding author. Address: Brain and Mind Research Institute, The University of Sydney, Room 438, Level 4, M04G, 100 Mallett Street, Camperdown, NSW 2050, Australia. Tel.: +61 2 9114 44251; fax: +61 2 9114 4254. E-mail address: [email protected] (M.C. Kiernan).
Evidence of upper motor neuron (UMN) dysfunction in the context of lower motor neuron (LMN) degeneration is fundamental to the diagnosis of amyotrophic lateral sclerosis (ALS; Brooks et al., 2000). Objective measures of UMN dysfunction may prove useful
http://dx.doi.org/10.1016/j.clinph.2014.07.029 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Simon NG et al. Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.07.029
2
N.G. Simon et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
to delineate clinical phenotypes, permit monitoring of treatment response, and thereby serve as a measure of disease progression in clinical trials (Hardiman et al., 2011; Kiernan et al., 2011; Turner et al., 2009). However, clinical evidence of UMN involvement may be difficult to detect, particularly when coexistent LMN signs are prominent (Swash, 2012; Turner et al., 2013). There is accumulating evidence of dysfunction of spinal circuits linked to the pathogenesis of ALS (Turner and Kiernan, 2012). Dysfunction of spinal motoneuronal circuits, specifically reduced recurrent inhibition, has been suggested as a mechanism underlying the development of UMN clinical signs in ALS patients (Raynor and Shefner, 1994). Renshaw cells mediate recurrent inhibition of spinal motoneurones and early dysfunction of these cells has been identified in experimental models of ALS (Mazzocchio and Rossi, 2010). In addition, histopathological studies have identified loss of both interneurons and motoneurons in the spinal anterior horn of ALS patients (Stephens et al., 2006). A premise of the present study was that insight into the pathogenesis of ALS may be derived from interrogating segmental motoneuronal excitability and specifically its relation to UMN and LMN abnormalities. As such, the pathophysiological changes in segmental motoneuronal excitability in ALS were evaluated using the H-reflex pathway, as a tool to probe motoneuronal excitability and thereby function (Koelman et al., 1993; Misiaszek, 2003; Pierrot-Deseilligny and Burke, 2012), relating changes to clinical evidence of UMN and LMN dysfunction. A further motivation was to develop an objective approach that could be readily applied in most clinical units to monitor progress in individual patients and any change in response to treatment. 2. Materials and methods Clinical and neurophysiological data were prospectively acquired from a cohort of ALS patients referred to a specialised clinic, each with a diagnosis of clinically probable or definite ALS according to the Awaji-Shima criteria (de Carvalho et al., 2008). Twenty-eight ALS patients were recruited (mean age 62.7 ± 2.0 years, mean height 169.7 ± 3.5 cm; 46% male). Assessments were undertaken prior to the commencement of riluzole in 32% and on riluzole in 68%. A group of 15 age- and height-matched healthy subjects (mean age 57.6 ± 2.3, mean height 166.8 ± 1.6 cm, 47% male) served as a control group for the ALS cohort. Subjects were excluded if there was a prior history or clinical evidence of peripheral neuropathy, lumbosacral radiculopathy, or lumbar spinal surgery. An UMN disease control group comprising 8 patients with spinal cord injury and prominent lower limb UMN clinical signs (mean age 41.2 years, 146–233 days post-injury) were used to compare H-reflex parameters. All subjects provided written informed consent, and the study was approved by the South East Sydney Area Health Service Human Research Ethics Committee. 2.1. Clinical measures Standardized clinical assessments were undertaken in each ALS patient prior to further neurophysiological investigations. Motor functional status and disability was assessed using the revised ALS Functional Rating Scale, with a maximal score of 48 suggesting minimally symptomatic disease (ALSFRS-R; Cedarbaum et al., 1999). The clinical region of onset and symptomatic body regions at the time of clinical review were also recorded, as was height for ALS and control subjects. Muscle strength was assessed using the standard Medical Research Council (MRC) rating scale (Medical Research Council,
1976). Muscle groups graded included shoulder abduction, elbow flexion, wrist extension, finger flexion, finger extension, finger abduction, hip flexion, knee extension, ankle dorsiflexion, and great toe extension, with scores from all groups incorporated into the modified MRC sum score (Burrell et al., 2011). Muscle strength scores from individual lower limbs were added to produce a lower limb MRC sum score. Plantar flexion was not included in the score as there are inherent difficulties in clinically estimating plantar flexion strength with standard manual muscle testing. Weakness of plantar flexion is typically under detected (Lunsford and Perry, 1995). The presence of UMN clinical involvement in each lower limb was assessed using a formal protocol to determine an UMN score (Supplementary Data S1). Knee and ankle reflexes were individually scored on a two-point scale with a score of one given for a brisk reflex without spread or a preserved reflex in a wasted muscle, and a score of two given for