Characteristics of motor evoked potentials in patients with peripheral vascular disease

TL;DR

This study shows that motor evoked potentials (MEPs) in patients with peripheral vascular disease (PVD) are similar to healthy individuals, suggesting PVD does not affect MEP monitoring during spinal surgeries.

Contribution

The study provides the first data on MEP characteristics in older patients with PVD, showing that PVD does not impair MEPs.

Findings

MEP amplitudes were similar between PVD patients and healthy volunteers for most muscles.

Disease severity in PVD patients did not correlate with MEP amplitude, except for the vastus lateralis.

MEP latencies were longer in older participants but not different between PVD patients and healthy volunteers for a given muscle.

Abstract

With an aging population, it is common to encounter people diagnosed with peripheral vascular disease (PVD). Some will undergo surgeries during which the spinal cord may be compromised and intraoperative neuromonitoring with motor evoked potentials (MEPs) is employed to help mitigate paralysis. No data exist on characteristics of MEPs in older, PVD patients, which would be valuable for patients undergoing spinal cord at-risk surgery or participating in neurophysiological research. Transcranial magnetic stimulation, which can be delivered to the awake patient, was used to stimulate the motor cortex of 20 patients (mean (±SD)) age 63.2yrs (±11.5) with confirmed PVD, every 10 minutes for one hour with MEPs recorded from selected upper and lower limb muscles. Data were compared to that from 20 healthy volunteers recruited for a protocol development study (28yrs (±7.6)). MEPs did not differ between patient’s symptomatic and asymptomatic legs. MEP amplitudes were not different for a given muscle between patients and healthy participants. Except for vastus lateralis, disease severity did not correlate with MEP amplitude. There were no differences over time in the coefficient of variation of MEP amplitude at each time point for any muscle in patients or in healthy participants. Although latencies of MEPs were not different between patients and healthy participants for a given muscle, they were longer in older participants. The results obtained suggest PVD alone does not impact MEPs; there were no differences between more symptomatic and less symptomatic legs. Further, in general, disease severity did not corelate with MEP characteristics. With an aging population, more patients with PVD and cardiovascular risk factors will be participating in neurophysiological studies or undergoing surgery where spinal cord integrity is monitored. Our data show that MEPs from these patients can be easily evoked and interpreted.

Figures

A) Illustration of circular coil placement over vertex, B) Electrode placement on PVD patients with cone coil shown for illustration.Ground electrode over olecranon process omitted. L = Left, R = Right, BR = brachioradialis, APB = abductor pollicis brevis, VL = vastus lateralis, TA = tibialis anterior and AH = abductor hallucis.

AvMEP amplitude (mean ± SD) for A) patient lower limb muscles VL, TA and AH comparing symptomatic (black bars) and asymptomatic (white bars) legs, and B) upper limb muscles BR and AHB and lower limb muscles VL, TA and AH comparing patients (black bars) and healthy participants (grey bars).NS represents no significant difference.

CV of AvMEP amplitude (mean ± SD) for A) patient lower limb muscles VL, TA and AH comparing symptomatic (black bars) and asymptomatic (white bars) legs, and B) upper limb muscles BR and AHB and lower limb muscles VL, TA and AH comparing patients (black bars) and healthy participants (grey bars).NS represents no significant difference.

AvMEP latency (mean ± SD) for A) patient lower limb muscles VL, TA and AH comparing symptomatic (black bars) and asymptomatic (white bars) legs, and B) upper limb muscles BR and AHB and lower limb muscles VL, TA and AH comparing patients (black bars) and healthy participants (grey bars).NS represents no significant difference.

CV of AvMEP latency (mean ± SD) for A) patient lower limb muscles VL, TA and AH comparing symptomatic (black bars) and asymptomatic (white bars) legs, and B) upper limb muscles BR and AHB and lower limb muscles VL, TA and AH comparing patients (black bars) and healthy participants (grey bars).NS represents no significant difference.

CV (mean ± SD) of MEP amplitude at each 10-minute time point for A) patients and B) healthy participants.BR = brachioradialis, APB = abductor pollicis brevis, VL = vastus lateralis, TA = tibialis anterior and AH = abductor hallucis.

