Hand transplantations have been performed in a few individuals, who have recovered several hand functions after the transplant (1). Sensory and motor recovery occurs in the transplanted hands of these patients even many years after the transplant (2, 3). The muscle fibers in a transplanted hand remain denervated for a long period of time after the transplant: the first signs of reinnervation of intrinsic hand muscles in one of these patients were observed 11 months postoperatively (2). Although the patients follow physiotherapy during the period of denervation, the intrinsic muscles of the hand experience some degree of atrophy due to inactivity (1). This prolonged inactivity may change the electrophysiological membrane properties of muscle fibers, as observed in long-term denervation (4, 5). Although we previously analyzed the time to reinnervation of motor units in a transplanted hand (2) and the variability in discharge of these motor units (6), there are no reports on the electrophysiological properties of the fiber membrane in muscles of a transplanted hand. Electrophysiological membrane properties can be investigated indirectly by measuring in vivo the velocity of propagation of the action potentials along the muscle fibers (muscle fiber conduction velocity [CV]). For these investigations, the noninvasiveness of the measuring technique is a necessary prerequisite because of the risk of infections in the transplanted hand. In this study, we focus on the detailed analysis of CV of muscle fibers in individual motor units of a transplanted hand and its dependence on the history of activation of the motor unit. This analysis complements our previous investigations that focused on the neural control of reinnervated motor units (2, 6).
Muscle fiber CV can be noninvasively estimated in individual motor units (7). In able-bodied individuals, the range of CV of motor unit action potentials is 2 to 6 m/sec (8) and depends on the time interval between discharges (9, 10). The CV of an action potential can be partly predicted by measuring the time interval separating the action potential from the preceding discharge and the baseline value of CV, that is, the CV of an isolated discharge, although the entire history of activation may have an influence. The relationship between CV and the interval separating two activations of the fiber membrane is known as the velocity-recovery function (VRF) of muscle fibers (10).
The aim of the study was to investigate the baseline value of muscle fiber CV and the dependence of CV on the discharge interval in individual motor units of an intrinsic muscle of a transplanted hand to determine whether the electrophysiological properties of the muscle fibers are consistent with able-bodied individuals.
One motor unit could be identified in all sessions after the first sign of reinnervation of the abductor digiti minimi muscle. Propagating action potentials with negligible changes in shape along some of the channels of the array were detected in all experimental sessions (Fig. 1). Motor unit activity in the abductor digiti minimi was observed starting from 11 months postoperatively (2), and thus in 10 of the 11 experimental sessions. A total of 13 ramp and 14 maximal effort contractions were analyzed from these 10 sessions. The other recordings were excluded from the analysis due to poor signal quality.
Figure 2 shows an example of multichannel recordings and the identified action potentials belonging to a single motor unit in a ramp contraction. Characteristics of the discharge rate of the analyzed motor units were previously described (6). Briefly, the average discharge rate increased from 17.4±4.3 pulses per second (pps) to 22.1±5.0 pps during the 60-sec ramp contraction and decreased from 27.1±8.4 pps to 17.2±2.9 pps during the maximal contractions. The coefficient of variation for interspike interval (ISI) was on average 35.9%±7.4% and 36.8%±10.8% for the ramp and maximal contractions, respectively, larger than the values usually observed in able-bodied individuals (6). This large variability provided a wide range of values for instantaneous discharge rate, necessary for a robust estimation of the dependence on discharge rate.
Muscle fiber CV was estimated from a number of channels in the range 3 to 6 (median value, 4 channels). The average value of CV in the ramp contractions (average over all sessions and discharges) was 4.6±1.1 m/sec, which is within the physiological range usually observed in able-bodied individuals (8). The CV estimated at the beginning of the maximal effort contractions was 4.7±1.0 m/sec; this value was obtained from discharges with an average ISI of 36.9 msec. This value of CV was not significantly different from the value estimated at similar discharge rates (interval bin, 30–40 msec) during the ramp contractions (4.5±1.1 m/sec).
