- Email: [email protected]
Single fiber electromyography
Handbook of Clinical Neurology, Vol. 160 (3rd series) Clinical Neurophysiology: Basis and Technical Aspects K.H. Levin and P. Chauvel, Editors https://doi.org/10.1016/B978-0-444-64032-1.00019-9 Copyright © 2019 Elsevier B.V. All rights reserved
Chapter 19
Single fiber electromyography VERN C. JUEL* Department of Neurology, Duke University School of Medicine, Durham, NC, United States
Abstract Single fiber electromyography (SFEMG) is a highly selective technique that permits assessment of individual muscle fiber action potentials (MFAPs). This selectivity is achieved with a specialized concentric needle electrode with a 25-mm diameter recording surface located in a side port 3 mm from the needle tip. Additional selectivity is achieved with 500-Hz low-frequency filtering. An oscilloscope with a trigger and delay line enables identification of time-locked MFAPs within the same motor unit. SFEMG techniques allow assessment of two important features of the motor unit: jitter and fiber density (FD). Neuromuscular jitter is a direct measure of neuromuscular transmission and reflects the temporal variation in end-plate potentials reaching threshold to elicit a MFAP. SFEMG may be used to assess paired jitter with voluntary activation or by axonal stimulation of motor nerve branches to individual end plates. SFEMG is the most sensitive clinical test for neuromuscular junction disease and is often abnormal in clinically unaffected muscles in patients with myasthenia gravis (MG) and Lambert–Eaton myasthenia (LEM). Normal jitter findings in a clinically weak muscle exclude neuromuscular junction disease as a cause for weakness in that muscle. FD measurements assess the local concentration of muscle fibers within a motor unit and provide a sensitive in vivo assessment of reinnervation.
INTRODUCTION
THE SINGLE FIBER ELECTRODE
Single fiber electromyography (SFEMG) is a highly selective technique that facilitates assessment of individual muscle fiber action potentials (MFAPs). This selectivity is achieved by use of a specialized concentric electrode with a very small recording surface in conjunction with recording equipment that filters low-frequency signals from distant muscle fibers. With SFEMG, MFAPs innervated by the same motor neuron can be recorded to provide valuable information regarding the structure and function of motor units. SFEMG is used to assess neuromuscular transmission through measurement of jitter and impulse blocking. Fiber density (FD) measurement by SFEMG may reveal early evidence for reinnervation in motor neuropathic processes.
The selectivity of the single fiber electrode is largely due to its small 25-mm diameter recording surface located in a side port 3 mm from the needle tip (Fig. 19.1). The small recording surface area (0.0005 mm2) of the single fiber electrode results in less shunting of electrical field from adjacent muscle fibers compared with the larger recording surfaces of conventional concentric needle electrodes (0.07 mm2) and of facial concentric needle electrodes (0.019 mm2) (Stålberg and Sanders, 2009). This selectivity is reflected in the marked radial decline in signal amplitude seen with increasing distance from the recording surface with single fiber electrodes, in comparison to conventional concentric, monopolar, and macro EMG electrodes (Fig. 19.2) (Stålberg, 1983).
*Correspondence to: Vern C. Juel, M.D., 40 Duke Medicine Circle, Clinic 1L, Room 1255, DUMC 3403, Duke University Medical Center, Durham, NC 27710, United States. Tel: +919-684-4044, Fax: +919-660-3853, E-mail: [email protected]
304
V.C. JUEL
Fig. 19.1. SFEMG electrode and needle holder. The selectivity of SFEMG is attributed primarily to the 25-mm diameter recording surface at the side port 3 mm from the needle tip. Copyright © 2011, DB Sanders.
1000
AMPL (mV)
800
600
400 SF el
200 conc el macro el
0 0
400
800
1200
1600
2000
2400
Distance (mm)
Fig. 19.2. Radial decline of signal amplitude over distance with single fiber, concentric, and macro electrodes. Note the sharp decline in the signal amplitude with the single fiber electrode. From Sta˚lberg EV (1983). Macro EMG. Muscle Nerve 6: 619–630, with permission from John Wiley and Sons.
INSTRUMENTATION The selectivity afforded by the single fiber electrode is enhanced by the use of a 500-Hz low-frequency filter to reject signals from distant muscle fibers. A highfrequency filter setting of 10 kHz is also used. Amplifier gain settings of 0.2–0.5 mV/division are optimal for recording MFAPs; higher settings introduce more background noise into the recordings. An oscilloscope sweep speed of 0.5 ms/division facilitates assessment of the temporal variability of the MFAP signals.
A trigger and delay line is used to identify signal components that precede and follow the triggered signal. Voltage- or amplitude-level triggers have traditionally been used in SFEMG recording and initiate the oscilloscope sweep when a potential exceeds the voltage level set by the operator. With the peak detection triggering technique, the operator also sets a voltage level, and the peaks of the triggering and nontriggering potentials are extrapolated from the intersection of computercalculated slopes of the rising and falling phases of the MFAPs. While the techniques produce comparable jitter values, the peak detection method is less affected by artifactually increased jitter values related to overlapping or “riding” signals or to baseline variability (Stålberg and Sanders, 2009).
FIBER DENSITY MEASUREMENTS FD measurements assess the local concentration of muscle fibers within a motor unit. Increased FD is analogous to histopathologic muscle fiber type grouping seen following reinnervation from collateral sprouting of motor nerves. When measured with a single fiber electrode, the amplitude of an MFAP falls to 200 mV when the electrode is 300 mm from the muscle fiber. Accordingly, MFAPs with amplitudes >200 mV and rise times <300 ms are generated by muscle fibers within 300 mm of the recording surface (Stålberg et al., 2010). FD assessment must be performed with actual single fiber electrodes, as the recording characteristics of conventional concentric needles do not reliably permit signal recordings from individual muscle fibers. To assess FD, the single fiber electrode is positioned to maximize the amplitude of a single MFAP. After triggering on this potential, the mean number of timelocked potentials (with amplitude >200 mV and rise time <300 ms) observed in this recording position is counted. In addition to the triggering potential, all clear notches involving the potentials are included in the total count of MFAPs. A total of 20 different sites within the muscle are sampled via three to four separate needle insertions. The mean FD is then calculated for the muscle and compared to muscle and age-matched normal values (Gilchrist et al., 1992; Bromberg and Scott, 1994).
JITTER In the 1960s, Stålberg and Ekstedt originally observed variability in signal latencies recorded with novel, custom electrodes designed to record potentials from individual muscle fibers. Although this variability was initially attributed to a defective oscilloscope (hence the term jitter, referring to signal deviations from true periodicity in an electronic device), it became clear that this variability was related to neuromuscular
SINGLE FIBER ELECTROMYOGRAPHY transmission following recordings from a patient with myasthenia gravis and subsequent experiments with curare (Stålberg et al., 2010). Neuromuscular jitter is a direct measure of the effectiveness of neuromuscular transmission and represents the temporal variation for muscle end-plate potentials (EPPs) to reach threshold to produce an MFAP. Neuromuscular junction disease lowers the safety factor by reduced quantal release of acetylcholine (ACh) in presynaptic disease or by reduced end-plate responsiveness to ACh in postsynaptic disorders. In either case, increased jitter occurs when EPPs are delayed in reaching threshold, and impulse blocking occurs when EPPs fail to elicit a MFAP. Clinical weakness in a given muscle occurs when a critical number of end plates are blocked in that muscle. In SFEMG, jitter is measured as the variability in the arrival time of individual MFAPs as they reach the recording electrode.