a reflex with evidence of pathological spread. A score of one point each was given for the presence of an extensor plantar response (using the Babinski manoeuvre) and ankle clonus (>4 beats) for a maximal score of 6 per lower limb. Measures of muscle tone were not incorporated into the UMN score because their validity and reliability have not been confirmed (Fleuren et al., 2010). The UMN score was specifically devised as a global score of UMN dysfunction in the lower limb, to allow quantification of clinical UMN signs in the presence of concomitant LMN features, and to produce a range of values from low to high to provide robust data for the regression analyses. The global burden of LMN involvement in each lower limb was estimated using a quantitative scale grading muscle wasting in the thigh and leg (Burrell et al., 2011). Each region was graded on a 3 point scale to determine a Wasting Score out of 6 for each lower limb as follows: No wasting = 0 points; trace wasting = 1 point; moderate wasting = 2 points; severe wasting = 3 points. 2.2. Neurophysiological protocols H-reflex studies of the tibial nerve were undertaken with the subject seated in an adjustable armchair, with the hip flexed to 120°, the knee flexed to 160° and the ankle at 110° of plantar flexion (Burke et al., 1999; Lin et al., 2002; McNulty et al., 2008). Subjects were directed to keep their leg relaxed throughout the procedure with visual feedback used to identify any background EMG activity. The tibial nerve was stimulated at the popliteal fossa using bipolar electrodes. H-reflex responses and the soleus compound muscle action potential (M-wave) were recorded from the soleus muscle using Ag/AgCl electrodes (4620M, Unomedical Ltd., Birkerød, Denmark). The active electrode (G1) was placed in the midline of the muscle immediately below the gastrocnemius, and the reference (G2) was placed 4 cm distal to G1. An adhesive electrosurgical grounding plate (2406M, Unomedical Ltd., Birkerød, Denmark) was placed over the proximal fibula. The H-reflex and M-wave recruitment curves were recorded using an automated computerised system (QTRACÓ Institute of Neurology, Queens Square, UK) and a multifunction data acquisition system (NI PCI-6221, National Instruments, Austin, TX, USA). Stimulation was computer controlled and converted to current using an isolated linear bipolar constant-current stimulator (maximal output ± 50 mA; DS5, Digitimer, Welwyn Garden City, UK). Square-wave pulses of 1-ms duration were delivered at a frequency of 0.33 Hz. This stimulation frequency would produce a minor degree of post-activation depression in control subjects (Crone and Nielsen, 1989), but this phenomenon was less prominent in patients with spasticity in the lower limbs (Nielsen et al., 1993), such as might be expected in patients with ALS. The stimulation frequency was specifically selected to minimize the impact
Please cite this article in press as: Simon NG et al. Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.07.029
N.G. Simon et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
3
of post-activation depression while maintaining the tolerability of the procedure for the subjects involved. Responses were amplified (ICP511 AC amplifier, Grass Technologies, West Warwick, RI, USA) with mains frequency noise removed (Hum Bug 50/60 Hz Noise Eliminator, Quest Scientific Instruments, North Vancouver, Canada). Initially, the optimal stimulation position was identified at which the H-reflex was elicited with the lowest possible stimulus intensity. Subsequently, windows were adjusted to capture the latency and duration of the H and M waveforms to determine the maximal peak-to-peak amplitude. The stimulus intensity required to produce the maximal M-wave amplitude was determined (Mmaxint). H and M recruitment curves were then recorded with stimulation commencing at 3% of Mmaxint and increasing in 3% increments thereafter to a maximum of 50 mA. Three stimuli at each intensity level were delivered, and the resulting H-reflex and M-wave amplitudes were averaged to produce each point on the H and M recruitment curves (Fig. 1).
Corporation, Redmond, WA, USA). The following neurophysiological parameters were analysed (Fig. 2): Maximal H-reflex amplitude (Hmax), maximal compound muscle action potential amplitude (Mmax), minimal stimulus intensities to produce a H-reflex (Hthresh) and M-response (Mthresh), stimulus intensity at Hmax (Hmaxint), latency of Hmax (Hlat), latency of Mmax (Mlat), and Hlat minus Mlat (HMlat). All H-reflex and M-wave amplitudes were subsequently normalised to Mmax, and all stimulus intensities were normalised to Mthresh. A line of best fit was applied to the linear ascending component of the H and M recruitment curves using a method adapted from Funase et al. (1994). The linear regression equation was determined and used to calculate the slope angle in degrees (Hh and Mh) using the formula;
2.3. Data analysis
2.4. Statistical analysis
Calculations were performed off-line with data exported to Microsoft ExcelÓ for Mac version 14.2.3, 2011 (Microsoft
All descriptive values were presented as mean ± SEM. Statistical analysis was performed using SPSS version 20, 2011 (IBM, Armonk,
h¼
180
p
tan1 ðmÞ
where m is the slope calculated in the linear regression. Subsequently the slope ratio (Hh/Mh) was calculated.