The following results were obtained from the ramp contractions, unless otherwise indicated. The CV value did not depend on the experimental session after reinnervation, neither when all discharges in each session were pooled together nor when the discharges were separated in the defined ISI bins. Thus, values for CV were grouped for all experimental sessions. When grouped together, values of CV estimated in ISI bins containing intervals more than 1000 msec did not differ between each other. Thus, it was assumed that for ISIs more than 1000 msec the action potentials were isolated. The average CV of the isolated action potentials was 3.8±0.3 m/sec and was considered as the baseline value. The dependence on ISI was computed by normalizing (%) the CV values to the baseline value for each contraction. The minimum ISI in the computation of the VRF was 4.7 msec and the mean value for ISIs in the first bin was 7.7 msec. This value corresponded to the closest occurrence of pairs of action potentials. The mean value of the ISIs in the last bin considered was 1440 msec.
CV changed according to the instantaneous discharge rate during each contraction. Figure 3 shows a representative example of instantaneous discharge rate and CV computed for each discharge identified during one of the contractions. In this example, CV was almost linearly related to discharge rate, which is an observation in accordance with the effect of the VRF in able-bodied subjects.
Figure 4 shows the dependence of CV on ISI estimated from the patient using all discharges from the 13 ramp contractions. A one-way analysis of variance (ANOVA) of normalized CV with factor the bin of ISI was significant (F=19.7, P<0.0001; Fig. 4), indicating a dependency of CV on ISI (logarithmic fitting; R2=0.95).
Individual motor units were identified from the abductor digiti minimi muscle of a patient with a transplanted hand. The CV with which the action potentials of single motor units propagated was on average approximately 3.8 m/sec, which is within the range of physiological values observed in able-bodied individuals. For example, Farina et al. (7) reported values of approximately 3 m/sec for CV of motor units in the abductor digiti minimi muscle of healthy individuals at low discharge rates. Similar values have been obtained in other hand muscles of healthy subjects (11, 12).
The measured baseline CV in the transplanted hand was not different among experimental sessions; since the first sign of reinnervation, CV values were within the physiological range (8). This is at variance with results in patients after prolonged denervation, who present with CV values lower than healthy subjects (5), with a progressive increase after reinnervation (4). Similarly, muscle fiber CV decreases after a period of immobilization (4). However, there are several methodological differences between the present study and those from denervated or immobilized muscles. In this study, CV was assessed only from the motor units that were reinnervated and not from a random sampling of muscle fibers obtained with invasive electrical stimulation at various stages of denervation, as in other studies (4, 5). The monitoring of denervated muscle fibers was not possible in this study because invasive techniques could not be applied. The first values of CV were obtained during the experimental session in which the first action potentials were recorded; this session was 4 months after the first measure. The present results thus indicate that in less than 4 months the reinnervated muscle fibers reached values of CV of their action potentials similar to those observed in able-bodied individual and these values were then stable over repeated measures.
The observation that values of CV in this patient are similar to healthy individuals is interesting also considering that the patient was under chronic administration of prednisone (6). It has been demonstrated that muscle fiber CV may be reduced after glucocorticoid administration in healthy (13) and pathological subjects (14) and in patients with Cushing's syndrome (15), which is partly in contradiction with the present findings. Nevertheless, in those studies, the muscles investigated were the biceps brachii and lower limb muscles, whereas there is lack of data on the glucocorticoid actions on the small muscles of the hand. However, it is well known that in the classic (chronic) form of steroid myopathy, facial, sphincter, and hand muscles are usually spared, whereas patients note difficulties in the proximal muscles of the upper and lower limb, which is in agreement with the current findings of unchanged CV in the patient investigated.