JITTER MEASUREMENT WITH VOLUNTARY ACTIVATION Jitter measurements are most technically straightforward with the voluntary activation method in cooperative patients. With this technique, paired jitter measurements are made between a triggering potential and a nontriggering potential to reflect combined jitter in two end plates. Initially, the recording needle is inserted in a muscle and positioned to record clear, sharp signals from two or more muscle fibers within the same motor unit. This positioning is different from that used for FD assessment where the amplitude of a single MFAP is maximized. The subject minimally and steadily contracts the muscle to maintain a constant firing rate. The oscilloscope is triggered on one of the potentials, and that triggering potential is displayed in a fixed position on the oscilloscope screen. Nontriggering potentials from the other muscle fibers are relatively time-locked to the triggering potential in the normal state, but become increasingly variable with neuromuscular junction disease. Typically, 100 discharges of the triggering potential are collected. After review of the tracings and editing of spurious triggers and other artifacts, at least 50 signals should be suitable for jitter analysis. The electrode is repositioned to record from 20 different fiber pairs in different areas of the muscle. About three needle insertions are required to collect this sample. For assessment of possible MG, the extensor digitorum communis (EDC), frontalis, and orbicularis oculi are the most commonly studied muscles.
JITTER MEASUREMENT WITH AXONAL STIMULATION Jitter measurements with axonal stimulation are technically demanding and are most useful for studying
305
patients unable to maintain constant muscle activation, such as patients with impaired consciousness, tremor, or very young children. Recordings with this technique also facilitate the assessment of diseases with ratedependent jitter such as Lambert–Eaton myasthenia (LEM). In this technique, axonal stimulation of an intramuscular motor nerve branch or of a motor nerve proximal to a muscle is used to trigger the oscilloscope, and jitter is measured between the stimulation pulse and one or more MFAPs from the same motor unit. Stimulation jitter recordings are typically performed in the EDC and orbicularis oculi muscles. For EDC recordings, a monopolar stimulating electrode is placed in the proximal end-plate region with a surface or second monopolar needle anode placed about 2 cm laterally (Trontelj et al., 1986). For orbicularis oculi recordings, the monopolar stimulating electrode is placed near one of the temporal branches of the facial nerve supplying the lateral quadrant of the orbicularis oculi muscle after establishing the location of the nerve branch with surface stimulation. A surface cathode is placed nearby over the zygomatic arch (Trontelj et al., 1988). Stimuli of very low intensity (<10 mA) and short duration are then applied at 2 Hz to elicit very slight muscle twitching. A single fiber electrode is then inserted in the twitching portion of the muscle and optimally positioned to record single fiber potentials. The stimulus frequency is then increased to a physiologic firing rate of 10 Hz, and the stimulus intensity is increased by about 15% to a supramaximal level. Achieving supramaximal axonal stimulation is a critical issue in this technique, as subthreshold or liminal stimulation can elicit false jitter and impulse blocking (Trontelj and Stålberg, 1992a). Excessive stimulation and twitching in the EDC during recordings can dislodge the stimulating needle and result in liminal stimulation. Jitter values less than 5 ms likely reflect direct muscle fiber stimulation. As with the voluntary activation method, at least 50 signals are collected for each MFAP. The single fiber needle is repositioned several times during the study to ensure that recordings are made from different muscle fibers. A minimum of 30 end plates should be studied to demonstrate normal findings in a muscle.
CRITERIA FOR ACCEPTABLE POTENTIALS During SFEMG procedures, MFAP signals should be collected with a noise-free baseline and a stable electrode position, as movement of 100 mm may change the MFAP amplitude by >20% (Stålberg et al., 2018). Positional stability can be assessed in superimposition mode to verify a constant MFAP shape (Fig. 19.3). In paired jitter studies, the trigger should remain stable on the very same
306
V.C. JUEL
5 ms 100 mV
Fig. 19.3. Normal paired jitter (MCD ¼ 31 ms) recorded in the extensor digitorum communis muscle. Copyright © 2011, VC Juel.
MFAP throughout data collection. Impulse blocking should be noted and confirmed by inspecting rastered tracings. Blocked pairs are observed when jitter values exceed about 100 ms, and blocking noted at lower jitter values suggests a technical issue. Paired jitter recordings should be made between different potential pairs verified by visual inspection of the signal waveforms and by differences in firing rate and in interpotential interval (IPI). Artifacts to recognize in SFEMG recordings include triangular injury potentials observed when an MFAP is followed by an unstable waveform resembling a positive sharp wave. Injury potentials arise from the same muscle fiber as the initial MFAP and should never be included in jitter analysis (Stålberg et al., 2010). Bimodal jitter or “flip-flop” occurs when two distinct IPI values are observed due to intermittent delayed firing of the triggering potential (Thiele and Stålberg, 1974). Abnormally low jitter (<5 ms) suggests a signal without neuromuscular transmission and is observed with muscle fiber splitting in chronic myopathy; these values should also not be included in jitter calculations.
JITTER CALCULATION Jitter is calculated as the mean difference between consecutive IPIs, or MCD, as follows: MCD ¼ |IPI1 IPI2 | + |IPI2 IPI3 | + ⋯ + |IPIn1 IPIn |=ðn 1Þ: where n ¼ the total number of IPIs. In axonal stimulation studies, thepffiffiMCD value is transformed by dividing the ffi MCD by 2 to account for the measurement of jitter at a single end plate (Trontelj et al., 1986). The MCD calculation helps to reduce the influence of slow trends in the IPIs unrelated to neuromuscular transmission over the course of data collection. Causes of such slow trends include changes in MFAP
configuration due to slight movement of the single fiber electrode or changes in muscle fiber propagation velocity. During paired jitter measurements with voluntary activation, variability in the MFAP firing rate or interdischarge interval (IDI) may influence the IPI due to changes in MFAP propagation. The effect of previous MFAPs on the speed of subsequent MFAPs that spread along the muscle fiber is known as the velocity recovery function (VRF) (Stålberg, 1966). The VRF is related to restoration of the membrane resting potential following muscle fiber depolarization and to fiber diameter. With variable firing rates in jitter measurements with voluntary activation or when the stimulation rate is changed in jitter measurements with axonal stimulation, the influence of the VRF can increase IPI variability and lead to an apparent increase in MCD. Potential pairs with long IPI values (>4 ms) should not be included in jitter analysis due to the VRF and influence of the previous IDI. Use of the mean sorted data difference (MSD) helps to reduce the effect of variable firing rate on the jitter calculation. MSD ¼ |IPIa IPIb | + |IPIb IPIc | + ⋯ + |IPIy IPIz |=ðn 1Þ: where IDIa < IDIb < IDIc < ⋯ < IDIz. When reporting results for paired jitter, the lower of the calculated MCD or MSD values should be used to reduce the contribution of nonjunctional influences that may artifactually increase jitter. Stimulated jitter should not be calculated using data collected at transition periods between different stimulation frequencies.
JITTER ANALYSIS In MG, normal jitter, abnormal jitter without blocking, and abnormal jitter with blocking may be observed in
SINGLE FIBER ELECTROMYOGRAPHY
307
5 ms 500 mV
Fig. 19.4. Increased paired jitter (MCD ¼ 118 ms) with impulse blocking in the extensor digitorum communis muscle in a patient with moderate generalized myasthenia gravis (MGFA 3A). Copyright © 2011, VC Juel.
the same muscle (Fig. 19.4). Several reporting parameters help to accurately characterize the population of end plates studied in a muscle with SFEMG. The mean jitter value for all potentials or potential pairs is a reasonable measure of central tendency and reporting the median jitter value mitigates the influence of outliers with very high jitter. The percentage of normal fiber pairs, abnormal fiber pairs without blocking, and abnormal fiber pairs with blocking should also be reported. Sequential histograms are helpful to demonstrate trends in data, such as the effect of firing rate (Stålberg et al., 2018). A jitter study is abnormal in a muscle when >10% of individual fiber pairs (or end plates with axonal stimulation) are abnormal or when the mean jitter exceeds normal limits (American Association of Electrodiagnostic Medicine Quality Assurance Committee, 2001a). Jitter normally ranges from 10–50 ms depending on the muscle and age of the subject. Jitter is increased with advanced age and in distal muscles. A multicenter collaboration developed normative values for paired jitter with voluntary activation and FD for several muscles and age ranges (Gilchrist et al., 1992; Bromberg and Scott, 1994). Normative values for end plates studied with axonal stimulation have also been published (Stålberg et al., 2010). In comparing the techniques of jitter measurement with voluntary activation and axonal stimulation, the voluntary activation method is technically easier to perform in cooperative patients with fewer artifacts and more straightforward assessment of impulse blocking. However, variable firing rates may influence the MCD and false triggers may confound data collection. Although the axonal stimulation technique is very challenging, it can facilitate jitter assessment in patients unable to provide sustained muscle activation and allows assessment of the effect of firing rate. However, a number of common technical artifacts can seriously corrupt the measurements, including liminal stimulation with false jitter and blocking and direct muscle fiber stimulation with abnormally low jitter values.