Fig. 1. The H-reflex circuit and modulating spinal inputs. The H-reflex is comprised of a monosynaptic communication between Ia afferents originating in muscle spindles and aMNs in the spinal anterior horn. The H-reflex is modulated by several local and descending inputs, including other sensory inputs, interneuronal influences and descending spinal tracts. Changes in the physiology of spinal neuronal elements and interneurons have been identified in ALS patients (marked in red) including Renshaw cells (recurrent inhibition), inhibitory interneurons, anterior horn cells and descending tracts. Data from a representative ALS patient are depicted with H and M recruitment curves and corresponding raw data showing changing waveforms with increasing stimulation intensity (a. to e.). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Please cite this article in press as: Simon NG et al. Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.07.029
4
N.G. Simon et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
3.2. Baseline H-reflex parameters in ALS
Fig. 2. H-reflex and M-wave recruitment curves and derived parameters. Representative H-reflex and M-wave recruitment curves demonstrating the parameters measured in the present study.
NY, USA). Shapiro–Wilk tests were used to assess normality of the individual parameters. One-way ANOVA tests were used to compare means between the ALS, healthy control, and disease control groups, and correlation analyses were performed using Pearson product-moment and Spearman’s rank-order tests as appropriate for normally-distributed and not normally distributed data, respectively. Multiple linear regression models were constructed to assess the degree to which differences in Hmax/Mmax and slope ratio were predicted by variation in clinical UMN dysfunction (UMN Score) and age. The model for Hh/Mh also incorporated Mmax. Analyses were preplanned to minimise Type I statistical error.
H-reflexes were recordable in all control limbs and in 92% of ALS limbs (absent in 4 limbs). In the four limbs with absent H-reflexes there was a marked reduction of the soleus M-wave amplitude (mean = 0.48 ± 0.28 mV). In these limbs, UMN signs were modest (UMN score 0 or 1), and LMN signs were more prominent (wasting score 2–4). Each of these limbs was weak (MRC sum score 10–17), and EMG studies identified denervation in gastrocnemius. As such, the limbs with absent H-reflexes were associated with a greater burden of LMN disease, and the absent reflexes could reflect an altered Ia afferent input due to atrophic changes involving intrafusal muscle (Swash and Fox, 1974; Swash, 2012). Mmax was significantly reduced in ALS patients when compared with controls (ALS = 10.2 ± 1.2 mV, control = 15.9 ± 1.2 mV, p = 0.001), and this was still so after exclusion of the data for the 4 limbs in which no H-reflex was recordable. Hmax was also reduced in ALS patients (ALS = 3.8 ± 0.5 mV, control = 6.3 ± 0.9 mV, p < 0.05), such that there was no difference in the mean Hmax/Mmax ratios (0.38 ± 0.04 in ALS, 0.37 ± 0.05 in controls, p = 0.88). When ALS limbs with absent reflexes were excluded from analysis, Hmax/ Mmax was slightly greater (0.43 ± 0.05) but not significantly different from control values. Hlat (ALS = 33.4 ± 0.5 ms, control = 30.9 ± 0.5 ms, p < 0.001), Mlat (ALS = 5.7 ± 0.4 ms, control = 4.7 ± 0.2 ms, p < 0.001) and HMlat (ALS = 27.7 ± 0.4 ms, control = 26.2 ± 0.3 ms, p < 0.05) were all prolonged in ALS patients despite similar mean height. The findings are consistent with these reports of standard motor nerve conduction studies (Feinberg et al., 1999; Mills and Nithi, 1998), and the prolongation of H–M latency difference suggests loss of large, rapidly conducting fibres rather than slower conduction in sprouting terminal axons as a mechanism of the prolongation of H-reflex latency. 3.3. Influence of UMN dysfunction on H-reflex parameters
3. Results 3.1. Clinical characteristics Assessment of disease onset from the present cohort established that their disease started clinically in the upper limb in 39%, bulbar region in 32%, lower limb in 25%, and thoracic region in 4%. The mean duration of disease between the onset of weakness and the study session was 25.8 ± 4.0 months. Mean ALSFRS-R score at the time of assessment was 39.8 ± 1.3 (range 23–48, where 48 represents minimal disability), indicating that ALS patients typically described a mild to moderate disease burden. The ALS and control groups were not significantly different in terms of age and height. Lower limb UMN Score varied from 0 to 6 out of a maximum of 6 (mean 2.6 ± 0.4) and wasting score from 0 to 4 out of a maximum of 6 (mean 1.2 ± 0.3). In terms of specific clinical UMN signs, an extensor plantar response was identified in 8 limbs, and clonus was identified in 3 limbs, but no limbs demonstrated both clonus and an extensor plantar response. The strength of individual lower limb muscles ranged from moderately weak to normal (lower limb MRC sum score 10–20 out of 20, mean 17.6 ± 0.7). As such, the ALS group was heterogeneous in the extent of clinical UMN and LMN signs, and there was generally mild or moderate UMN and LMN involvement clinically. In 58% of ALS limbs, EMG of medial gastrocnemius muscle identified active denervation, defined as the presence of fibrillations, positive sharp waves or fasciculations and neurogenic motor unit changes. Fasciculation potentials without neurogenic motor unit changes were identified in 36% of those limbs without ‘‘active’’ denervation.