In the analyzed patient, it was possible to estimate the dependence of CV on ISI for a relatively large range of ISIs (on average, 7.7–1440 msec). This is usually not possible in able-bodied individuals during voluntary contractions because of the difficulty in identifying action potentials of individual motor units with surface recordings. Thus, data on the variation in CV with ISI in healthy subjects were previously mainly obtained with electrical stimulation (9). Using intramuscular microstimulation of muscle fibers, Mihelin et al. (9) showed that in healthy individuals CV values were below baseline for interstimulus intervals smaller than approximately 6 msec (subnormal phase) followed by a peak value and an exponential decrease to baseline, similar to previous results by Stålberg (10). The subnormal phase in the CV was not observed in the current study because there were no action potentials with a distance smaller than 4 msec. Moreover, the measured CV did not show a peak value but a monotonic decrease with increasing ISI (Fig. 4), which is also expected from results in healthy individuals. In healthy subjects, the peak in the VRF was indeed reported to occur at ISIs in the range 5 to 15 msec (8) and 8 to 50 msec (9) and showed a large variability among the muscle fibers analyzed. In a study using transcutaneous electrical stimulation (16), the peak in the VRF assessed from the compound action potential occurred for ISIs in the range 6 to 50 msec across subjects. The monotonic decrease in CV observed in this study is explained by the range of investigated ISIs, whose lower limit was on average 7.7 msec, probably larger than the interval corresponding to the peak value. It is, however, worth noting that, as it has been shown recently (17), the characteristics of the VRF obtained by paired electrical stimuli do not necessarily reflect the activation history during voluntary contractions. Indeed, during voluntary activation, it is the entire history of discharge that influence the value of CV for each discharge (17) and not only the previous ISI. When pooling together several discharges with different ISIs and different discharge histories, it is likely that on average the association between CV and ISI would resemble some of the properties of the VRF estimated from paired stimuli with some variability. This may explain the rather large variability observed in the association between CV and ISI in this study.
The observed maximum value for the VRF is consistent with previous observations in healthy subjects. CV was shown to reach values higher than baseline by 12.5% to 24% (10) and by 4.86% to 15.7% (9) in single muscle fibers, and by 4.6% to 15% in the evoked compound potential (16). Accordingly, the maximum value recorded in this study was approximately 25% higher than baseline (Fig. 4). Finally, the duration of the ISI for which the effect of the previous action potential became negligible was similar with respect to observations in healthy individuals (10).
The relationship between CV and ISI was interpolated with a logarithmic function. This relationship implies a direct correlation between instantaneous discharge rate and CV, as it was visually observed in all recordings (e.g., Fig. 3). The relationship was nonlinear and corresponded to larger sensitivity of CV to discharge rate for low rates and a saturation (expressed by the logarithmic function CV (%)= 4.4 · ln(DR)+101.5) for high discharge rates. For example, the variation in CV with discharge rate would be reduced from 0.44%/pps at an average rate of 10 pps to 0.11%/pps at an average rate of 40 pps.
The sensitivity of CV to discharge rate was smaller than in some previous single motor unit studies during voluntary contractions in able-bodied individuals, although the variability of this relationship in healthy subjects is large (7). In able-bodied individuals, an average slope of this relation of approximately 1.5%/pps was observed in the abductor digiti minimi muscle (7) and of approximately 1.3%/pps in the biceps brachii muscle (18). However, these values strongly depend on the range of discharge rates analyzed, which should be sufficiently large to reduce the variability of the measure. In studies where a large range of discharge rates was analyzed, the slope of the relationship was similar to this study (e.g., approximately 0.6%/pps for neck muscles ).
In summary, the baseline value of CV and the dependence of CV on ISI in individual motor units of the abductor digiti minimi muscle of a transplanted hand were similar as in able-bodied individuals. The mechanisms affecting the propagation of action potentials along the muscle fibers and its dependency on discharge rate are thus preserved in the transplanted hand even after several months of denervation.
MATERIALS AND METHODS
The measures were performed on a man who had lost his right hand (dominant) in a farming accident when he was 13 years of age. At the age of 35 years, he underwent unilateral hand transplantation (2). The deceased donor was a 43-year-old man who had died from a stroke. The surgical procedure and immunosuppressive regimen were described in detail previously (20). The Italian Hand Transplantation Program was approved by the Italian Health Department and the Ethical Committee of the University of Milan-Bicocca. The measures described in this study have been approved by the Ethical Committee of the University of Milan-Bicocca.
Physiotherapy began after the swelling subsided and included standard rehabilitation for flexor and extensor tendons, including passive mobilization of the wrist and hand, active-assisted movements and active wrist- extension. When performance of active tasks of the fingers and the wrist improved, the rehabilitation focused on increasing tendon gliding, flexor-extensor balance, and grip strength. Electrotherapy was also applied but only to extrinsic muscles.