Single fiber electrodes are typically sterilized and reused due to their high cost and must be periodically cleaned and sharpened to maintain proper signalrecording characteristics. Electrode maintenance procedures are detailed elsewhere (Stålberg et al., 2010).
JITTER MEASUREMENT WITH CONCENTRIC NEEDLE ELECTRODES In light of contemporary restrictions on use of resterilized material by medical institutions, disposable concentric needles are becoming more widely used to measure jitter. The primary issue related to this technique is that there is significant loss of selectivity when a concentric needle with a much larger recording surface is used. Conventional concentric needles have a recording surface about 140 times larger than the single fiber electrode, and the smallest commercially available “facial” concentric needles have recording surfaces about 38 times larger than single fiber electrodes (Stålberg and Sanders, 2009). The signals recorded with concentric needle electrodes or “apparent single fiber action potentials” are often composite potentials that represent a summation of several individual MFAPs (Ertas et al., 2000; Stålberg and Sanders, 2009). As such, FD measurements cannot be made using concentric needle electrodes. The jitter recorded from a compound signal is smaller than that of its component signals, and jitter measured with concentric needles is about 5 ms lower than when measured with a true single fiber electrode (Stålberg et al., 2010). An international consortium recently established reference values for jitter recorded by concentric needle electrodes in healthy controls. Normal values for paired jitter recordings with voluntary activation as well as for jitter recordings with axonal stimulation were established in the orbicularis oculi, frontalis, and EDC muscles (Stålberg et al., 2016).
V.C. JUEL
When performing jitter studies with concentric needle electrodes, the electrode with the smallest available recording surface should be used (currently the Alpine DCF-25 concentric needle electrode with a cannula diameter of 0.30 mm and recording surface of 0.019 mm2) (Stålberg and Sanders, 2009). This electrode was used by the international consortium to generate the reference values noted here, and it should therefore be used by operators intending to reference patient study data to these normative values. Increasing low-frequency filtering to 1000Hz helps to reduce the influence of muscle fiber potentials distant from the recording surface, but this filtering attenuates the amplitudes of the recorded potentials. Further increases in low-frequency filtering may generate “ringing” or add extra phases to recorded potentials. Jitter measurements in these extra phases will yield abnormally low jitter values (<5 ms) (Stålberg and Sanders, 2009). Acceptable potentials for jitter analysis recorded with concentric needles should have one positive and one negative peak, and the initial rising phase should exhibit a constant shape with no shoulders or notches. Signal peaks should be well defined with constant shape and no gross amplitude variation as verified in superimposition mode. Due to the effect of filtering on amplitude, potentials with signal amplitudes of >50mV are acceptable in concentric jitter recordings. There should be a stable baseline preceding and following the potentials to avoid contamination by other activated motor units (Stålberg et al., 2016). As opposed to jitter measurements using a true single fiber electrode, concentric jitter measurements require additional sampling to capture acceptable potentials and post hoc editing and exclusion of composite potentials with shoulders or variable morphology. Using disposable concentric needles for jitter measurement eliminates the issue of electrode maintenance and sterilization, and the operator can be assured that the needles will be sharp and provide good signals. However, the concentric needle sacrifices significant selectivity and some sensitivity in assessment of jitter, and it cannot be used to assess FD. Acquisition and editing times can be relatively longer in order to reject potentials unsuitable for jitter analysis.
FINDINGS IN DISEASE SFEMG is the most sensitive in vivo test for neuromuscular junction disease. When weakness in a given muscle is due to neuromuscular junction disease, abnormal jitter and blocking will be observed in that muscle. When feasible, jitter studies should be performed in affected muscles. Although SFEMG could be performed in almost any skeletal muscle, the muscles most readily studied include the EDC, frontalis, and orbicularis oculi.
Cooperative patients are easily able to maintain low degrees of steady activation in these muscles, thus enabling paired jitter assessments. In addition, normative data are available for all three of these muscles for jitter studies performed with voluntary activation (Gilchrist et al., 1992; Bromberg and Scott, 1994), axonal stimulation (Stålberg et al., 2010), and concentric needle recording (Stålberg et al., 2016).
MYASTHENIA GRAVIS In MG, abnormal jitter is consistently observed in weak muscles and frequently in clinically unaffected muscles with normal strength, and the number of abnormal end plates is proportional to the severity of disease (Sanders et al., 1979). A large series documented abnormal jitter in EDC in 86% of patients with MG; when jitter was assessed in a second muscle (usually frontalis), the diagnostic sensitivity increased to 99% (Sanders and Howard, 1986). Several studies and a retrospective review have documented the increased sensitivity of SFEMG over repetitive nerve stimulation (RNS) studies and acetylcholine receptor (AChR) antibody testing for the diagnosis of MG (Sanders et al., 1979; Sanders and Howard, 1986; Oh et al., 1992; American Association of Electrodiagnostic Medicine Quality Assurance Committee, 2001b). A summary of 844 MG patients assessed prior to treatment documented the high relative sensitivity of SFEMG testing in both generalized and ocular MG compared to RNS testing performed in a hand and shoulder muscle and to AChR antibody assay, particularly in ocular MG (Fig. 19.5) (Stålberg et al., 2010). 99
97
89 76
80 60
% Abnormal
308
48
SF SF ANY EDC
RNS
AChR Ab
Generalized
SF SF ANY EDC
RNS
55
AChR Ab
Ocular
Fig. 19.5. Sensitivity of diagnostic tests in 844 patients with generalized vs ocular MG prior to treatment. SFANY¼ increased jitter in any tested muscle; SFEDC ¼ increased jitter in extensor digitorum communis; RNS¼ abnormal decrement in a hand and/or shoulder muscle; ARA ¼ elevated AChR antibody titer (Sanders DB, Massey JM, Juel VC, Hobson-Webb L, unpublished). From Sta˚lberg E, Trontelj JV, Sanders DB (2010). Single fiber electromyography: studies in healthy and diseased muscle, third edn. Fiskeb€ackskil, Sweden: Edshagen Publishing House, with permission from Edshagen Publishing House.
SINGLE FIBER ELECTROMYOGRAPHY Jitter varies in parallel with the severity of disease in MG, as demonstrated by serial jitter studies (Howard and Sanders, 1981; Sanders and Howard, 1986; Stålberg et al., 2010). Muscle selection for SFEMG testing in MG should be based on the clinical distribution of weakness (Sanders, 2002). When limb or bulbar weakness is present, the EDC should be examined first. If the initial findings are normal and clinical suspicion for MG remains high, the frontalis muscle may also be examined. Patients with pure ocular weakness may undergo initial SFEMG testing in the frontalis muscle; if findings are normal and clinical suspicion for MG is strong, the orbicularis oculi may then be examined.
LAMBERT–EATON MYASTHENIA In LEM, high degrees of jitter and impulse blocking are observed even in the setting of relatively mild clinical weakness (Fig. 19.6). The jitter and blocking may be rate dependent and decrease with higher firing or stimulation rates, though this is not observed in all end plates or in all patients (Trontelj and Stålberg, 1991; Sanders, 1992; Trontelj and Stålberg, 1992b). Improvement in jitter and blocking in LEM parallels clinical response to treatment (Phillips, 1982; Kim et al., 1998).
BOTULISM Jitter studies are more sensitive than RNS testing in botulism. Markedly increased jitter with blocking is
309
observed in weak muscles, and rate-dependent jitter that decreases with increased firing rate may also be observed (Schiller and Stålberg, 1978). In an outbreak of foodborne botulism, SFEMG with voluntary activation demonstrated markedly increased jitter with blocking in all reported cases despite normal RNS findings (Padua et al., 1999). However, in a study of several cases of wound botulism utilizing axonal stimulation, an increased stimulation rate did not frequently or consistently reduce jitter and blocking (Mandler and Maselli, 1996).
