Abstract

With the development of microsurgical techniques, replantation has become a feasible alternative to stump treatment after the amputation of an extremity. It is known that amputation often induces phantom limb pain and cortical reorganization within the corresponding somatosensory areas. However, whether replantation reduces the risk of comparable persisting pain phenomena as well as reorganization of the primary somatosensory cortex is still widely unknown. Therefore, the present study aimed to investigate the potential development of persistent pain and cortical reorganization of the hand and lip areas within the sensory cortex by means of magnetoencephalographic dipole analyses after replantation of a traumatically amputated upper limb proximal to the radiocarpal joint. Cortical reorganization was investigated in 13 patients with limb replantation using air puff stimulation of the phalanges of both thumbs and both corners of the lower lip. Displacement of the centre of gravity of lip and thumb representations and increased cortical activity were found in the limb and face areas of the primary somatosensory cortex contralateral to the replanted arm when compared to the ipsilateral hemisphere. Thus, cortical reorganization in the primary somatosensory cortex also occurs after replantation of the upper extremity. Patients’ reports of pain in the replanted body part were negatively correlated with the amount of cortical reorganization, i.e. the more pain the patients reported, the less reorganization of the subjects’ hand representation within the primary somatosensory cortex was observed. Longitudinal studies in patients after macroreplantation are necessary to assess whether the observed reorganization in the primary somatosensory cortex is a result of changes within the representation of the replanted arm and/or neighbouring representations and to assess the relationship between the development of persistent pain and reorganization.

Introduction

Traumatic amputation of the upper extremity is a severe injury that affects many aspects of the subject’s life and often results in painful phenomena such as phantom limb pain (Faggin et al., 1997; Weiss et al., 1998, 1999; Kooijman et al., 2000; Flor et al., 2006). With the development of microsurgical techniques, replantation of the amputated extremity has become a valuable alternative to amputation and its numerous negative consequences. Although the functionality of smaller replanted body parts, such as fingers, can be restored successfully, macroreplantation of an upper extremity proximal to the radiocarpal joint constitutes a large challenge for restoration of sensory and motor functions (Hirche and Germann, 2010).

It is widely known that deafferentation often results in persistent pain in the deafferented body part. More than 80% of amputees experience painful phenomena in the deafferented region (Flor, 2002). Research on patients with replantations at the upper extremity has shown that most of them suffer from persistent and at least moderate pain in the replanted hand or arm (Wiech et al., 2000). Nevertheless, recent estimates suggest that in ~10% of patients who suffer from a traumatic limb amputation, the development of severe chronic pain and mental health problems can be avoided by replantation (Hirche and Germann, 2010). Furthermore, the functional results and the regained sensation of a replanted extremity are equal or superior to a prosthesis (Hierner and Berger, 2005). Moreover, limb loss is accompanied by depression and anxiety in a substantial proportion of amputees (Darnall et al., 2005; Desmond, 2007). Therefore, the maintenance of the body integrity by macroreplantation justifies the choice of replantation over prosthesis as it allows for a considerable increase in quality of life (Graham et al., 1998; Hierner and Berger, 2005). Therefore, macroreplantation is considered the option of choice after traumatic loss of a limb.

Although microsurgery has considerably improved during the past decades, the sensory and functional outcome of replantation surgery do not parallel the improvements in microsurgical techniques (Lundborg, 2003). Replantation causes enormous changes both in the peripheral nervous system and in cortex (Röricht et al., 2001; Brenneis et al., 2005; Björkman et al., 2007a, b; Piza-Katzer et al., 2007; Eickhoff et al., 2008). At present, few studies have investigated the neural consequences of deafferentation in patients whose traumatically dissected or amputated limb or other body parts have become successfully replanted. Björkman et al. (2007a) investigated by means of functional MRI the time course of responses in primary (SI) and secondary (SII) somatosensory cortices at 1, 2, 4, 8 and 12 months to mechanical sensory stimulation applied to the hand of a patient who had experienced replantation of his traumatically lost hand at wrist level. The authors found that within the first month after replantation, stimulation of the replanted hand exclusively activated structures of the ipsilateral SI. Two months later, activation of SI was observed bilaterally. As more time elapsed, activation increasingly switched to the contralateral SI. Although changes in activation within sensory and motor cortices have been observed after replantation, little is known about possible changes of the homuncular organization of sensorimotor cortices caused by the replantation and the potential and gradual reinnervation of the replanted extremity. Besides the observations of Björkman et al., 2007a, b, long-term effects of macroreplantations of upper extremities or other body parts on the cortex are still unknown. However, these results also imply that peripheral mechanisms might not fully account for unsatisfactory restoration of sensory and motor functions following replantation. Central processes of plasticity taking place after replantation of traumatically amputated body parts might also play an important role in replantation success.