The motor unit analysis was performed with surface electromyography (EMG) recordings. The recording method has been previously described (2, 6). Multichannel surface EMG signals were detected from several intrinsic muscles of the transplanted hand (2). The first set of recordings was obtained 7 months postoperatively, followed by a second evaluation at 11 months and then monthly thereafter, until 10 sessions had been completed. An additional session was performed 4 months after the 10th session, for a total of 11 sessions. As in a previous report (6), this study describes results from the abductor digiti minimi, which was the only muscle of the transplanted hand where the discharges of individual motor units could be identified in repetitive sessions after reinnervation.
The skin overlying the muscle was lightly abraded with paste to improve the quality of the skin-electrode contact. A linear array of 16 silver electrodes (1 mm diameter, 2.5 mm interelectrode distance) was held over the muscle by an operator during the recordings. The electrode location and direction were chosen to detect signals showing propagation of motor-unit action potentials along the array with minimal shape changes during propagation. The surface EMG signals were amplified by a multichannel amplifier (EMG 16, LISiN; Ottino Bioelettronica, Torino, Italy) and band-pass filtered (−3 dB bandwidth, 10–500 Hz), sampled at 2048 samples/sec, converted to digital data by a 12-bit A/D converter board, displayed in real time, and stored on a PC for further processing. The EMG signals were processed on-line by a custom-made instrument that generated an analog signal proportional to the discharge rate of the identified motor unit. The discharge rate signal was displayed on an oscilloscope as visual feedback to the subject.
The subject was asked to perform three 60-sec contractions of maximal effort and three 60-sec contractions that involved linearly increasing the muscle activity from the minimum to maximum using the visual feedback on motor unit discharge rate (ramp contractions). The subject rested 5 min between contractions.
The action potentials discharged by single motor units were identified off-line from the EMG signal (21). The accuracy in the detection of action potentials was approximately 100% in this application because of the small number of reinnervated motor units (6). The muscle fiber CV was estimated for each identified discharge with a multichannel delay estimation algorithm (22). The channels chosen for CV estimation were those for which the shape of the action potential remained maximally stable (visual inspection of the signals), which is a necessary condition for accurate determination of CV (23). To assess the dependence of CV on ISI, the CV estimated from each action potential was expressed as a function of the interval of time separating the potential from the previous discharge (Fig. 1). Because CV was not affected by the time delay from the previous discharge when this was more than 1000 msec (see Results), and the baseline value for CV was estimated as the average velocity of the action potentials separated by previous action potentials by more than 1000 msec. This condition approximated isolated discharges. The CV values were normalized (%) to the baseline value to obtain the relation with ISI (10).
Because a potential effect of fatigue on fiber membrane properties would be less pronounced during ramp with respect to maximal effort contractions, the VRF was estimated from the recordings during the ramp contractions only. The maximal effort contractions were used to exclude a significant effect of fatigue during the ramp contractions. For this purpose, CV estimated at the beginning of the maximal effort contractions was compared with CV values estimated during the ramp contractions at similar discharge rates. Discharge rates similar to those achieved at the beginning of the maximal contractions were reached only at the end of the ramp contractions, and thus any differences in estimated CV between the two contractions would indicate an effect of fatigue during the ramp contractions.
The values for CV obtained in different experimental sessions were compared with a one-way ANOVA with factor the experimental session. For investigating the dependence of CV on the ISI, the ISI was divided in 24 interval bins. The interval bins for ISI had duration of 10 msec in the range 0 to 100 msec, 25 msec between 125 and 200 msec, 50 msec between 250 and 500 msec, and 250 msec between 1000 and 1500 msec. Thus, the first two interval bins contained discharges occurring at less than 20 msec from each other. A one-way ANOVA was used to assess the dependency of CV on the interval bin. The function expressing CV for these interval bins was interpolated with a logarithmic fit. When ANOVA was significant, pair-wise comparisons were obtained with the post hoc Student-Newman-Keuls test. The value of CV in the first 2 sec of the maximal contractions was compared with the value of CV during the ramp contractions at the same discharge rate with a paired Student's t test, to verify the absence of fatigue. Results are presented as mean and SD in the text and mean and SE in the figures.