MOTOR NEURON DISEASE AND PERIPHERAL NEUROPATHY
In parallel with collateral sprouting and reinnervation, FD is increased in neuropathy and motor neuron disorders. Increased FD is an early and sensitive measure of reinnervation. In a cohort of patients with motor neuron disease, increased jitter and FD were observed in 70% of muscles that had no abnormality by conventional EMG (Massey et al., 1985). Increased FD may be seen within 3–4 weeks following nerve injury and precede reinnervation findings on conventional EMG and muscle histopathology (Schwartz et al., 1976). As reinnervation of the motor unit becomes more complete, increased jitter values return toward normal though FD remains increased. In progressive neuropathy and motor neuron disease, both jitter and FD remain increased.
MYOPATHY Increased FD has been observed in some muscle disorders, though not to the degree seen with reinnervation in neuropathy or motor neuron disease. Fiber splitting, grouping, and loss, and ephaptic recruitment of muscle fibers by other motor units are processes that are likely responsible for the increased FD (Stålberg and Trontelj, 1992; Sanders and Stålberg, 1996). Jitter may also be increased in some myopathies, presumably on the basis of end-plate region pathology and/or immature innervation of regenerating muscle fibers. Though clinically distinct from MG, chronic progressive external ophthalmoplegia may exhibit increased jitter and blocking (Krendel et al., 1987).
SUMMARY Fig. 19.6. Markedly increased jitter (MCD ¼ 1260 ms) with impulse blocking on jitter recordings with axonal stimulation recorded in the orbicularis oculi muscle in a patient with Lambert–Eaton myasthenia. From Sta˚lberg E, Trontelj JV, Sanders DB (2010). Single fiber electromyography: studies in healthy and diseased muscle, third edn. Fiskeb€ackskil, Sweden: Edshagen Publishing House, with permission from Edshagen Publishing House.
SFEMG is a powerful technique that provides significant insight regarding the structure and function of the motor unit through its selective ability to record individual MFAPs. The assessment of jitter provides a highly sensitive assessment of neuromuscular transmission and is particularly helpful when less sensitive diagnostic tests such as RNS testing and AChR antibody testing are negative in the evaluation for neuromuscular junction
310
V.C. JUEL
disease, such as MG. FD is a reflection of the local concentration of muscle fibers within a motor unit and represents a sensitive and early measure of reinnervation. SFEMG testing requires specialized equipment and technical expertise that are not widely available. Intensive quality monitoring is necessary to ensure collection of acceptable signals to prevent misdiagnosis.
REFERENCES American Association of Electrodiagnostic Medicine Quality Assurance Committee (2001a). Practice parameter for repetitive nerve stimulation and single fiber EMG evaluation of adults with suspected myasthenia gravis or Lambert–Eaton myasthenic syndrome: summary statement. Muscle Nerve 24: 1236–1238. American Association of Electrodiagnostic Medicine Quality Assurance Committee (2001b). Literature review of the usefulness of repetitive nerve stimulation and single fiber EMG in the electrodiagnostic evaluation of patients with suspected myasthenia gravis or Lambert–Eaton myasthenic syndrome. Muscle Nerve 24: 1239–1247. Bromberg MB, Scott DM (1994). Single fiber EMG reference values: reformatted in tabular form. Muscle Nerve 17: 820–821. Ertas M, Baslo MB, Yildiz N et al. (2000). Concentric needle electrode for neuromuscular jitter analysis. Muscle Nerve 23: 715–719. Gilchrist JM, Barkhaus PE, Bril V et al. (1992). Single fiber EMG reference values: a collaborative effort. Muscle Nerve 15: 151–161. Howard JF, Sanders DB (1981). Serial single-fiber EMG studies in myasthenic patients treated with corticosteroids and plasma exchange therapy. Muscle Nerve 4: 254. Kim DS, Claussen GC, Oh SJ (1998). Single-fiber electromyography improvement with 3,4-diaminopyridine in Lambert–Eaton myasthenic syndrome. Muscle Nerve 21: 1107–1108. Krendel DA, Sanders DB, Massey JM (1987). Single fiber electromyography in chronic progressive external ophthalmoplegia. Muscle Nerve 10: 299–302. Mandler RN, Maselli RA (1996). Stimulated single-fiber electromyography in wound botulism. Muscle Nerve 19: 1171–1173. Massey JM, Sanders DB, Nandedkar SD (1985). Sensitivity of various EMG techniques in motor neurone disease. Electroencephalogr Clin Neurophysiol 61: 574–575. Oh SJ, Kim DE, Kuruoglu R et al. (1992). Diagnostic sensitivity of the laboratory tests in myasthenia gravis. Muscle Nerve 15: 720–724. Padua L, Aprile I, Lo Monaco M et al. (1999). Neurophysiological assessment in the diagnosis of botulism: usefulness of single-fiber EMG. Muscle Nerve 22: 1388–1392.
Phillips LH (1982). The use of single fiber EMG to monitor the response of Lambert–Eaton myasthenic syndrome to steroid therapy. Muscle Nerve 5: 555. Sanders DB (1992). The effect of firing rate on neuromuscular jitter in Lambert–Eaton myasthenic syndrome. Muscle Nerve 15: 256–258. Sanders DB (2002). Clinical impact of single-fiber electromyography. Muscle Nerve (Suppl 11): S15–S20. Sanders DB, Howard JF (1986). Single-fiber electromyography in myasthenia gravis. Muscle Nerve 9: 809–819. Sanders DB, Howard JF, Johns TR (1979). Single-fiber electromyography in myasthenia gravis. Neurology 29: 68–76. Sanders DB, Sta˚lberg EV (1996). Single-fiber electromyography. Muscle Nerve 19: 1069–1083. Schiller HH, Sta˚lberg E (1978). Human botulism studied with single fibre electromyography. Arch Neurol 35: 346–349. Schwartz MS, Sta˚lberg E, Schiller HH et al. (1976). The reinnervated motor unit in man. A single fibre EMG multielectrode investigation. J Neurol Sci 27: 303–312. Sta˚lberg E (1966). Propagation velocity in human single muscle fibres in situ. Thesis Acta Physiol Scand 70 (Suppl. 287): 1–112. Sta˚lberg EV (1983). Macro EMG. Muscle Nerve 6: 619–630. Sta˚lberg EV, Sanders DB (2009). Jitter recordings with concentric needle electrodes. Muscle Nerve 40: 331–339. Sta˚lberg E, Sanders DB, Ali S et al. (2016). Reference values for jitter recorded by concentric needle electrodes in healthy controls: a multicenter study. Muscle Nerve 53: 351–362. Sta˚lberg E, Trontelj JV (1992). Clinical neurophysiology: the motor unit in myopathy. In: LP Rowland, S DiMauro (Eds.), Handbook of clinical neurology, myopathies. Elsevier, Amsterdam, pp. 49–84. Sta˚lberg E, Trontelj JV, Sanders DB (2010). Single fiber electromyography: studies in healthy and diseased muscle, third edn. Edshagen Publishing House, Fiskeb€ackskil, Sweden. Sta˚lberg E, Trontelj JV, Sanders DB (2018). www.sfemg.info. Accessed March 15, 2018. Thiele B, Sta˚lberg E (1974). The bimodal jitter: a single fibre electromyographic finding. J Neurol Neurosurg Psychiatry 37: 403–411. Trontelj JV, Khuraibet A, Mihelin M (1988). The jitter in stimulated orbicularis oculi muscle: technique and normal values. J Neurol Neurosurg Psychiatry 51: 814–819. Trontelj JV, Mihelin M, Fernandez JM et al. (1986). Axonal stimulation for endplate jitter studies. J Neurol Neurosurg Psychiatry 49: 67–685. Trontelj JV, Sta˚lberg E (1991). Single motor end-plates in myasthenia gravis and LEMS at different firing rates. Muscle Nerve 14: 226–232. Trontelj JV, Sta˚lberg E (1992a). Jitter measurement by axonal micro-stimulation. Guidelines and technical notes. Electroencephalogr Clin Neurophysiol 85: 30–37. Trontelj JV, Sta˚lberg E (1992b). The effect of firing rate on neuromuscular jitter in Lambert–Eaton myasthenic syndrome: a reply. Muscle Nerve 15: 258.