In order to assess the effects of upper motor neuron dysfunction on H-reflex parameters, separate regression analyses were undertaken with dependent factors of Hmax/Mmax and slope ratio (Table 1). Changes in both Hmax/Mmax and the slope ratio were closely linked with variation in UMN signs in ALS patients. Hmax/Mmax was predicted by the UMN score (b = 0.47, p < 0.001), but was also predicted by age (b = 0.41, p = 0.001) with increasing age associated with lower values of Hmax/Mmax. The slope ratio was strongly predicted by the UMN score (b = 0.72, p < 0.001), but not by age or Mmax. A significant increase in the ratio of the slope of the initial portion of the H and M recruitment curves has been documented in the affected limb of patients with spasticity (Funase et al., 1996; Higashi et al., 2001). In the present study, the slope ratio was greater in ALS patients than in healthy controls (ALS = 1.07 ± 0.09, control = 0.74 ± 0.08, p = 0.02; Fig. 3A). However the increase was due to a significant reduction of the Mh (ALS = 34.9 ± 2.1°, control = 43.4 ± 2.9°, p = 0.02, Fig. 3B), rather than an increase in Hh (ALS = 34.1 ± 3.2°, control = 32.2 ± 3.9°, p = 0.60; Fig. 3B). Similarly, in UMN disease control group the slope ratio was significantly increased relative to healthy controls (1.18 ± 0.13, p < 0.05) but not ALS patients. In disease controls, neither Hh nor Mh were significantly different from healthy control values, and mean Mh was lower in the disease control than healthy control group (Table 2). Further analysis of the relationship between Hh/Mh and other H-reflex parameters identified a correlation between Mh and the stimulation intensity required to produce the M-wave (Mthresh, Spearman’s rho = 0.62, p < 0.001, Fig. 3C), with a lower slope of the M-wave recruitment curve associated with a lower Mthresh
Please cite this article in press as: Simon NG et al. Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.07.029
5
N.G. Simon et al. / Clinical Neurophysiology xxx (2014) xxx–xxx Table 1 Regression models incorporating H-reflex parameters and upper motor neurone scores. Dependent
Variable
Beta
t
p
Model statistics Adj R2
F
p
Hmax/Mmax
UMNS Age
0.47 0.41
4.10 3.51
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis Neil G. Simon a,b, Cindy S.-Y. Lin b, Michael Lee a, James Howells c, Steve Vucic a,d, David Burke c, Matthew C. Kiernan a,e,⇑ a
Neuroscience Research Australia, Barker St, Randwick, NSW 2031, Australia Prince of Wales Clinical School, University of New South Wales, Randwick, NSW 2031, Australia Sydney Medical School, K25 – Medical Foundation Building, The University of Sydney, NSW 2006, Australia d Westmead Clinical School, C24 Westmead Hospital, The University of Sydney, NSW 2006, Australia e Brain and Mind Research Institute, The University of Sydney, Mallett St, Camperdown, Australia b c
a r t i c l e
i n f o
Article history: Accepted 29 July 2014 Available online xxxx Keywords: Amyotrophic lateral sclerosis Segmental motoneurone Spinal cord circuitry Pathophysiology H-reflex
h i g h l i g h t s H-reflex parameters correlated strongly with clinical upper motor neurone (UMN) dysfunction in ALS
patients. The Hmax/Mmax ratio may be attenuated by increased collision in ALS limbs, reflecting altered dynam-
ics of motor unit recruitment. Alterations of motor unit recruitment related to UMN signs, raising the possibility of transsynaptic
lower motor neurone modulation after UMN damage.