Extensive research has revealed that any surgical deafferentation of peripheral nerves is followed by functional and structural changes both of cortical areas and lower parts of the neuropil that process sensory information from the amputated body part. In animals whose upper extremity, or parts thereof, is amputated, an invasion of the representation zone of the amputated body parts by neighbouring structures has been reported (Merzenich et al., 1984; Pons et al., 1991; Florence and Kaas, 1995). Evidence has been provided by the combination of magnetoencephalography (MEG) and functional MRI that reorganization of SI also occurs in humans after amputation (Elbert et al., 1994; Yang et al., 1994a, b; Knecht et al., 1996, 1998; Flor et al., 1998; Weiss et al., 1998, 2000). As functional and morphological changes of the deafferented area itself cannot be directly addressed by peripheral stimulation in amputees, most studies have assessed such modulations indirectly by investigating expansions of the functional activation of neighbouring representation areas into the representation area of the amputated limb and its relation to pain and other consequences of deafferentation. Most of these studies indicated that the functional and morphological shrinkage of an amputated limb’s representation area within SI and the invasion of activity into this area from neighbouring areas was significantly and positively related to the amount of the subjects’ pain following amputation and named this maladaptive plasticity (Flor et al., 1995; Birbaumer et al., 1997; Lotze et al., 1999). Recently, Makin et al. (2013) and Preißler et al. (2013) suggested the co-existence of two processes: (i) maladaptive plasticity driven by deafferentation; and (ii) processes driven by chronic pain. In the persistent representation model proposed by Makin et al. (2013) it is suggested that the strength of phantom sensation or phantom limb pain is positively associated with the activity of the brain region and the size of the representation area corresponding to the deafferented or amputated body part, as would be the case with experience-dependent plasticity. Previous evidence has shown that increased use or more frequent sensory stimulation of a body part results in an enlarged representation zone of this body part in the contralateral SI (Elbert et al., 1995; Sterr et al., 1998a, b), whereas reduced use of a body part is known to result in a decrease of its cortical map in SI and often in an expansion of the representation area of adjacent body parts (Candia et al., 2003; Lissek et al., 2009; Langer et al., 2012).

Greater knowledge concerning the central consequences of amputation and replantation of lost body parts represent key factors with regard to improvement of the functional and pain-related outcome of replantations (Lundborg, 2003). A better understanding of the association between pain phenomena and neural plasticity following replantation of the lost body part might thus significantly improve the functional and psychological outcome of replantation as well as the patients’ satisfaction with such a treatment. Patients who have undergone replantation immediately after a traumatic loss of an upper extremity are particularly interesting and suitable for the investigation of the interplay between different mechanisms contributing to cortical reorganization. As the extremity is replanted at its original body site, replantation patients provide researchers with the opportunity for examining changes within the formerly deafferented area itself.

Based on clinical and empirical evidence in patients with amputations we expect that patients who received replantation of their lost body parts might experience pain in the replanted body part even several years after replantation. We further expect that there are two processes modulating the functional representation of the replanted upper extremity: (i) loss of sensory input into the hand area of SI directly after amputation leads to an expansion of the neighbouring representation area (e.g. face as tested with lip stimulation) into the deafferented representations (similar to amputees without consecutive replantation and phantom limb pain); and (ii) gradual and partial reinnervation enables reorganization of the formerly deafferented area representing the replanted arm (as tested with stimulation of the thumb).