Chapter 19
Single fiber electromyography VERN C. JUEL* Department of Neurology, Duke University School of Medicine, Durham, NC, United States
Abstract Single fiber electromyography (SFEMG) is a highly selective technique that permits assessment of individual muscle fiber action potentials (MFAPs). This selectivity is achieved with a specialized concentric needle electrode with a 25-mm diameter recording surface located in a side port 3 mm from the needle tip. Additional selectivity is achieved with 500-Hz low-frequency filtering. An oscilloscope with a trigger and delay line enables identification of time-locked MFAPs within the same motor unit. SFEMG techniques allow assessment of two important features of the motor unit: jitter and fiber density (FD). Neuromuscular jitter is a direct measure of neuromuscular transmission and reflects the temporal variation in end-plate potentials reaching threshold to elicit a MFAP. SFEMG may be used to assess paired jitter with voluntary activation or by axonal stimulation of motor nerve branches to individual end plates. SFEMG is the most sensitive clinical test for neuromuscular junction disease and is often abnormal in clinically unaffected muscles in patients with myasthenia gravis (MG) and Lambert–Eaton myasthenia (LEM). Normal jitter findings in a clinically weak muscle exclude neuromuscular junction disease as a cause for weakness in that muscle. FD measurements assess the local concentration of muscle fibers within a motor unit and provide a sensitive in vivo assessment of reinnervation.
INTRODUCTION
THE SINGLE FIBER ELECTRODE
Single fiber electromyography (SFEMG) is a highly selective technique that facilitates assessment of individual muscle fiber action potentials (MFAPs). This selectivity is achieved by use of a specialized concentric electrode with a very small recording surface in conjunction with recording equipment that filters low-frequency signals from distant muscle fibers. With SFEMG, MFAPs innervated by the same motor neuron can be recorded to provide valuable information regarding the structure and function of motor units. SFEMG is used to assess neuromuscular transmission through measurement of jitter and impulse blocking. Fiber density (FD) measurement by SFEMG may reveal early evidence for reinnervation in motor neuropathic processes.
The selectivity of the single fiber electrode is largely due to its small 25-mm diameter recording surface located in a side port 3 mm from the needle tip (Fig. 19.1). The small recording surface area (0.0005 mm2) of the single fiber electrode results in less shunting of electrical field from adjacent muscle fibers compared with the larger recording surfaces of conventional concentric needle electrodes (0.07 mm2) and of facial concentric needle electrodes (0.019 mm2) (Stålberg and Sanders, 2009). This selectivity is reflected in the marked radial decline in signal amplitude seen with increasing distance from the recording surface with single fiber electrodes, in comparison to conventional concentric, monopolar, and macro EMG electrodes (Fig. 19.2) (Stålberg, 1983).
*Correspondence to: Vern C. Juel, M.D., 40 Duke Medicine Circle, Clinic 1L, Room 1255, DUMC 3403, Duke University Medical Center, Durham, NC 27710, United States. Tel: +919-684-4044, Fax: +919-660-3853, E-mail: [email protected]
304
V.C. JUEL
Fig. 19.1. SFEMG electrode and needle holder. The selectivity of SFEMG is attributed primarily to the 25-mm diameter recording surface at the side port 3 mm from the needle tip. Copyright © 2011, DB Sanders.
1000
AMPL (mV)
800
600
400 SF el
200 conc el macro el
0 0
400
800
1200
1600
2000
2400
Distance (mm)
Fig. 19.2. Radial decline of signal amplitude over distance with single fiber, concentric, and macro electrodes. Note the sharp decline in the signal amplitude with the single fiber electrode. From Sta˚lberg EV (1983). Macro EMG. Muscle Nerve 6: 619–630, with permission from John Wiley and Sons.
INSTRUMENTATION The selectivity afforded by the single fiber electrode is enhanced by the use of a 500-Hz low-frequency filter to reject signals from distant muscle fibers. A highfrequency filter setting of 10 kHz is also used. Amplifier gain settings of 0.2–0.5 mV/division are optimal for recording MFAPs; higher settings introduce more background noise into the recordings. An oscilloscope sweep speed of 0.5 ms/division facilitates assessment of the temporal variability of the MFAP signals.
A trigger and delay line is used to identify signal components that precede and follow the triggered signal. Voltage- or amplitude-level triggers have traditionally been used in SFEMG recording and initiate the oscilloscope sweep when a potential exceeds the voltage level set by the operator. With the peak detection triggering technique, the operator also sets a voltage level, and the peaks of the triggering and nontriggering potentials are extrapolated from the intersection of computercalculated slopes of the rising and falling phases of the MFAPs. While the techniques produce comparable jitter values, the peak detection method is less affected by artifactually increased jitter values related to overlapping or “riding” signals or to baseline variability (Stålberg and Sanders, 2009).
FIBER DENSITY MEASUREMENTS FD measurements assess the local concentration of muscle fibers within a motor unit. Increased FD is analogous to histopathologic muscle fiber type grouping seen following reinnervation from collateral sprouting of motor nerves. When measured with a single fiber electrode, the amplitude of an MFAP falls to 200 mV when the electrode is 300 mm from the muscle fiber. Accordingly, MFAPs with amplitudes >200 mV and rise times <300 ms are generated by muscle fibers within 300 mm of the recording surface (Stålberg et al., 2010). FD assessment must be performed with actual single fiber electrodes, as the recording characteristics of conventional concentric needles do not reliably permit signal recordings from individual muscle fibers. To assess FD, the single fiber electrode is positioned to maximize the amplitude of a single MFAP. After triggering on this potential, the mean number of timelocked potentials (with amplitude >200 mV and rise time <300 ms) observed in this recording position is counted. In addition to the triggering potential, all clear notches involving the potentials are included in the total count of MFAPs. A total of 20 different sites within the muscle are sampled via three to four separate needle insertions. The mean FD is then calculated for the muscle and compared to muscle and age-matched normal values (Gilchrist et al., 1992; Bromberg and Scott, 1994).
JITTER In the 1960s, Stålberg and Ekstedt originally observed variability in signal latencies recorded with novel, custom electrodes designed to record potentials from individual muscle fibers. Although this variability was initially attributed to a defective oscilloscope (hence the term jitter, referring to signal deviations from true periodicity in an electronic device), it became clear that this variability was related to neuromuscular
SINGLE FIBER ELECTROMYOGRAPHY transmission following recordings from a patient with myasthenia gravis and subsequent experiments with curare (Stålberg et al., 2010). Neuromuscular jitter is a direct measure of the effectiveness of neuromuscular transmission and represents the temporal variation for muscle end-plate potentials (EPPs) to reach threshold to produce an MFAP. Neuromuscular junction disease lowers the safety factor by reduced quantal release of acetylcholine (ACh) in presynaptic disease or by reduced end-plate responsiveness to ACh in postsynaptic disorders. In either case, increased jitter occurs when EPPs are delayed in reaching threshold, and impulse blocking occurs when EPPs fail to elicit a MFAP. Clinical weakness in a given muscle occurs when a critical number of end plates are blocked in that muscle. In SFEMG, jitter is measured as the variability in the arrival time of individual MFAPs as they reach the recording electrode.
JITTER MEASUREMENT WITH VOLUNTARY ACTIVATION Jitter measurements are most technically straightforward with the voluntary activation method in cooperative patients. With this technique, paired jitter measurements are made between a triggering potential and a nontriggering potential to reflect combined jitter in two end plates. Initially, the recording needle is inserted in a muscle and positioned to record clear, sharp signals from two or more muscle fibers within the same motor unit. This positioning is different from that used for FD assessment where the amplitude of a single MFAP is maximized. The subject minimally and steadily contracts the muscle to maintain a constant firing rate. The oscilloscope is triggered on one of the potentials, and that triggering potential is displayed in a fixed position on the oscilloscope screen. Nontriggering potentials from the other muscle fibers are relatively time-locked to the triggering potential in the normal state, but become increasingly variable with neuromuscular junction disease. Typically, 100 discharges of the triggering potential are collected. After review of the tracings and editing of spurious triggers and other artifacts, at least 50 signals should be suitable for jitter analysis. The electrode is repositioned to record from 20 different fiber pairs in different areas of the muscle. About three needle insertions are required to collect this sample. For assessment of possible MG, the extensor digitorum communis (EDC), frontalis, and orbicularis oculi are the most commonly studied muscles.