a b s t r a c t Objectives: There is accumulating evidence of dysfunction of spinal circuits in the pathogenesis of amyotrophic lateral sclerosis (ALS). Methods: The present study was undertaken to characterise the pathophysiological changes in segmental motoneuronal excitability in 28 ALS patients, using recruitment curves of the soleus H-reflex and M-wave, compared with clinical assessments of upper motor neuron (UMN) and lower motor neuron dysfunction. Results: H-reflex recruitment curves established that Hmax/Mmax and slope (Hh/Mh) ratios predicted clinical UMN dysfunction (p < 0.001). Changes in Hh/Mh were driven by reduced Mh. Assessment of Hmax/Mmax was similar in the ALS and control groups, and was affected by overlap of the H and M recruitment curves in ALS patients. Conclusion: Changes in the slope ratio (Hh/Mh) in ALS suggested that alterations in peripheral motor nerve excitability following UMN damage may affect the recorded H-reflex. Increased collision of reflex discharges with antidromically-conducted motor impulses may be exacerbated in ALS due to preferential loss of large-caliber a-motoneurones, which may explain the similarities in Hmax/Mmax between groups. Significance: Findings from the present study provide further insight into the pathophysiology of ALS, specifically the relative contributions of premotoneuronal and segmental motoneuronal dysfunction. Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction ⇑ Corresponding author. Address: Brain and Mind Research Institute, The University of Sydney, Room 438, Level 4, M04G, 100 Mallett Street, Camperdown, NSW 2050, Australia. Tel.: +61 2 9114 44251; fax: +61 2 9114 4254. E-mail address: [email protected] (M.C. Kiernan).
Evidence of upper motor neuron (UMN) dysfunction in the context of lower motor neuron (LMN) degeneration is fundamental to the diagnosis of amyotrophic lateral sclerosis (ALS; Brooks et al., 2000). Objective measures of UMN dysfunction may prove useful
http://dx.doi.org/10.1016/j.clinph.2014.07.029 1388-2457/Ó 2014 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Simon NG et al. Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.07.029
2
N.G. Simon et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
to delineate clinical phenotypes, permit monitoring of treatment response, and thereby serve as a measure of disease progression in clinical trials (Hardiman et al., 2011; Kiernan et al., 2011; Turner et al., 2009). However, clinical evidence of UMN involvement may be difficult to detect, particularly when coexistent LMN signs are prominent (Swash, 2012; Turner et al., 2013). There is accumulating evidence of dysfunction of spinal circuits linked to the pathogenesis of ALS (Turner and Kiernan, 2012). Dysfunction of spinal motoneuronal circuits, specifically reduced recurrent inhibition, has been suggested as a mechanism underlying the development of UMN clinical signs in ALS patients (Raynor and Shefner, 1994). Renshaw cells mediate recurrent inhibition of spinal motoneurones and early dysfunction of these cells has been identified in experimental models of ALS (Mazzocchio and Rossi, 2010). In addition, histopathological studies have identified loss of both interneurons and motoneurons in the spinal anterior horn of ALS patients (Stephens et al., 2006). A premise of the present study was that insight into the pathogenesis of ALS may be derived from interrogating segmental motoneuronal excitability and specifically its relation to UMN and LMN abnormalities. As such, the pathophysiological changes in segmental motoneuronal excitability in ALS were evaluated using the H-reflex pathway, as a tool to probe motoneuronal excitability and thereby function (Koelman et al., 1993; Misiaszek, 2003; Pierrot-Deseilligny and Burke, 2012), relating changes to clinical evidence of UMN and LMN dysfunction. A further motivation was to develop an objective approach that could be readily applied in most clinical units to monitor progress in individual patients and any change in response to treatment. 2. Materials and methods Clinical and neurophysiological data were prospectively acquired from a cohort of ALS patients referred to a specialised clinic, each with a diagnosis of clinically probable or definite ALS according to the Awaji-Shima criteria (de Carvalho et al., 2008). Twenty-eight ALS patients were recruited (mean age 62.7 ± 2.0 years, mean height 169.7 ± 3.5 cm; 46% male). Assessments were undertaken prior to the commencement of riluzole in 32% and on riluzole in 68%. A group of 15 age- and height-matched healthy subjects (mean age 57.6 ± 2.3, mean height 166.8 ± 1.6 cm, 47% male) served as a control group for the ALS cohort. Subjects were excluded if there was a prior history or clinical evidence of peripheral neuropathy, lumbosacral radiculopathy, or lumbar spinal surgery. An UMN disease control group comprising 8 patients with spinal cord injury and prominent lower limb UMN clinical signs (mean age 41.2 years, 146–233 days post-injury) were used to compare H-reflex parameters. All subjects provided written informed consent, and the study was approved by the South East Sydney Area Health Service Human Research Ethics Committee. 2.1. Clinical measures Standardized clinical assessments were undertaken in each ALS patient prior to further neurophysiological investigations. Motor functional status and disability was assessed using the revised ALS Functional Rating Scale, with a maximal score of 48 suggesting minimally symptomatic disease (ALSFRS-R; Cedarbaum et al., 1999). The clinical region of onset and symptomatic body regions at the time of clinical review were also recorded, as was height for ALS and control subjects. Muscle strength was assessed using the standard Medical Research Council (MRC) rating scale (Medical Research Council,