Full text

Introduction

Motor evoked potentials (MEPs) are used extensively as a research tool to explore neuromuscular function in health and disease [1–3]. Changes in characteristics of MEPs have been observed in many neurological and neuromuscular pathologies, such as stroke and muscular dystrophies [3, 4] and for intra-operative neuromonitoring (IONM) to diagnose and mitigate spinal cord injury during spinal surgery or thoraco-abdominal aortic aneurysm (TAAA) repair [5, 6].

Patients undergoing spinal surgery and TAAA repair are often older and frequently have cardiovascular risk factors such as high blood pressure, hypercholesterolaemia and smoke [7, 8]. As such, they frequently have concomitant peripheral vascular disease (PVD) in addition to their presenting pathology, where the arteries of the lower limbs are atherosclerotic, leading to occlusion and subsequent chronic limb ischaemia (CLI) [9]. This in turn leads to tissue loss and damage of the limb nerves and muscles.

MEP characteristics have been exhaustively studied in heathy volunteers, and in specific spinal or neurological disease [10–13]. The MEPs of PVD patients, however, have not previously been examined. Understanding changes that occur due to CLI is not only beneficial for IONM, but for any research or clinical situation where MEPs need to be characterised in patients with cardiovascular risk factors or recognised PVD. With an aging population and increasing rates of obesity and cardiovascular disease [14], the co-incidence of PVD will inevitably increase.

Current standards of IONM utilise MEPs generated by transcranial electrical stimulation (TES) [15]. However, TES in the awake participant can be prohibitively painful. Transcranial magnetic stimulation (TMS) is an alternative technique for generating MEPs, with the benefit of being pain-free, [16] thus it is frequently employed in neurophysiology research [17]. Further, MEP changes seen pre- and intra-operatively with TMS, correlate with TES-induced MEP changes during surgery and clinically after surgery [17, 18]. TMS is therefore suitable to characterise MEPs in non-anaesthetised PVD patients in the research and surgical settings.

The aim of this study was to determine the characteristics of TMS-induced MEPs from limb muscles in PVD patients and compare these to data from participants without cardiovascular risk, or PVD, which formed part of a protocol development study.

To our knowledge, this not been previously studied and would provide essential baseline data for future neurophysiological study of patients with cardiovascular disease and risk factors, with possible clinical applications, including intraoperative and post-operative monitoring.

Material & methods

Participants

20 patients with PVD (17 males), mean (±SD) age of 63.2 (±11.5) years took part in the study. All patients were symptomatic with radiologically confirmed PVD under the care of a vascular surgical team. Data were also collected prior to this study from 20 participants (8 females; 28 (±7.6) years of age) with no co-morbidities or cardiovascular risk factors. Recruitment occurred between 15^th^ November 2016 and 20^th^ November 2018.

A TMS safety questionnaire was undertaken to ensure appropriateness for inclusion [19]. The inclusion and additional PVD-specific exclusion criteria for patients are listed in Table 1.

Table 1: Patient-specific inclusion and exclusion criteria.

This study was approved by the Imperial College London research ethics committee (number: 16IC3553) and National Health Service health regulatory authority (number: 17/LO/1034) and performed with adherence to the Declaration of Helsinki guidance on conducting research in human participants. Written informed consent was obtained from all participants.

Disease severity questionnaire (VascuQol)

Patients completed the King’s College Hospital’s Vascular Quality of Life Questionnaire (VascuQol); this validated questionnaire determines disease severity with 25 questions exploring the effects of PVD on physical and emotional well-being of patients [20]. Each question is graded from 0 to 7, with 0 indicating PVD affects a particular aspect of their life all of the time (i.e. no ability to perform task). A mean score of 4.5 or less is associated with symptomatic intermittent claudication due to CLI [21].

Electromyography

With participants supine, pairs of disposable Ag/AgCl electrodes (25mm diameter, 1041PTS, Henleys Medical Supplies Ltd., UK) were applied to the skin 20mm apart, overlying the muscles of interest, after skin preparation with alcohol [22]. To explore potential differences in disease severity between the left and right leg, patients had electrodes placed over left and right vastus lateralis (VL), tibialis anterior (TA) and abductor hallucis (AH), as well as left brachioradialis (BR) and left abductor pollicis brevis (APB), see Fig 1. A ground electrode was placed over the ulna olecranon process. These limb muscles were chosen as they are supplied by different major peripheral nerves.