JITTER MEASUREMENT WITH AXONAL STIMULATION Jitter measurements with axonal stimulation are technically demanding and are most useful for studying
305
patients unable to maintain constant muscle activation, such as patients with impaired consciousness, tremor, or very young children. Recordings with this technique also facilitate the assessment of diseases with ratedependent jitter such as Lambert–Eaton myasthenia (LEM). In this technique, axonal stimulation of an intramuscular motor nerve branch or of a motor nerve proximal to a muscle is used to trigger the oscilloscope, and jitter is measured between the stimulation pulse and one or more MFAPs from the same motor unit. Stimulation jitter recordings are typically performed in the EDC and orbicularis oculi muscles. For EDC recordings, a monopolar stimulating electrode is placed in the proximal end-plate region with a surface or second monopolar needle anode placed about 2 cm laterally (Trontelj et al., 1986). For orbicularis oculi recordings, the monopolar stimulating electrode is placed near one of the temporal branches of the facial nerve supplying the lateral quadrant of the orbicularis oculi muscle after establishing the location of the nerve branch with surface stimulation. A surface cathode is placed nearby over the zygomatic arch (Trontelj et al., 1988). Stimuli of very low intensity (<10 mA) and short duration are then applied at 2 Hz to elicit very slight muscle twitching. A single fiber electrode is then inserted in the twitching portion of the muscle and optimally positioned to record single fiber potentials. The stimulus frequency is then increased to a physiologic firing rate of 10 Hz, and the stimulus intensity is increased by about 15% to a supramaximal level. Achieving supramaximal axonal stimulation is a critical issue in this technique, as subthreshold or liminal stimulation can elicit false jitter and impulse blocking (Trontelj and Stålberg, 1992a). Excessive stimulation and twitching in the EDC during recordings can dislodge the stimulating needle and result in liminal stimulation. Jitter values less than 5 ms likely reflect direct muscle fiber stimulation. As with the voluntary activation method, at least 50 signals are collected for each MFAP. The single fiber needle is repositioned several times during the study to ensure that recordings are made from different muscle fibers. A minimum of 30 end plates should be studied to demonstrate normal findings in a muscle.
CRITERIA FOR ACCEPTABLE POTENTIALS During SFEMG procedures, MFAP signals should be collected with a noise-free baseline and a stable electrode position, as movement of 100 mm may change the MFAP amplitude by >20% (Stålberg et al., 2018). Positional stability can be assessed in superimposition mode to verify a constant MFAP shape (Fig. 19.3). In paired jitter studies, the trigger should remain stable on the very same
306
V.C. JUEL
5 ms 100 mV
Fig. 19.3. Normal paired jitter (MCD ¼ 31 ms) recorded in the extensor digitorum communis muscle. Copyright © 2011, VC Juel.
MFAP throughout data collection. Impulse blocking should be noted and confirmed by inspecting rastered tracings. Blocked pairs are observed when jitter values exceed about 100 ms, and blocking noted at lower jitter values suggests a technical issue. Paired jitter recordings should be made between different potential pairs verified by visual inspection of the signal waveforms and by differences in firing rate and in interpotential interval (IPI). Artifacts to recognize in SFEMG recordings include triangular injury potentials observed when an MFAP is followed by an unstable waveform resembling a positive sharp wave. Injury potentials arise from the same muscle fiber as the initial MFAP and should never be included in jitter analysis (Stålberg et al., 2010). Bimodal jitter or “flip-flop” occurs when two distinct IPI values are observed due to intermittent delayed firing of the triggering potential (Thiele and Stålberg, 1974). Abnormally low jitter (<5 ms) suggests a signal without neuromuscular transmission and is observed with muscle fiber splitting in chronic myopathy; these values should also not be included in jitter calculations.
JITTER CALCULATION Jitter is calculated as the mean difference between consecutive IPIs, or MCD, as follows: MCD ¼ |IPI1 IPI2 | + |IPI2 IPI3 | + ⋯ + |IPIn1 IPIn |=ðn 1Þ: where n ¼ the total number of IPIs. In axonal stimulation studies, thepffiffiMCD value is transformed by dividing the ffi MCD by 2 to account for the measurement of jitter at a single end plate (Trontelj et al., 1986). The MCD calculation helps to reduce the influence of slow trends in the IPIs unrelated to neuromuscular transmission over the course of data collection. Causes of such slow trends include changes in MFAP
configuration due to slight movement of the single fiber electrode or changes in muscle fiber propagation velocity. During paired jitter measurements with voluntary activation, variability in the MFAP firing rate or interdischarge interval (IDI) may influence the IPI due to changes in MFAP propagation. The effect of previous MFAPs on the speed of subsequent MFAPs that spread along the muscle fiber is known as the velocity recovery function (VRF) (Stålberg, 1966). The VRF is related to restoration of the membrane resting potential following muscle fiber depolarization and to fiber diameter. With variable firing rates in jitter measurements with voluntary activation or when the stimulation rate is changed in jitter measurements with axonal stimulation, the influence of the VRF can increase IPI variability and lead to an apparent increase in MCD. Potential pairs with long IPI values (>4 ms) should not be included in jitter analysis due to the VRF and influence of the previous IDI. Use of the mean sorted data difference (MSD) helps to reduce the effect of variable firing rate on the jitter calculation. MSD ¼ |IPIa IPIb | + |IPIb IPIc | + ⋯ + |IPIy IPIz |=ðn 1Þ: where IDIa < IDIb < IDIc < ⋯ < IDIz. When reporting results for paired jitter, the lower of the calculated MCD or MSD values should be used to reduce the contribution of nonjunctional influences that may artifactually increase jitter. Stimulated jitter should not be calculated using data collected at transition periods between different stimulation frequencies.
JITTER ANALYSIS In MG, normal jitter, abnormal jitter without blocking, and abnormal jitter with blocking may be observed in
SINGLE FIBER ELECTROMYOGRAPHY
307
5 ms 500 mV
Fig. 19.4. Increased paired jitter (MCD ¼ 118 ms) with impulse blocking in the extensor digitorum communis muscle in a patient with moderate generalized myasthenia gravis (MGFA 3A). Copyright © 2011, VC Juel.
the same muscle (Fig. 19.4). Several reporting parameters help to accurately characterize the population of end plates studied in a muscle with SFEMG. The mean jitter value for all potentials or potential pairs is a reasonable measure of central tendency and reporting the median jitter value mitigates the influence of outliers with very high jitter. The percentage of normal fiber pairs, abnormal fiber pairs without blocking, and abnormal fiber pairs with blocking should also be reported. Sequential histograms are helpful to demonstrate trends in data, such as the effect of firing rate (Stålberg et al., 2018). A jitter study is abnormal in a muscle when >10% of individual fiber pairs (or end plates with axonal stimulation) are abnormal or when the mean jitter exceeds normal limits (American Association of Electrodiagnostic Medicine Quality Assurance Committee, 2001a). Jitter normally ranges from 10–50 ms depending on the muscle and age of the subject. Jitter is increased with advanced age and in distal muscles. A multicenter collaboration developed normative values for paired jitter with voluntary activation and FD for several muscles and age ranges (Gilchrist et al., 1992; Bromberg and Scott, 1994). Normative values for end plates studied with axonal stimulation have also been published (Stålberg et al., 2010). In comparing the techniques of jitter measurement with voluntary activation and axonal stimulation, the voluntary activation method is technically easier to perform in cooperative patients with fewer artifacts and more straightforward assessment of impulse blocking. However, variable firing rates may influence the MCD and false triggers may confound data collection. Although the axonal stimulation technique is very challenging, it can facilitate jitter assessment in patients unable to provide sustained muscle activation and allows assessment of the effect of firing rate. However, a number of common technical artifacts can seriously corrupt the measurements, including liminal stimulation with false jitter and blocking and direct muscle fiber stimulation with abnormally low jitter values.