Materials and methods

Subjects

Nineteen patients (17 male, two female, 23–69 years old, mean age 44 years) who underwent immediate replantation following traumatic loss of an upper extremity proximal to the radiocarpal joint, participated in the study. Information about the composition of the sample and the dropout rate is provided in Fig. 1. Characteristics of subjects are shown in Table 1. Informed consent was received from each participant. All procedures were conducted in accordance with the Declaration of Helsinki on human experimentation. The study was approved by the Ethics Committee of Friedrich Schiller University.

Figure 1

Flow chart of included patients.

Figure 1

Flow chart of included patients.

Table 1

Demographic and replantation specific details

Patient Sex Age at replantation Age at study TSR Dominant side Replantation side Location of amputation Type of amputation Affected arm nerves Pain intensity Pain severity code MEG data MRI available 
01 54 68 38 Upper arm Total U, M, R 83 Excluded 
02 60 62 173 Elbow Subtotal U, R n.a. n.a. 
03 34 40 71 Forearm Subtotal U, M, R 17 Excluded 
04 53 55 35 Forearm Subtotal U, M, R 40 X* 
05 17 36 235 Hand Subtotal 40 Excluded 
06 43 56 171 Forearm Subtotal U, M 37 
07 14 23 106 Forearm Subtotal U, M,R 20 
08 25 29 51 Upper arm Nerve root avulsion U, M,R 33 Excluded 
09 49 61 149 Upper arm Nerve root avulsion U, M,R 37 Excluded 
10 51 55 42 Forearm Subtotal U, M,R 90 Excluded 
11 43 47 50 Forearm Subtotal U, M,R 60 n.a. 
12 36 38 28 Upper arm Nerve root avulsion 97 n.a. Excluded 
13 63 69 75 Elbow Subtotal U, M,R 37 
14 51 51 Upper arm Total U,M,R n.a. n.a. 
15 20 33 161 Forearm Subtotal U, M 
16 25 25 Forearm Total U, M,R n.a. 
17 33 46 172 Upper arm Subtotal U, M,R 53 n.a. 
18 18 27 117 Forearm Subtotal U, M,R 37 
19 18 24 68 Forearm Subtotal U, M, R 
Mean ± SD  37.21 ± 16.01 44.47 ± 15.30 92.16 ± 67.21          
Patient Sex Age at replantation Age at study TSR Dominant side Replantation side Location of amputation Type of amputation Affected arm nerves Pain intensity Pain severity code MEG data MRI available 
01 54 68 38 Upper arm Total U, M, R 83 Excluded 
02 60 62 173 Elbow Subtotal U, R n.a. n.a. 
03 34 40 71 Forearm Subtotal U, M, R 17 Excluded 
04 53 55 35 Forearm Subtotal U, M, R 40 X* 
05 17 36 235 Hand Subtotal 40 Excluded 
06 43 56 171 Forearm Subtotal U, M 37 
07 14 23 106 Forearm Subtotal U, M,R 20 
08 25 29 51 Upper arm Nerve root avulsion U, M,R 33 Excluded 
09 49 61 149 Upper arm Nerve root avulsion U, M,R 37 Excluded 
10 51 55 42 Forearm Subtotal U, M,R 90 Excluded 
11 43 47 50 Forearm Subtotal U, M,R 60 n.a. 
12 36 38 28 Upper arm Nerve root avulsion 97 n.a. Excluded 
13 63 69 75 Elbow Subtotal U, M,R 37 
14 51 51 Upper arm Total U,M,R n.a. n.a. 
15 20 33 161 Forearm Subtotal U, M 
16 25 25 Forearm Total U, M,R n.a. 
17 33 46 172 Upper arm Subtotal U, M,R 53 n.a. 
18 18 27 117 Forearm Subtotal U, M,R 37 
19 18 24 68 Forearm Subtotal U, M, R 
Mean ± SD  37.21 ± 16.01 44.47 ± 15.30 92.16 ± 67.21          

F = female; M = male; L = left; R = right; U = ulnar nerve; M = median nerve; R = radial nerve; TSR = time since replantation in months. Pain intensity refers to pain in the replanted extremity. Some patients were excluded for the following reasons: Subject 05: amputation distally of the wrist; Subjects 03 and 10: additional thumb amputation at the replanted arm; and Subjects 08, 09 and 12: cervical root avulsions.