1. Dubernard JM, Owen E, Lanzetta M, et al. What is happening with hand transplants? Lancet
2001; 357: 1711.
2. Lanzetta M, Pozzo M, Bottin A, et al. Reinnervation of motor units in intrinsic muscles of a transplanted hand. Neurosci Lett
2005; 373: 138.
3. Owen E, Dubernard JM, Lanzetta M, et al. Peripheral nerve regeneration in human hand transplantation. Transplant Proc
2001; 33: 1720.
4. Cruz-Martínez A, Arpa J. Muscle fiber conduction velocity
in situ (MFCV) in denervation, reinnervation and disuse atrophy. Acta Neurol Scand
1999; 100: 337.
5. Hofer C, Forstner C, Mödlin M, et al. In vivo assessment of conduction velocity and refractory period of denervated muscle fibers. Artif Organs
2005; 29: 436.
6. Farina D, Pozzo M, Lanzetta M, et al. Discharge variability of motor units in an intrinsic muscle of transplanted hand. J Neurophysiol
2008; 99: 2232.
7. Farina D, Gazzoni M, Camelia F. Low-threshold motor unit
membrane properties vary with contraction intensity during sustained activation with surface EMG visual feedback. J Appl Physiol
2004; 96: 1505.
8. Troni W, Cantello R, Rainero I. Conduction velocity along human muscle fibers in situ. Neurology
1983; 33: 1453.
9. Mihelin M, Trontelj J, Stålberg E. Muscle fiber recovery functions studied with double pulse stimulation. Muscle Nerve
1991; 14: 739.
10. Stålberg E. Propagation velocity in human muscle fibers in situ. Acta Physiol Scand
1966; 287: 1.
11. Gazzoni M, Camelia F, Farina D. Conduction velocity of quiescent muscle fibers decreases during sustained contraction. J Neurophysiol
2005; 94: 387.
12. Schulte E, Farina D, Rau G, et al. Single motor unit
analysis from spatially filtered surface electromyogram signals. Part 2: Conduction velocity estimation. Med Biol Eng Comput
2003; 41: 338.
13. Minetto MA, Botter A, Lanfranco F, et al. Muscle fiber conduction slowing and decreased levels of circulating muscle proteins after short-term dexamethasone administration in healthy subjects. J Clin Endocrinol Metab
14. van der Hoeven JH. Decline of muscle fiber conduction velocity
during short-term high-dose methylprednisolone therapy. Muscle Nerve
1996; 19: 100.
15. Minetto MA, Lanfranco F, Botter A, et al. Do muscle fiber conduction slowing and decreased levels of circulating muscle proteins represent sensitive markers of steroid myopathy? A pilot study in Cushing's disease. Eur J Endocrinol
2011; 164: 985.
16. Kamavuako EN, Hennings K, Farina D. Velocity recovery function of the compound muscle action potential assessed with doublet and triplet stimulation. Muscle Nerve
2007; 36: 190.
17. McGill KC, Lateva ZC. History dependence of human muscle-fiber conduction velocity during voluntary isometric contractions. J Appl Physiol
2011; 111: 630.
18. Nishizono H, Kurata H, Miyashita M. Muscle fiber conduction velocity
related to stimulation rate. Electroencephalogr Clin Neurophysiol
1989; 72: 529.
19. Farina D, Falla D. Effect of muscle-fiber velocity recovery function on motor unit
action potential properties in voluntary contractions. Muscle Nerve
2008; 37: 650.
20. Dubernard JM, Owen E, Herzberg G, et al. Human hand allograft: Report on first 6 months. Lancet
1999; 353: 1315.
21. Gazzoni M, Farina D, Merletti R. A new method for the extraction and classification of single motor unit
action potentials from surface EMG signals. J Neurosci Methods
2004; 136: 165.
22. Farina D, Muhammad W, Fortunato E, et al. Estimation of single motor unit
conduction velocity from surface electromyogram signals detected with linear electrode arrays. Med Biol Eng Comput
2001; 39: 225.
23. Farina D, Merletti R. Methods for estimating muscle fibre conduction velocity from surface electromyographic signals. Med Biol Eng Comput
2004; 42: 432.