Single fiber electrodes are typically sterilized and reused due to their high cost and must be periodically cleaned and sharpened to maintain proper signalrecording characteristics. Electrode maintenance procedures are detailed elsewhere (Stålberg et al., 2010).
JITTER MEASUREMENT WITH CONCENTRIC NEEDLE ELECTRODES In light of contemporary restrictions on use of resterilized material by medical institutions, disposable concentric needles are becoming more widely used to measure jitter. The primary issue related to this technique is that there is significant loss of selectivity when a concentric needle with a much larger recording surface is used. Conventional concentric needles have a recording surface about 140 times larger than the single fiber electrode, and the smallest commercially available “facial” concentric needles have recording surfaces about 38 times larger than single fiber electrodes (Stålberg and Sanders, 2009). The signals recorded with concentric needle electrodes or “apparent single fiber action potentials” are often composite potentials that represent a summation of several individual MFAPs (Ertas et al., 2000; Stålberg and Sanders, 2009). As such, FD measurements cannot be made using concentric needle electrodes. The jitter recorded from a compound signal is smaller than that of its component signals, and jitter measured with concentric needles is about 5 ms lower than when measured with a true single fiber electrode (Stålberg et al., 2010). An international consortium recently established reference values for jitter recorded by concentric needle electrodes in healthy controls. Normal values for paired jitter recordings with voluntary activation as well as for jitter recordings with axonal stimulation were established in the orbicularis oculi, frontalis, and EDC muscles (Stålberg et al., 2016).
V.C. JUEL
When performing jitter studies with concentric needle electrodes, the electrode with the smallest available recording surface should be used (currently the Alpine DCF-25 concentric needle electrode with a cannula diameter of 0.30 mm and recording surface of 0.019 mm2) (Stålberg and Sanders, 2009). This electrode was used by the international consortium to generate the reference values noted here, and it should therefore be used by operators intending to reference patient study data to these normative values. Increasing low-frequency filtering to 1000Hz helps to reduce the influence of muscle fiber potentials distant from the recording surface, but this filtering attenuates the amplitudes of the recorded potentials. Further increases in low-frequency filtering may generate “ringing” or add extra phases to recorded potentials. Jitter measurements in these extra phases will yield abnormally low jitter values (<5 ms) (Stålberg and Sanders, 2009). Acceptable potentials for jitter analysis recorded with concentric needles should have one positive and one negative peak, and the initial rising phase should exhibit a constant shape with no shoulders or notches. Signal peaks should be well defined with constant shape and no gross amplitude variation as verified in superimposition mode. Due to the effect of filtering on amplitude, potentials with signal amplitudes of >50mV are acceptable in concentric jitter recordings. There should be a stable baseline preceding and following the potentials to avoid contamination by other activated motor units (Stålberg et al., 2016). As opposed to jitter measurements using a true single fiber electrode, concentric jitter measurements require additional sampling to capture acceptable potentials and post hoc editing and exclusion of composite potentials with shoulders or variable morphology. Using disposable concentric needles for jitter measurement eliminates the issue of electrode maintenance and sterilization, and the operator can be assured that the needles will be sharp and provide good signals. However, the concentric needle sacrifices significant selectivity and some sensitivity in assessment of jitter, and it cannot be used to assess FD. Acquisition and editing times can be relatively longer in order to reject potentials unsuitable for jitter analysis.
FINDINGS IN DISEASE SFEMG is the most sensitive in vivo test for neuromuscular junction disease. When weakness in a given muscle is due to neuromuscular junction disease, abnormal jitter and blocking will be observed in that muscle. When feasible, jitter studies should be performed in affected muscles. Although SFEMG could be performed in almost any skeletal muscle, the muscles most readily studied include the EDC, frontalis, and orbicularis oculi.
Cooperative patients are easily able to maintain low degrees of steady activation in these muscles, thus enabling paired jitter assessments. In addition, normative data are available for all three of these muscles for jitter studies performed with voluntary activation (Gilchrist et al., 1992; Bromberg and Scott, 1994), axonal stimulation (Stålberg et al., 2010), and concentric needle recording (Stålberg et al., 2016).
MYASTHENIA GRAVIS In MG, abnormal jitter is consistently observed in weak muscles and frequently in clinically unaffected muscles with normal strength, and the number of abnormal end plates is proportional to the severity of disease (Sanders et al., 1979). A large series documented abnormal jitter in EDC in 86% of patients with MG; when jitter was assessed in a second muscle (usually frontalis), the diagnostic sensitivity increased to 99% (Sanders and Howard, 1986). Several studies and a retrospective review have documented the increased sensitivity of SFEMG over repetitive nerve stimulation (RNS) studies and acetylcholine receptor (AChR) antibody testing for the diagnosis of MG (Sanders et al., 1979; Sanders and Howard, 1986; Oh et al., 1992; American Association of Electrodiagnostic Medicine Quality Assurance Committee, 2001b). A summary of 844 MG patients assessed prior to treatment documented the high relative sensitivity of SFEMG testing in both generalized and ocular MG compared to RNS testing performed in a hand and shoulder muscle and to AChR antibody assay, particularly in ocular MG (Fig. 19.5) (Stålberg et al., 2010). 99
97
89 76
80 60
% Abnormal
308
48
SF SF ANY EDC
RNS
AChR Ab
Generalized
SF SF ANY EDC
RNS
55
AChR Ab
Ocular
Fig. 19.5. Sensitivity of diagnostic tests in 844 patients with generalized vs ocular MG prior to treatment. SFANY¼ increased jitter in any tested muscle; SFEDC ¼ increased jitter in extensor digitorum communis; RNS¼ abnormal decrement in a hand and/or shoulder muscle; ARA ¼ elevated AChR antibody titer (Sanders DB, Massey JM, Juel VC, Hobson-Webb L, unpublished). From Sta˚lberg E, Trontelj JV, Sanders DB (2010). Single fiber electromyography: studies in healthy and diseased muscle, third edn. Fiskeb€ackskil, Sweden: Edshagen Publishing House, with permission from Edshagen Publishing House.
SINGLE FIBER ELECTROMYOGRAPHY Jitter varies in parallel with the severity of disease in MG, as demonstrated by serial jitter studies (Howard and Sanders, 1981; Sanders and Howard, 1986; Stålberg et al., 2010). Muscle selection for SFEMG testing in MG should be based on the clinical distribution of weakness (Sanders, 2002). When limb or bulbar weakness is present, the EDC should be examined first. If the initial findings are normal and clinical suspicion for MG remains high, the frontalis muscle may also be examined. Patients with pure ocular weakness may undergo initial SFEMG testing in the frontalis muscle; if findings are normal and clinical suspicion for MG is strong, the orbicularis oculi may then be examined.
LAMBERT–EATON MYASTHENIA In LEM, high degrees of jitter and impulse blocking are observed even in the setting of relatively mild clinical weakness (Fig. 19.6). The jitter and blocking may be rate dependent and decrease with higher firing or stimulation rates, though this is not observed in all end plates or in all patients (Trontelj and Stålberg, 1991; Sanders, 1992; Trontelj and Stålberg, 1992b). Improvement in jitter and blocking in LEM parallels clinical response to treatment (Phillips, 1982; Kim et al., 1998).
BOTULISM Jitter studies are more sensitive than RNS testing in botulism. Markedly increased jitter with blocking is
309
observed in weak muscles, and rate-dependent jitter that decreases with increased firing rate may also be observed (Schiller and Stålberg, 1978). In an outbreak of foodborne botulism, SFEMG with voluntary activation demonstrated markedly increased jitter with blocking in all reported cases despite normal RNS findings (Padua et al., 1999). However, in a study of several cases of wound botulism utilizing axonal stimulation, an increased stimulation rate did not frequently or consistently reduce jitter and blocking (Mandler and Maselli, 1996).