A) Illustration of circular coil placement over vertex, B) Electrode placement on PVD patients with cone coil shown for illustration.Ground electrode over olecranon process omitted. L = Left, R = Right, BR = brachioradialis, APB = abductor pollicis brevis, VL = vastus lateralis, TA = tibialis anterior and AH = abductor hallucis.

Raw EMG signals were amplified (x1000), band-pass filtered (10Hz–1kHz) (ISO Dam, WPI, UK) and sampled (at 2kHz) using an analogue to digital converter (Power 1401, Cambridge Electronic Design [CED], UK). Signal software (CED, UK) was used to record to disk for subsequent analysis.

Transcranial magnetic stimulation

TMS was delivered using two Magstim 200^2^ magnetic stimulators (Magstim, Dyfed, UK), each connected to a different coil, which was positioned over the vertex. The location of the vertex was determined as the point on the scalp at which two lines, from nasion to inion and from tragus to tragus, bisect; this point was marked with an indelible pen to ensure consistent coil placement. A circular coil (90mm outer diameter), which is more commonly used to stimulate many upper limb muscles simultaneously, was positioned B-side up to induce anticlockwise current in the brain, thus preferentially activating the right hemisphere [23]. An angled double-cone coil (wing outer diameter 120mm), which is capable of stimulating deep into the motor cortex due to its construction and geometry, was employed to stimulate lower limb muscles. It was positioned over the vertex so that the induced current in the brain flowed in the posterior-to-anterior direction. Two ‘combined’ resting motor thresholds (cRMT) were determined for the upper and lower limb muscles for each participant, with the circular and cone coils, respectively. This was the lowest stimulation intensity that produced MEPs ≥50μV in amplitude [24] in at least three out of six stimuli in either all upper or lower limb muscles. Stimulator intensities of 120% of the cRMT for the muscle with the highest threshold were used [25]; these are shown in Table 2 for both study groups. For a given coil, there were no differences between intensities used between healthy participants and patients; circular coil P = 0.62, cone coil P = 0.19.

Table 2: Stimulation intensities (mean maximum stimulator output (%MSO) and range) for upper and lower limb muscles for healthy volunteers and patients.

Each block of stimulation consisted of six single stimuli from each coil, every 10 minutes over the course of 60 minutes (seven sets of twelve stimuli in total from 0 to 60 minutes). Participants were instructed to relax all muscles during stimulation to prevent facilitation [26].

Data analysis

Patient disease severity questionnaire (VascuQol)

The scores from 25 questions were averaged to give a mean value from 0 to 7. Microsoft Excel 2016 (Microsoft, USA) was used to perform calculations.

Motor evoked potentials

Signal Software (CED, UK) was used for data analysis by visual positioning of cursors at the start and finish of each MEP. Peak-to-peak amplitude and latency (time from stimulation to first MEP deflection) were subsequently measured.

Given the known variability of TMS-induced MEPs in relaxed muscles [27] the variability both within a time point and across the whole experiment was calculated. The variability between measurements was determined by the coefficient of variation (CV); CV = standard deviation/mean.

Variability within a time point. To determine the variability at a given time point, the amplitude of each MEP was measured, and the CV of the 6 MEPs calculated. The CV of the MEP amplitude was calculated for each time point and for each muscle.

Variability across the whole experiment. The averaged MEP (avMEP) is a digital averaging process, where the 6 MEPs at a given time point are amalgamated to produce a single MEP; from this the amplitude and latency are measured.

To determine the variability of MEPs across the duration of the experiment, the amplitude of the avMEP at each of the 10-minute time points was measured and the overall CV calculated. This was performed for each muscle and each participant to calculate the mean CV of the avMEP amplitude; comparisons over time for a given muscle were performed.

Statistical analysis

Statistical analyses were carried out using GraphPad Prism v10.1.2 (GraphPad Software LLC, Boston, MA, USA). Data were tested for normality (Shapiro-Wilk) and appropriate tests (non-parametric) were performed with post-hoc multiple comparisons with Dunn’s corrected P values reported.