*Subject 04 had to be excluded from the group analyses of the lip representations as no distinct dipole solution could be determined after lip stimulation.

Pain assessment

Pain was assessed using a composite pain score which comprises three questions concerning pain intensities. On a scale ranging from 0, no pain, to 10, pain as bad as it ever could be, subjects were asked to rate their current pain intensity, average pain, and the pain maximum during the preceding 4 weeks. The mean value in response to these three questions multiplied by 10, constituted the average pain score.

Recordings of somatosensory-evoked magnetic fields

Somatosensory-evoked magnetic fields were recorded using a whole-head MEG (Elekta Neuromag®) with 306 independent sensors. The sensors are organized in 102 sensor locations, each consisting of two planar gradiometers and one magnetometer, resulting in three independent measures at each spatial sensor position. The planar gradiometers detect the strongest signal above the local source area while the magnetometers provide two maxima on the opposite sides of the source that characterize opposite field directions. During somatosensory-evoked magnetic field recordings, patients were lying on a bed inside a magnetically shielded room with their head placed in the mould of the dewar. Patients received light tactile stimulation at the phalanges of both thumbs (D1) and at both corners of the lower lip using a pneumatically driven stimulator (Somatosensory Stimulus Generator, 4-D NeuroImaging Inc.). The stimulator consisted of a thin rubber membrane with a diameter of 10 mm that tapped the skin when expanded by an air pressure pulse (25 psi). D1 and lips received ∼400 stimuli in random order at an interstimulus interval ranging from 800 to 1600 ms (0.625–1.25 Hz). Somatosensory-evoked magnetic fields were recorded for each single stimulus as single trial epochs of 500 ms including a 100 ms prestimulus interval. Somatosensory-evoked magnetic fields were sampled with 1000 Hz using a band-pass filter (0.1–330 Hz). The exact location of the head in relation to the MEG sensors was determined using four indicator coils that were positioned at specific sites on the subjects’ head. For detection of eye-movement artefacts, an electrooculogram was recorded. Two Ag/AgCl electrodes were attached vertically as well as horizontally. Trials with amplitudes exceeding 150 μV were excluded from all further analysis.

Somatosensory-evoked magnetic field data analysis was performed offline including Maxwell filtering (Taulu et al., 2005; Maxfilter Version 2.0.21) together with a time domain extension, baseline correction from −90 to 0 ms, and band-pass filtering between 0.1 and 100 Hz. Only half of the channels were used for data analysis, specifically sensors located over the hemisphere opposite to the stimulation. An example of averaged somatosensory-evoked magnetic fields to each of the stimulations is shown in Fig. 2; all other waveforms are shown in Supplementary Fig. 1. For dipole source analysis of the averaged somatosensory-evoked magnetic fields, a spherical head model was used as previous studies have shown that it provides good reliability when sources are located in SI (Schaefer et al., 2002). As in previous work (Gotz et al., 2011), an equivalent current dipole (ECD) model was applied to estimate the location and strength of the source. The identified dipoles were superimposed on individual MRIs. MRI recording was carried out with a 3 T system (Siemens MAGNETOM Trio, TIM System; 256 sagittal magnetic resonance images of 1-mm slice thickness, 256 × 256 matrix, 1 × 1 mm in-plane resolution). Magnetic resonance data were only available for 13 patients as four patients had magnetic implants, one was pregnant, and one subject refused the magnetic resonance participation because of claustrophobia within the scanner. In these six cases without MRI, the Elekta Neuromag internal device origin with Cartesian coordinates of 0, 0, 40 was used as the centre of the head sphere. Before the somatosensory-evoked magnetic field recordings, an individual head coordinate system was defined with a 3D digitizer (Isotrack, Polhemus Navigation Sciences) to locate the head position in the MEG dewar and to align the coordinate systems of MEG and MRI. The x-axis was determined by the auricular points, the y-axis passed through the nasion and the z-axis pointed upwards.