1976). Muscle groups graded included shoulder abduction, elbow flexion, wrist extension, finger flexion, finger extension, finger abduction, hip flexion, knee extension, ankle dorsiflexion, and great toe extension, with scores from all groups incorporated into the modified MRC sum score (Burrell et al., 2011). Muscle strength scores from individual lower limbs were added to produce a lower limb MRC sum score. Plantar flexion was not included in the score as there are inherent difficulties in clinically estimating plantar flexion strength with standard manual muscle testing. Weakness of plantar flexion is typically under detected (Lunsford and Perry, 1995). The presence of UMN clinical involvement in each lower limb was assessed using a formal protocol to determine an UMN score (Supplementary Data S1). Knee and ankle reflexes were individually scored on a two-point scale with a score of one given for a brisk reflex without spread or a preserved reflex in a wasted muscle, and a score of two given for a reflex with evidence of pathological spread. A score of one point each was given for the presence of an extensor plantar response (using the Babinski manoeuvre) and ankle clonus (>4 beats) for a maximal score of 6 per lower limb. Measures of muscle tone were not incorporated into the UMN score because their validity and reliability have not been confirmed (Fleuren et al., 2010). The UMN score was specifically devised as a global score of UMN dysfunction in the lower limb, to allow quantification of clinical UMN signs in the presence of concomitant LMN features, and to produce a range of values from low to high to provide robust data for the regression analyses. The global burden of LMN involvement in each lower limb was estimated using a quantitative scale grading muscle wasting in the thigh and leg (Burrell et al., 2011). Each region was graded on a 3 point scale to determine a Wasting Score out of 6 for each lower limb as follows: No wasting = 0 points; trace wasting = 1 point; moderate wasting = 2 points; severe wasting = 3 points. 2.2. Neurophysiological protocols H-reflex studies of the tibial nerve were undertaken with the subject seated in an adjustable armchair, with the hip flexed to 120°, the knee flexed to 160° and the ankle at 110° of plantar flexion (Burke et al., 1999; Lin et al., 2002; McNulty et al., 2008). Subjects were directed to keep their leg relaxed throughout the procedure with visual feedback used to identify any background EMG activity. The tibial nerve was stimulated at the popliteal fossa using bipolar electrodes. H-reflex responses and the soleus compound muscle action potential (M-wave) were recorded from the soleus muscle using Ag/AgCl electrodes (4620M, Unomedical Ltd., Birkerød, Denmark). The active electrode (G1) was placed in the midline of the muscle immediately below the gastrocnemius, and the reference (G2) was placed 4 cm distal to G1. An adhesive electrosurgical grounding plate (2406M, Unomedical Ltd., Birkerød, Denmark) was placed over the proximal fibula. The H-reflex and M-wave recruitment curves were recorded using an automated computerised system (QTRACÓ Institute of Neurology, Queens Square, UK) and a multifunction data acquisition system (NI PCI-6221, National Instruments, Austin, TX, USA). Stimulation was computer controlled and converted to current using an isolated linear bipolar constant-current stimulator (maximal output ± 50 mA; DS5, Digitimer, Welwyn Garden City, UK). Square-wave pulses of 1-ms duration were delivered at a frequency of 0.33 Hz. This stimulation frequency would produce a minor degree of post-activation depression in control subjects (Crone and Nielsen, 1989), but this phenomenon was less prominent in patients with spasticity in the lower limbs (Nielsen et al., 1993), such as might be expected in patients with ALS. The stimulation frequency was specifically selected to minimize the impact
Please cite this article in press as: Simon NG et al. Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.07.029
N.G. Simon et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
3
of post-activation depression while maintaining the tolerability of the procedure for the subjects involved. Responses were amplified (ICP511 AC amplifier, Grass Technologies, West Warwick, RI, USA) with mains frequency noise removed (Hum Bug 50/60 Hz Noise Eliminator, Quest Scientific Instruments, North Vancouver, Canada). Initially, the optimal stimulation position was identified at which the H-reflex was elicited with the lowest possible stimulus intensity. Subsequently, windows were adjusted to capture the latency and duration of the H and M waveforms to determine the maximal peak-to-peak amplitude. The stimulus intensity required to produce the maximal M-wave amplitude was determined (Mmaxint). H and M recruitment curves were then recorded with stimulation commencing at 3% of Mmaxint and increasing in 3% increments thereafter to a maximum of 50 mA. Three stimuli at each intensity level were delivered, and the resulting H-reflex and M-wave amplitudes were averaged to produce each point on the H and M recruitment curves (Fig. 1).