AvMEP amplitudes, CVs of the avMEP amplitudes, AvMEP latencies and CVs of the avMEP latencies were collapsed over the hour and compared between patient-reported symptomatic and asymptomatic sides in patients. Where both legs were symptomatic, data were evenly split between groups to allow equal group sizes for analysis. No differences were found between symptomatic and asymptomatic legs with respect to these parameters, hence data were averaged between legs and compared between patients and healthy participants using Kruskal-Wallis tests.

To examine changes in variability of MEP amplitudes within a time point over the hour protocol, CV of MEP amplitudes were compared over time and between symptomatic and asymptomatic lower limb muscles using Kruskal-Wallis tests. Since no differences were found between symptomatic and asymptomatic legs, data were averaged between legs. To determine relationships between clinical score (VascuQol) and MEP characteristics in patients and between age and MEP characteristics across all participants, and between avMEP amplitudes and CV of avMEP amplitudes, Spearman’s correlations were performed.

The level of statistical significance was set at P<0.05, or appropriate for multiple comparisons. Data are presented as mean ± SD in the text and figures.

Results

VascuQol score in PVD group

The VascuQol score was 4.3±1.2, representing symptomatic PVD not requiring surgical intervention (as per the main inclusion criteria). This score indicates patients experience cramp-like pain in the lower leg when walking due to muscle ischaemia.

Amplitude of avMEP

There were no differences in avMEP amplitudes between symptomatic and asymptomatic legs in patients (VL, 0.21±0.19mV vs 0.36±0.41mV, P>0.99; TA, 0.66±0.62mV vs 0.54±0.30mV, P>0.99; AH, 0.75±0.60mV vs 0.73±0.37mV, P>0.99), see Fig 2A; data were therefore averaged between sides.

AvMEP amplitude (mean ± SD) for A) patient lower limb muscles VL, TA and AH comparing symptomatic (black bars) and asymptomatic (white bars) legs, and B) upper limb muscles BR and AHB and lower limb muscles VL, TA and AH comparing patients (black bars) and healthy participants (grey bars).NS represents no significant difference.

There were no differences between avMEP amplitudes of patients and healthy participants for a given muscle (Fig 2B).

CV of avMEP amplitude

There were no differences in CV of avMEP amplitudes between symptomatic and asymptomatic legs in patients (VL, 0.38±0.21 vs 0.26±0.10, P>0.99; TA, 0.32±0.20 vs 0.28±0.14, P>0.99; AH, 0.22±0.11 vs 0.20±0.10, P>0.99), see Fig 3A; data were therefore averaged between sides.

CV of AvMEP amplitude (mean ± SD) for A) patient lower limb muscles VL, TA and AH comparing symptomatic (black bars) and asymptomatic (white bars) legs, and B) upper limb muscles BR and AHB and lower limb muscles VL, TA and AH comparing patients (black bars) and healthy participants (grey bars).NS represents no significant difference.

There were no differences between CV of avMEP amplitudes of patients and healthy participants for a given muscle (Fig 3B).

Latency of avMEP

There were no differences in avMEP latencies between symptomatic and asymptomatic legs in patients (VL, 25.62±3.00ms vs 25.63±3.05ms, P>0.99; TA, 31.74±4.04ms vs 30.93±3.94ms, P>0.99; AH, 44.37±6.50ms vs 43.78±5.88ms, P>0.99), see Fig 4A; data were therefore averaged between sides.

AvMEP latency (mean ± SD) for A) patient lower limb muscles VL, TA and AH comparing symptomatic (black bars) and asymptomatic (white bars) legs, and B) upper limb muscles BR and AHB and lower limb muscles VL, TA and AH comparing patients (black bars) and healthy participants (grey bars).NS represents no significant difference.

There were no differences between avMEP latencies of patients and healthy participants for a given muscle (Fig 4B).

CV of avMEP latency

There were no differences in CV of avMEP latency between symptomatic and asymptomatic legs in patients (VL, 0.04±0.04 vs 0.03±0.03, P>0.99; TA, 0.03±0.03 vs 0.03±0.03, P>0.99; AH, 0.03±0.04 vs 0.02±0.02, P>0.99), see Fig 5A; data were therefore averaged between sides.

CV of AvMEP latency (mean ± SD) for A) patient lower limb muscles VL, TA and AH comparing symptomatic (black bars) and asymptomatic (white bars) legs, and B) upper limb muscles BR and AHB and lower limb muscles VL, TA and AH comparing patients (black bars) and healthy participants (grey bars).NS represents no significant difference.