Figure 2

MEG waveforms (butterfly plots) for a typical subject. Left: somatosensory-evoked fields over the hemisphere contralateral to replantation. Right: somatosensory-evoked magnetic field over the hemisphere ipsilateral to replantation. Top left: somatosensory-evoked magnetic field to stimulation of the thumb (D1) on the replanted hand; upper right: somatosensory-evoked magnetic field to stimulation of D1 on the non-replanted hand. Bottom left: somatosensory-evoked magnetic field to lip stimulation on the side of replantation. Bottom right: somatosensory-evoked magnetic field to lip stimulation on the side contralateral to replantation (control side).

Figure 2

MEG waveforms (butterfly plots) for a typical subject. Left: somatosensory-evoked fields over the hemisphere contralateral to replantation. Right: somatosensory-evoked magnetic field over the hemisphere ipsilateral to replantation. Top left: somatosensory-evoked magnetic field to stimulation of the thumb (D1) on the replanted hand; upper right: somatosensory-evoked magnetic field to stimulation of D1 on the non-replanted hand. Bottom left: somatosensory-evoked magnetic field to lip stimulation on the side of replantation. Bottom right: somatosensory-evoked magnetic field to lip stimulation on the side contralateral to replantation (control side).

Table 2

Dipole solutions for lip representations in SI

Patient Hemisphere
 
Time Coordinates
 
GoF (%) Q (nAm) V (mm3
Regarding replant. Body side (ms) x y z 
Vp02 Contralateral 42 47.2 23.3 77.7 89.4 10 95.8 
Ipsilateral  71 −37.9 27.6 61.1 95.7 25.1 36.1 
Vp06 Contralateral 21 −37.8 26.1 67.7 91.9 12.9 184.1 
Ipsilateral  22 42.3 30.3 72.2 82.3 10.2 209.5 
Vp07 Contralateral 26 −48.8 31.4 79.3 89.5 9.8 41.4 
Ipsilateral  24 30.8 39.4 87.9 78.8 11.3 68.7 
Vp13 Contralateral 38 47.4 38.7 73 83.3 15.6 9.9 
Ipsilateral  38 −29.4 11.1 59.9 85.4 16.2 117.4 
Vp15 Contralateral 52 −41.3 34.4 72.5 91.4 16.5 8.9 
Ipsilateral  44 53.1 32.7 65.9 91.9 15.7 7.7 
Vp16 Contralateral 20 45.3 22.7 69.1 85.6 10.8 138.3 
Ipsilateral  18 −38 37.4 78.7 75.1 5.4 548.2 
Vp17 Contralateral 22 51.9 37.8 74.1 78.1 4.4 363.1 
Ipsilateral  22 −30.7 30.5 71.7 91.3 27.9 35.1 
Vp18 Contralateral 22 23.3 17.4 83.7 86.8 20.4 97.3 
Ipsilateral  22 −39.9 34.1 91.5 81.9 5.4 123.0 
Vp19 Contralateral 18 37.6 25.7 80.6 88.2 11.6 59.5 
Ipsilateral  20 −27.4 42.8 86.1 95.5 18.6 17.1 
Patient Hemisphere
 