Corporation, Redmond, WA, USA). The following neurophysiological parameters were analysed (Fig. 2): Maximal H-reflex amplitude (Hmax), maximal compound muscle action potential amplitude (Mmax), minimal stimulus intensities to produce a H-reflex (Hthresh) and M-response (Mthresh), stimulus intensity at Hmax (Hmaxint), latency of Hmax (Hlat), latency of Mmax (Mlat), and Hlat minus Mlat (HMlat). All H-reflex and M-wave amplitudes were subsequently normalised to Mmax, and all stimulus intensities were normalised to Mthresh. A line of best fit was applied to the linear ascending component of the H and M recruitment curves using a method adapted from Funase et al. (1994). The linear regression equation was determined and used to calculate the slope angle in degrees (Hh and Mh) using the formula;
2.3. Data analysis
2.4. Statistical analysis
Calculations were performed off-line with data exported to Microsoft ExcelÓ for Mac version 14.2.3, 2011 (Microsoft
All descriptive values were presented as mean ± SEM. Statistical analysis was performed using SPSS version 20, 2011 (IBM, Armonk,
h¼
180
p
tan1 ðmÞ
where m is the slope calculated in the linear regression. Subsequently the slope ratio (Hh/Mh) was calculated.
Fig. 1. The H-reflex circuit and modulating spinal inputs. The H-reflex is comprised of a monosynaptic communication between Ia afferents originating in muscle spindles and aMNs in the spinal anterior horn. The H-reflex is modulated by several local and descending inputs, including other sensory inputs, interneuronal influences and descending spinal tracts. Changes in the physiology of spinal neuronal elements and interneurons have been identified in ALS patients (marked in red) including Renshaw cells (recurrent inhibition), inhibitory interneurons, anterior horn cells and descending tracts. Data from a representative ALS patient are depicted with H and M recruitment curves and corresponding raw data showing changing waveforms with increasing stimulation intensity (a. to e.). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Please cite this article in press as: Simon NG et al. Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.07.029
4
N.G. Simon et al. / Clinical Neurophysiology xxx (2014) xxx–xxx
3.2. Baseline H-reflex parameters in ALS
Fig. 2. H-reflex and M-wave recruitment curves and derived parameters. Representative H-reflex and M-wave recruitment curves demonstrating the parameters measured in the present study.
NY, USA). Shapiro–Wilk tests were used to assess normality of the individual parameters. One-way ANOVA tests were used to compare means between the ALS, healthy control, and disease control groups, and correlation analyses were performed using Pearson product-moment and Spearman’s rank-order tests as appropriate for normally-distributed and not normally distributed data, respectively. Multiple linear regression models were constructed to assess the degree to which differences in Hmax/Mmax and slope ratio were predicted by variation in clinical UMN dysfunction (UMN Score) and age. The model for Hh/Mh also incorporated Mmax. Analyses were preplanned to minimise Type I statistical error.
H-reflexes were recordable in all control limbs and in 92% of ALS limbs (absent in 4 limbs). In the four limbs with absent H-reflexes there was a marked reduction of the soleus M-wave amplitude (mean = 0.48 ± 0.28 mV). In these limbs, UMN signs were modest (UMN score 0 or 1), and LMN signs were more prominent (wasting score 2–4). Each of these limbs was weak (MRC sum score 10–17), and EMG studies identified denervation in gastrocnemius. As such, the limbs with absent H-reflexes were associated with a greater burden of LMN disease, and the absent reflexes could reflect an altered Ia afferent input due to atrophic changes involving intrafusal muscle (Swash and Fox, 1974; Swash, 2012). Mmax was significantly reduced in ALS patients when compared with controls (ALS = 10.2 ± 1.2 mV, control = 15.9 ± 1.2 mV, p = 0.001), and this was still so after exclusion of the data for the 4 limbs in which no H-reflex was recordable. Hmax was also reduced in ALS patients (ALS = 3.8 ± 0.5 mV, control = 6.3 ± 0.9 mV, p < 0.05), such that there was no difference in the mean Hmax/Mmax ratios (0.38 ± 0.04 in ALS, 0.37 ± 0.05 in controls, p = 0.88). When ALS limbs with absent reflexes were excluded from analysis, Hmax/ Mmax was slightly greater (0.43 ± 0.05) but not significantly different from control values. Hlat (ALS = 33.4 ± 0.5 ms, control = 30.9 ± 0.5 ms, p < 0.001), Mlat (ALS = 5.7 ± 0.4 ms, control = 4.7 ± 0.2 ms, p < 0.001) and HMlat (ALS = 27.7 ± 0.4 ms, control = 26.2 ± 0.3 ms, p < 0.05) were all prolonged in ALS patients despite similar mean height. The findings are consistent with these reports of standard motor nerve conduction studies (Feinberg et al., 1999; Mills and Nithi, 1998), and the prolongation of H–M latency difference suggests loss of large, rapidly conducting fibres rather than slower conduction in sprouting terminal axons as a mechanism of the prolongation of H-reflex latency. 3.3. Influence of UMN dysfunction on H-reflex parameters