There were no differences between CV of avMEP amplitudes of patients and healthy participants for a given muscle (Fig 5B).

Mean CV of MEP amplitude at each time point

There were no differences over time for any muscle in either the patients or the healthy participants (P>0.99 in all cases), see Fig 6.

CV (mean ± SD) of MEP amplitude at each 10-minute time point for A) patients and B) healthy participants.BR = brachioradialis, APB = abductor pollicis brevis, VL = vastus lateralis, TA = tibialis anterior and AH = abductor hallucis.

Correlations

VascuQol score and MEP amplitudes

There was no correlation between VascuQol score and MEP amplitudes (ρ range -0.37 to 0.07, P>0.05 in all cases) for any muscle, except VL (ρ -0.78, P<0.001). For VL, disease severity score was inversely proportional to MEP amplitude, i.e. worse patient-reported symptoms were associated with larger MEP amplitudes.

VascuQol score and MEP latency

There was no correlation between VascuQol score and MEP latency (ρ range -0.41 to -0.08 P>0.05 in all cases).

Age and avMEP amplitude

There was no correlation between age and avMEP amplitudes (ρ range- 0.31 to -0.10, P>0.05 in all cases) for any muscle except BR (ρ -0.55, P<0.001), where increases in age were associated with smaller avMEP amplitudes.

Age and avMEP latency

All muscles showed a significant positive correlation between increasing age and avMEP latency (ρ range 0.39 to 0.61, P<0.05 in all cases), i.e. latencies were longer in older participants.

avMEP amplitude and CV of avMEP amplitude

There were variable relationships between avMEP amplitudes and CVs of avMEP amplitudes. Significant relationships were found for APB and AH in healthy participants and for VL and TA in patients. The slopes of the significant correlations were always negative, i.e. increased avMEP amplitudes were associated with smaller CVs (APB healthy participants, ρ—0.47, P = 0.04; AH healthy participants ρ—0.54, P = 0.01; VL patients, ρ—0.52, P = 0.02; TA patients ρ—0.50, P = 0.02).

Discussion

The results of this study have demonstrated MEP characteristics of patients with PVD. When compared to healthy, younger participants, they appear to be similar, with differences likely attributable to age rather than disease. In the following, these findings will be discussed further, along with their potential impact on future neurophysiological research and clinical application.

MEP amplitude and latency

The data from this study showed no differences in the MEP amplitude and latency between PVD patients and healthy participants. When MEP amplitudes and latency were correlated with age for all participants, only BR showed a negative correlation with age, whilst all muscle latencies increased with increasing age. Although the healthy participants were not aged matched to PVD patients, this cohort was younger than PVD patients. Therefore, any differences observed appear to be due to the effects of aging, rather than due to pathology. In support of this, no differences in MEP amplitudes or latencies were seen between the symptomatic (presumably more diseased) and asymptomatic legs of PVD patients, allowing the data to be averaged.

Aging can have variable effects on MEP characteristics, with results in the literature being inconsistent. Some studies which compared healthy younger participants with matched healthy older participants, have found both at rest and during contraction, older participants have smaller MEP amplitudes with longer latencies [28, 29], the latter in keeping with the results of the current study. However, other studies have demonstrated that the effect of age on MEP amplitudes can vary between muscles [30–32]. One study, for example, showed input-output curves (stimulus intensity vs MEP amplitudes) of MEPs were different in an upper limb muscle, the first dorsal interosseous muscle, between younger and older adults but were the same for a lower limb muscle, vastus lateralis [31]. This is consistent with our data, where with increasing age, BR (an upper limb muscle) MEP amplitude appears to decrease, whilst no such difference was seen in the amplitude of lower limb muscle MEPs.

Further, disease severity scores showed no correlation between more severe disease and effects on MEP characteristics. A single significant correlation was found between the disease severity score and avMEP amplitude; higher VascuQol scores were associated with smaller VL amplitude. The VascuQol is a functional scoring system, assessing the patient’s ability to perform common daily activities, such as bathing; a maximum score of 7 would indicate ‘maximum’ function with no impact on function due to PVD. Therefore, with more severe disease (i.e. a lower VascuQol score), one would predict smaller MEPs secondary to potential tissue damage. No such correlations were found, and the single negative correlation is difficult to explain physiologically. It is therefore proposed symptomatic PVD, which impairs activities of daily living, appears to have little impact on limb muscle MEP characteristics.