Time Coordinates
 
GoF (%) Q (nAm) V (mm3
Regarding replant. Body side (ms) x y z 
Vp02 Contralateral 42 47.2 23.3 77.7 89.4 10 95.8 
Ipsilateral  71 −37.9 27.6 61.1 95.7 25.1 36.1 
Vp06 Contralateral 21 −37.8 26.1 67.7 91.9 12.9 184.1 
Ipsilateral  22 42.3 30.3 72.2 82.3 10.2 209.5 
Vp07 Contralateral 26 −48.8 31.4 79.3 89.5 9.8 41.4 
Ipsilateral  24 30.8 39.4 87.9 78.8 11.3 68.7 
Vp13 Contralateral 38 47.4 38.7 73 83.3 15.6 9.9 
Ipsilateral  38 −29.4 11.1 59.9 85.4 16.2 117.4 
Vp15 Contralateral 52 −41.3 34.4 72.5 91.4 16.5 8.9 
Ipsilateral  44 53.1 32.7 65.9 91.9 15.7 7.7 
Vp16 Contralateral 20 45.3 22.7 69.1 85.6 10.8 138.3 
Ipsilateral  18 −38 37.4 78.7 75.1 5.4 548.2 
Vp17 Contralateral 22 51.9 37.8 74.1 78.1 4.4 363.1 
Ipsilateral  22 −30.7 30.5 71.7 91.3 27.9 35.1 
Vp18 Contralateral 22 23.3 17.4 83.7 86.8 20.4 97.3 
Ipsilateral  22 −39.9 34.1 91.5 81.9 5.4 123.0 
Vp19 Contralateral 18 37.6 25.7 80.6 88.2 11.6 59.5 
Ipsilateral  20 −27.4 42.8 86.1 95.5 18.6 17.1 

GoF = Goodness of Fit statistics; Q = dipole strength; V = confidence volume; R = right hemisphere; L = left hemisphere.

Table 3

Distribution parameters of the dipole coordinates for lip representations (n = 9)

 Mean Median SD Minimum Maximum 
xrs 42.29 45.30 8.64 23.30 51.90 
xnrs 36.61 37.90 8.09 27.40 53.10 
yrs 28.61 26.10 7.34 17.40 38.70 
ynrs 31.87 32.70 9.16 11.10 42.80 
zrs 75.30 74.10 5.38 67.70 83.70 
znrs 75.00 72.20 11.72 59.90 91.50 
 Mean Median SD Minimum Maximum 
xrs 42.29 45.30 8.64 23.30 51.90 
xnrs 36.61 37.90 8.09 27.40 53.10 
yrs 28.61 26.10 7.34 17.40 38.70 
ynrs 31.87 32.70 9.16 11.10 42.80 
zrs 75.30 74.10 5.38 67.70 83.70 
znrs 75.00 72.20 11.72 59.90 91.50 

rs = side contralateral to replantation; nrs = side ipsilateral to replantation (not replanted).

The ECD of each stimulated body site was assessed at the earliest peak of magnetic activity in a time window from 20 ms to 80 ms with regard to D1 stimulation and from 10 ms to 60 ms with reference to lip stimulation, excluding a 35 ms delay of the pressure onset to the trigger. To ensure that the same components were compared within subjects, the earliest peak of magnetic activity that was present in both hemispheres was chosen. Dipole locations were accepted if the goodness of fit was better than r = 0.75, the ECD strength exceeded 4 nAm and the confidence volume was <1000 mm3.

To ensure comparability between subjects, only 13 of 19 participants were included in the MEG data analysis. Exclusion criteria were amputations distally of the wrist (one subject), additional thumb amputation at the replanted arm (two subjects), and cervical root avulsions (three subjects). ECD solutions after D1 stimulation could only be computed for 10 subjects as the implants of three participants interfered with the MEG measurement because of their magnetic features (cf. Table 1). In addition to these three patients, one further subject had to be excluded from the group analyses of the lip representations as no distinct dipole solution could be determined after lip stimulation.

Assessment of cortical reorganization

Euclidean distances between EDC locations of body parts adjacent to the amputated extremity have previously been used to demonstrate maladaptive plasticity reorganization in numerous studies in amputees (Yang et al., 1994a, b; Elbert et al., 1995; Knecht et al., 1998; Grüsser et al., 2004). Therefore, we used the Euclidean distance between mirrored left lip ECD localization and the right lip ECD localization to assess the extent of cortical reorganization triggered by sensory loss in the extremity. The y–z plane of the internal coordinate system of the MEG system served as the mirror plane. This plane divides the brain along the longitudinal fissure, so that mirroring left ECD simply changed the sign of the x coordinate.