MOTOR NEURON DISEASE AND PERIPHERAL NEUROPATHY
In parallel with collateral sprouting and reinnervation, FD is increased in neuropathy and motor neuron disorders. Increased FD is an early and sensitive measure of reinnervation. In a cohort of patients with motor neuron disease, increased jitter and FD were observed in 70% of muscles that had no abnormality by conventional EMG (Massey et al., 1985). Increased FD may be seen within 3–4 weeks following nerve injury and precede reinnervation findings on conventional EMG and muscle histopathology (Schwartz et al., 1976). As reinnervation of the motor unit becomes more complete, increased jitter values return toward normal though FD remains increased. In progressive neuropathy and motor neuron disease, both jitter and FD remain increased.
MYOPATHY Increased FD has been observed in some muscle disorders, though not to the degree seen with reinnervation in neuropathy or motor neuron disease. Fiber splitting, grouping, and loss, and ephaptic recruitment of muscle fibers by other motor units are processes that are likely responsible for the increased FD (Stålberg and Trontelj, 1992; Sanders and Stålberg, 1996). Jitter may also be increased in some myopathies, presumably on the basis of end-plate region pathology and/or immature innervation of regenerating muscle fibers. Though clinically distinct from MG, chronic progressive external ophthalmoplegia may exhibit increased jitter and blocking (Krendel et al., 1987).
SUMMARY Fig. 19.6. Markedly increased jitter (MCD ¼ 1260 ms) with impulse blocking on jitter recordings with axonal stimulation recorded in the orbicularis oculi muscle in a patient with Lambert–Eaton myasthenia. From Sta˚lberg E, Trontelj JV, Sanders DB (2010). Single fiber electromyography: studies in healthy and diseased muscle, third edn. Fiskeb€ackskil, Sweden: Edshagen Publishing House, with permission from Edshagen Publishing House.
SFEMG is a powerful technique that provides significant insight regarding the structure and function of the motor unit through its selective ability to record individual MFAPs. The assessment of jitter provides a highly sensitive assessment of neuromuscular transmission and is particularly helpful when less sensitive diagnostic tests such as RNS testing and AChR antibody testing are negative in the evaluation for neuromuscular junction
310
V.C. JUEL
disease, such as MG. FD is a reflection of the local concentration of muscle fibers within a motor unit and represents a sensitive and early measure of reinnervation. SFEMG testing requires specialized equipment and technical expertise that are not widely available. Intensive quality monitoring is necessary to ensure collection of acceptable signals to prevent misdiagnosis.
REFERENCES American Association of Electrodiagnostic Medicine Quality Assurance Committee (2001a). Practice parameter for repetitive nerve stimulation and single fiber EMG evaluation of adults with suspected myasthenia gravis or Lambert–Eaton myasthenic syndrome: summary statement. Muscle Nerve 24: 1236–1238. American Association of Electrodiagnostic Medicine Quality Assurance Committee (2001b). Literature review of the usefulness of repetitive nerve stimulation and single fiber EMG in the electrodiagnostic evaluation of patients with suspected myasthenia gravis or Lambert–Eaton myasthenic syndrome. Muscle Nerve 24: 1239–1247. Bromberg MB, Scott DM (1994). Single fiber EMG reference values: reformatted in tabular form. Muscle Nerve 17: 820–821. Ertas M, Baslo MB, Yildiz N et al. (2000). Concentric needle electrode for neuromuscular jitter analysis. Muscle Nerve 23: 715–719. Gilchrist JM, Barkhaus PE, Bril V et al. (1992). Single fiber EMG reference values: a collaborative effort. Muscle Nerve 15: 151–161. Howard JF, Sanders DB (1981). Serial single-fiber EMG studies in myasthenic patients treated with corticosteroids and plasma exchange therapy. Muscle Nerve 4: 254. Kim DS, Claussen GC, Oh SJ (1998). Single-fiber electromyography improvement with 3,4-diaminopyridine in Lambert–Eaton myasthenic syndrome. Muscle Nerve 21: 1107–1108. Krendel DA, Sanders DB, Massey JM (1987). Single fiber electromyography in chronic progressive external ophthalmoplegia. Muscle Nerve 10: 299–302. Mandler RN, Maselli RA (1996). Stimulated single-fiber electromyography in wound botulism. Muscle Nerve 19: 1171–1173. Massey JM, Sanders DB, Nandedkar SD (1985). Sensitivity of various EMG techniques in motor neurone disease. Electroencephalogr Clin Neurophysiol 61: 574–575. Oh SJ, Kim DE, Kuruoglu R et al. (1992). Diagnostic sensitivity of the laboratory tests in myasthenia gravis. Muscle Nerve 15: 720–724. Padua L, Aprile I, Lo Monaco M et al. (1999). Neurophysiological assessment in the diagnosis of botulism: usefulness of single-fiber EMG. Muscle Nerve 22: 1388–1392.
Phillips LH (1982). The use of single fiber EMG to monitor the response of Lambert–Eaton myasthenic syndrome to steroid therapy. Muscle Nerve 5: 555. Sanders DB (1992). The effect of firing rate on neuromuscular jitter in Lambert–Eaton myasthenic syndrome. Muscle Nerve 15: 256–258. Sanders DB (2002). Clinical impact of single-fiber electromyography. Muscle Nerve (Suppl 11): S15–S20. Sanders DB, Howard JF (1986). Single-fiber electromyography in myasthenia gravis. Muscle Nerve 9: 809–819. Sanders DB, Howard JF, Johns TR (1979). Single-fiber electromyography in myasthenia gravis. Neurology 29: 68–76. Sanders DB, Sta˚lberg EV (1996). Single-fiber electromyography. Muscle Nerve 19: 1069–1083. Schiller HH, Sta˚lberg E (1978). Human botulism studied with single fibre electromyography. Arch Neurol 35: 346–349. Schwartz MS, Sta˚lberg E, Schiller HH et al. (1976). The reinnervated motor unit in man. A single fibre EMG multielectrode investigation. J Neurol Sci 27: 303–312. Sta˚lberg E (1966). Propagation velocity in human single muscle fibres in situ. Thesis Acta Physiol Scand 70 (Suppl. 287): 1–112. Sta˚lberg EV (1983). Macro EMG. Muscle Nerve 6: 619–630. Sta˚lberg EV, Sanders DB (2009). Jitter recordings with concentric needle electrodes. Muscle Nerve 40: 331–339. Sta˚lberg E, Sanders DB, Ali S et al. (2016). Reference values for jitter recorded by concentric needle electrodes in healthy controls: a multicenter study. Muscle Nerve 53: 351–362. Sta˚lberg E, Trontelj JV (1992). Clinical neurophysiology: the motor unit in myopathy. In: LP Rowland, S DiMauro (Eds.), Handbook of clinical neurology, myopathies. Elsevier, Amsterdam, pp. 49–84. Sta˚lberg E, Trontelj JV, Sanders DB (2010). Single fiber electromyography: studies in healthy and diseased muscle, third edn. Edshagen Publishing House, Fiskeb€ackskil, Sweden. Sta˚lberg E, Trontelj JV, Sanders DB (2018). www.sfemg.info. Accessed March 15, 2018. Thiele B, Sta˚lberg E (1974). The bimodal jitter: a single fibre electromyographic finding. J Neurol Neurosurg Psychiatry 37: 403–411. Trontelj JV, Khuraibet A, Mihelin M (1988). The jitter in stimulated orbicularis oculi muscle: technique and normal values. J Neurol Neurosurg Psychiatry 51: 814–819. Trontelj JV, Mihelin M, Fernandez JM et al. (1986). Axonal stimulation for endplate jitter studies. J Neurol Neurosurg Psychiatry 49: 67–685. Trontelj JV, Sta˚lberg E (1991). Single motor end-plates in myasthenia gravis and LEMS at different firing rates. Muscle Nerve 14: 226–232. Trontelj JV, Sta˚lberg E (1992a). Jitter measurement by axonal micro-stimulation. Guidelines and technical notes. Electroencephalogr Clin Neurophysiol 85: 30–37. Trontelj JV, Sta˚lberg E (1992b). The effect of firing rate on neuromuscular jitter in Lambert–Eaton myasthenic syndrome: a reply. Muscle Nerve 15: 258.