The protocol was designed to include patients who had stable disease with symptomatic leg pain on exercise, not severe enough to require surgery; the mean VascuQol score reflects this. These patients have an arterial blood supply to their lower limbs which is adequate provided no exertional stress is placed upon them; there is a state of intermittent ischaemia. Based on our results, the severity of PVD in our patient cohort, and the severity of the ischaemia, does not seem to result in any deterioration of nerve function. The normal MEP characteristics is especially surprising given the muscle atrophy and sarcopenia that is observed many PVD patients, who find the claudication pain limits even simple daily tasks and physical activity [33]. It may be that patients with more severe PVD, such as those with critical limb ischaemia presenting with pain at rest and tissue necrosis, or sudden acute limb ischaemia following limb trauma for example, may have nerve dysfunction which manifests in a reduction in MEP amplitude or greatly prolonged latency. It has been shown experimentally that acute limb ischaemia following a tourniquet-induced ischaemic nerve block results in smaller MEPs with increased latencies [34, 35]. To the best of our knowledge however, the impact of moderate PVD on MEP characteristics has not been investigated before and provides valuable baseline data for older patients with known cardiovascular morbidity, with or without diagnosed PVD (given identical risk factors) undergoing neurophysiological investigation or surgery where IONM is employed.

We were aware of the potential effect of MEP amplitude on the variability of MEPs within a muscle. The design of our protocol meant there was a disparity between the relative stimulation intensity for each muscle given a single ‘combined’ test intensity per coil was used. A stimulus intensity of 120% of motor threshold of the muscle with the highest threshold was used; therefore, the intensity would be greater for the remaining muscles with lower motor thresholds. Previous work has shown greater stimulation intensities result in MEPs with not only larger amplitudes but also less variability, likely through consistent activation of most of the available motor units [25, 36, 37]. However, our results showed no consistent significant correlations between the mean CV of avMEP and the mean amplitude of avMEP for either group. However, significant correlations were always in the expected direction, with larger MEP amplitudes being associated with smaller CVs. Given the differences in strength of the circular coil and the cone coil, it is not surprising that the stimulus intensities used for the circular and cone coils were equivalent. However, the results also reveal that for a given coil, the intensities used for the two groups were not different, demonstrating that the thresholds for the patients were not higher.

Amplitude is a key factor to take into consideration when considering the use of MEPs for clinical purposes. A sudden reduction in amplitude or loss of MEPs is highly suggestive of injury during spinal surgery [38]. Muscles with small MEP amplitudes can make it difficult to detect injury-induced decreases. Furthermore, the depressive effects anaesthetic agents significantly reduce MEP amplitudes and prolong latencies [18, 39]. In many IONM guidelines, usually a hand and foot muscle are used as control and monitoring muscle, respectively [40]. The results of this study suggest (Fig 2B) that MEPs in APB (hand) and AH (foot) in PVD patients produced the largest mean amplitudes of the upper and lower limb muscles, respectively, similarly observed in healthy participants; thus they would likely be suitable for monitoring purposes in the PVD population as per standard IOMN guidelines [40].

Reliability

Many neurophysiological studies have investigated the variability of TMS-induced MEPs [10, 27, 36, 41–43]. No previous study to our knowledge has compared the variability of different muscles over a prolonged time period. As can be seen from the current data, it is evident that there is variability in TMS-induced MEP characteristics between stimuli at each epoch and over time. Our interest lay in the extent of this variation over time within a given muscle.

During surgeries where the spinal cord is vulnerable, there will be certain surgical interventions or time points where there is a higher SCI risk. At these crucial timepoints, there will be a need for repeated trains of stimuli to continually assess spinal cord integrity, hence the need to assess variability within trains at a given time point (CV of mean MEP). Determining the variability of TMS-induced MEPs over time is also necessary particularly where frequent but intermittent monitoring is required for a longer duration of time, hence the need to assess variability over an extended period (CV of avMEP). This was achieved with a protocol recording trains of six MEPs at regular intervals over a 1-hour period.