Furthermore, it has been shown repeatedly that somatosensory deafferentation is accompanied by increased excitability of the area surrounding the deafferented body part (Rossini et al., 1994; Buchner et al., 1999; Karl et al., 2001; Weiss et al., 2004; Sens et al., 2012). Thus, we used the difference between the ECD strengths to left versus right lip stimulation as a second measure for the extent of maladaptive reorganization in SI.

Similarly, changes in hand representation were assessed by computing both Euclidean distances between the mirrored ECD localization for left D1 stimulation and the ECD localization for right D1 stimulation and the difference between ECD strengths for D1 of the replanted versus healthy arm. These parameters might critically depend on current pain, which was assessed by bivariate correlations.

MEG data were analysed with DANA Software (Release 3, Elekta AB). The locations of thumb and lip representations in the hemisphere contralateral to the replantation were considered to be significantly deviating from the corresponding representation in the hemisphere contralateral to the uninjured arm, if the Euclidean distances differ from each other by >6 mm on average. The critical value of 6 mm was chosen as former research with healthy subjects showed discrepancies of 3–6 mm with regard to the location of the representations of equal body parts in the two hemispheres when mirrored along the z-axis (Gallen et al., 1993; Grüsser et al., 2001; Schaefer et al., 2004).

Statistical analysis

All statistical tests were performed using SPSS Statistics (version 19.0 for Windows; IBM Corporation). Hypothesis testing was carried out using t-tests when data were normally distributed. In cases in which the data were not normally distributed, non-parametric tests (Wilcoxon signed-rank tests) were performed. For all the tests, we present test statistics, P-values, and estimate effect sizes (labelled with r). Relationships between variables were assessed by correlation analyses. For normally distributed data, Pearson correlations were computed, labelled by r. In case of non-normally distributed data, Spearman correlations were used, labelled by rs. The power values (1 − β) presented for the correlations were calculated using the free-ware program G*Power. Test family were t-tests, statistical test was correlation. As type of power analysis we used the post hoc test to compute the achieved power. Given the according correlation coefficient r and α = 0.05 as well as a sample size of 9 respective 10 the program calculated the power for the one-tailed correlation analysis.

Results

Pain in the replanted extremity

A one-sample t-test computed against a test value of 0, representing ‘no pain’, showed that subjects reported strong average pain score in the replanted body part during the preceding four weeks [mean = 40.00, standard error (SE) = 7.17, t(16) = 5.76, P < 0.001, r = 0.81]. Average pain score of 17 subjects was included in the analysis as one subject refused to fill out any questionnaires and one subject had to be excluded from this analysis as he was under acute pain medication.

Cortical reorganization

Cortical reorganization of the lip representation

Euclidean distances between mirrored left lip ECD localization and the right lip ECD localization varied in location by >20 mm [mean = 20.59 mm, standard deviation (SD) = 7.66 mm] significantly exceeding the assessed maximum for natural variability [t(8) = 5.72, P < 0.005, r = 0.90]. Dipole strengths of lip representations were compared between the two hemispheres [t(8) = -0.69, P = 0.54]. The dipole strength in the hemisphere contralateral to the replantation was mean = 12.44 nAm (SD = 4.62 nAm) compared with mean = 15.09 nAm (SD = 7.94 nAm) in the hemisphere ipsilateral to the replantation.

Cortical reorganization of the hand representation

Euclidean distance between mirrored ECD localization for left D1 stimulation and the ECD localization for right D1 stimulation revealed discrepancies of the location of the D1 in SI of >10 mm on average (mean = 10.51 mm, SD = 5.35 mm). A one-sided t-test against a test value of 6 mm was computed to assess whether the difference in the locations of the D1 representations was beyond natural variability, and could be considered a result of reorganizational processes. Euclidean distances between the bilateral D1 representations significantly exceeded the critical deviation of 6 mm [mean = 10.51 mm, SE = 1.69 mm, t(9) = 2.66, P = 0.026, r = 0.66, Fig. 3].