3. Results 3.1. Clinical characteristics Assessment of disease onset from the present cohort established that their disease started clinically in the upper limb in 39%, bulbar region in 32%, lower limb in 25%, and thoracic region in 4%. The mean duration of disease between the onset of weakness and the study session was 25.8 ± 4.0 months. Mean ALSFRS-R score at the time of assessment was 39.8 ± 1.3 (range 23–48, where 48 represents minimal disability), indicating that ALS patients typically described a mild to moderate disease burden. The ALS and control groups were not significantly different in terms of age and height. Lower limb UMN Score varied from 0 to 6 out of a maximum of 6 (mean 2.6 ± 0.4) and wasting score from 0 to 4 out of a maximum of 6 (mean 1.2 ± 0.3). In terms of specific clinical UMN signs, an extensor plantar response was identified in 8 limbs, and clonus was identified in 3 limbs, but no limbs demonstrated both clonus and an extensor plantar response. The strength of individual lower limb muscles ranged from moderately weak to normal (lower limb MRC sum score 10–20 out of 20, mean 17.6 ± 0.7). As such, the ALS group was heterogeneous in the extent of clinical UMN and LMN signs, and there was generally mild or moderate UMN and LMN involvement clinically. In 58% of ALS limbs, EMG of medial gastrocnemius muscle identified active denervation, defined as the presence of fibrillations, positive sharp waves or fasciculations and neurogenic motor unit changes. Fasciculation potentials without neurogenic motor unit changes were identified in 36% of those limbs without ‘‘active’’ denervation.
In order to assess the effects of upper motor neuron dysfunction on H-reflex parameters, separate regression analyses were undertaken with dependent factors of Hmax/Mmax and slope ratio (Table 1). Changes in both Hmax/Mmax and the slope ratio were closely linked with variation in UMN signs in ALS patients. Hmax/Mmax was predicted by the UMN score (b = 0.47, p < 0.001), but was also predicted by age (b = 0.41, p = 0.001) with increasing age associated with lower values of Hmax/Mmax. The slope ratio was strongly predicted by the UMN score (b = 0.72, p < 0.001), but not by age or Mmax. A significant increase in the ratio of the slope of the initial portion of the H and M recruitment curves has been documented in the affected limb of patients with spasticity (Funase et al., 1996; Higashi et al., 2001). In the present study, the slope ratio was greater in ALS patients than in healthy controls (ALS = 1.07 ± 0.09, control = 0.74 ± 0.08, p = 0.02; Fig. 3A). However the increase was due to a significant reduction of the Mh (ALS = 34.9 ± 2.1°, control = 43.4 ± 2.9°, p = 0.02, Fig. 3B), rather than an increase in Hh (ALS = 34.1 ± 3.2°, control = 32.2 ± 3.9°, p = 0.60; Fig. 3B). Similarly, in UMN disease control group the slope ratio was significantly increased relative to healthy controls (1.18 ± 0.13, p < 0.05) but not ALS patients. In disease controls, neither Hh nor Mh were significantly different from healthy control values, and mean Mh was lower in the disease control than healthy control group (Table 2). Further analysis of the relationship between Hh/Mh and other H-reflex parameters identified a correlation between Mh and the stimulation intensity required to produce the M-wave (Mthresh, Spearman’s rho = 0.62, p < 0.001, Fig. 3C), with a lower slope of the M-wave recruitment curve associated with a lower Mthresh
Please cite this article in press as: Simon NG et al. Segmental motoneuronal dysfunction is a feature of amyotrophic lateral sclerosis. Clin Neurophysiol (2014), http://dx.doi.org/10.1016/j.clinph.2014.07.029
5
N.G. Simon et al. / Clinical Neurophysiology xxx (2014) xxx–xxx Table 1 Regression models incorporating H-reflex parameters and upper motor neurone scores. Dependent
Variable
Beta
t
p
Model statistics Adj R2
F
p
Hmax/Mmax
UMNS Age
0.47 0.41
4.10 3.51