TMS-induced MEPs have been shown to vary in amplitude [37]; this variability is greater with TMS than with TES, since motor cortical activation occurs largely trans-synaptically [27, 44]. TES induces MEPs through direct activation of corticospinal neurons and generates more consistent results [45]. It is for this reason that surgical neuromonitoring guidelines usually recommend TES [15]. However, there are practical disadvantages to its use. TES routinely utilises invasive needle or screw stimulating electrodes, compared to the non-invasive hand-held stimulating coil employed for TMS use. The use of high current stimulus intensities of TES is also very painful, usually restricting clinical use to the anaesthetised patient. This prevents interrogation of the motor pathways post-operatively in the lightly sedated or awake patient who may develop late SCI sometime after a TAAA repair; as many as 19% of patients develop paralysis 24 hours after this surgery, outside of the operating theatre, usually when spinal cord monitoring has ended [46].

Our results have shown that the CVs of the mean avMEP amplitude are comparable in patients and healthy participants. The mean CV of the avMEP amplitude gives an indication of the variation of all MEPs generated in each muscle over the course of the 1-hour protocol. This is especially true for BR and AH, which had the lowest CVs of the avMEP amplitude for upper and lower limb muscles in both groups of participants. These results are in keeping with previous studies, showing TMS-related measurements have good intra- and inter-investigator [41] and test-retest [43] reliabilities in healthy young and old participants [10, 37, 47]. Given there were no significant differences between healthy participants and PVD patient CV values in this study, the variability of normative MEP data in the literature where predominantly young, healthy participants are included, should be applicable to patients with cardiovascular disease or risk factors participating in research or neurophysiological testing.

Some of the variability in MEPs could be attributed to methodology [23, 48, 49], such as stimulus intensity or number of stimuli delivered, [10] level of muscle contraction at time of stimulation, [26, 50] data analysis procedures including removal of initial MEPs when averaging [36] and technical factors relating to coil position [51, 52]. To limit this, we delivered trains of 6 stimuli for our averaging calculations, maintained strict coil orientation and position, performed stimulation at rest (as confirmed by pre-stimulus EMG levels) and removed only data where there was muscle contraction. As a result, MEP measures in this investigation, including variability, are comparable to previously published work [27, 37, 53].

Limitations

The authors acknowledge the comparison with data from younger healthy volunteers, rather than healthy age-matched participants, may be viewed as a limitation. The healthy volunteer data were generated from a protocol development study, undertaken before the current study. The comparison presented here is to demonstrate differences between diseased and disease-free participants. PVD is a disease of aging and associated with lifestyle factors, with the atherosclerotic process increasing with age; age is an independent risk factor, irrespective of health status [54]. Thus, it would be difficult to find an older person who would be truly disease-free and “healthy” with respect to their arterial wall integrity. Confirmation would require advanced functional assessments and blood flow imaging studies, [55] which would have been practically prohibitive for this study. Data exist for healthy older participants and our data are compared to the literature in the discussion.

The use of a cRMT results in a stimulation intensity of 120% for the muscle with the highest motor threshold, but greater for the other muscles. It is convention to use a specific coil and test intensity to target each muscle but this would have been impractical in the confines of a time critical protocol. However, previous work has shown only small differences in motor threshold between limb muscles, [3, 56] thus the use of a single test intensity is unlikely to have a major impact on the results obtained.

In many neurophysiological studies, MEP amplitudes are often normalised to the maximum M-waves (mMax), the maximal EMG response to direct motor nerve stimulation; [57] this allows comparisons to be made across muscles and between participants with different MEP amplitudes. However, since a key focus of the current investigation was variability of MEPs over time (as assessed by CV), normalisation to mMax was not deemed necessary as CV is a form of normalisation. Further, due to the time critical nature of this study design, it was not feasible to include measurement of MMax for all muscles studies.

Conclusions & recommendations

MEPs were recordable from upper and lower limb muscles in PVD patients and were found to have similar characteristics to data from our healthy participants. Our results suggest there are no additional effects of pathological limb ischaemia on MEP characteristics beyond those expected by aging. With an increasingly aging population, there will be greater numbers of patients with PVD and cardiovascular risk factors participating in neurophysiological studies or undergoing surgery where spinal cord integrity is monitored [58] and MEPs from these patients can be easily evoked and interpreted.