Figure 3

Coronar slice of a T1-weighted image including dipole projections for dipoles determined after thumb stimulation in one subject. The blue triangle represents the dipole determined after stimulation of the non-affected right thumb while the red square shows the dipole determined after stimulation of the left thumb which was part of the replanted extremity. The solid blue and red lines indicate the dipole orientations. The dashed line indicates that the dipole determined after stimulation of the non-affected right thumb was mirrored along the y–z plane of the internal coordinate system of the MEG system, which divides the brain along the longitudinal fissure. Within the contralateral hemisphere, the Euclidean distance between the dipole locations is computed as indicated by the white arrow.

Figure 3

Coronar slice of a T1-weighted image including dipole projections for dipoles determined after thumb stimulation in one subject. The blue triangle represents the dipole determined after stimulation of the non-affected right thumb while the red square shows the dipole determined after stimulation of the left thumb which was part of the replanted extremity. The solid blue and red lines indicate the dipole orientations. The dashed line indicates that the dipole determined after stimulation of the non-affected right thumb was mirrored along the y–z plane of the internal coordinate system of the MEG system, which divides the brain along the longitudinal fissure. Within the contralateral hemisphere, the Euclidean distance between the dipole locations is computed as indicated by the white arrow.

Mean dipole strength of the D1 dipole was 54.98 nAm (SD = 33.12 nAm) in the hemisphere contralateral to the replantation and 21.98 nAm (SD = 12.24 nAm) in the ipsilateral hemisphere, respectively. Dipole strength of the D1 dipole contralateral to the replanted extremity was significantly stronger than the dipole strength that was determined after stimulation of the uninjured D1 (median = 27.20, z = −2.90, P = 0.001, r = −0.87).

Relationship between measures of cortical reorganization and pain scores

Relationship between cortical reorganization of the lip area and pain severity

No significant correlation was found between the extent of cortical reorganization of the lip area, as measured by Euclidean distance, and the average pain score [r = 0.29, P (one-tailed) = 0.246, n = 8; R2 = 0.06, 1 − β = 0.21; cf. Fig. 4]. Moreover, there was no significant relationship between the difference of the dipole strengths of the lip representations and the average pain score [r = −0.28, P (one-tailed) = 0.250, n = 8; R2 = 0.08, 1 − β = 0.20; cf. Fig. 4].

Figure 4

Relationship between reorganization in the primary somatosensory cortex (SI) and average pain score. Top: reorganization of the SI representation of the thumb (D1). Bottom: reorganization of the SI representation of the lip. Reorganization was assessed as Euclidean distance (ED) between locations of ECD to the stimulation of lip/D1 contralateral to replantation and the mirrored ECD to the stimulation of lip/D1 contralateral to replantation (left) and as difference of ECD strength to these stimulations (right). Note that only the correlation between Euclidean distance between D1 ECDs and average pain score (top left) reached significance (P < 0.05).

Figure 4

Relationship between reorganization in the primary somatosensory cortex (SI) and average pain score. Top: reorganization of the SI representation of the thumb (D1). Bottom: reorganization of the SI representation of the lip. Reorganization was assessed as Euclidean distance (ED) between locations of ECD to the stimulation of lip/D1 contralateral to replantation and the mirrored ECD to the stimulation of lip/D1 contralateral to replantation (left) and as difference of ECD strength to these stimulations (right). Note that only the correlation between Euclidean distance between D1 ECDs and average pain score (top left) reached significance (P < 0.05).

Relationship between cortical reorganization of thumb representations and pain scores

Pearson correlations revealed a significant relationship between the average pain score and Euclidean distance between mirrored ECD localization for left D1 stimulation and the ECD localization for right D1 stimulation [r = −0.61, P (one-tailed) = 0.04, n = 9; R2 = 0.53, 1 − β = 0.72; cf. Fig. 4]. By contrast, the average pain score did not correlate significantly with the difference between the dipole strengths [rs = −0.48, P (one-tailed) = 0.09, n = 9; R2 = 0.23, 1 − β = 0.48; cf. Fig. 4].

We also controlled the correlations for the effect of third variables such as age at replantation, time since replantation, level of amputation (upper arm, elbow, and forearm), type of amputation (total versus subtotal), and affected arm nerves (median, ulnar, and radial nerve). This was achieved by computing two-tailed first-order partial correlations between the measures for cortical reorganization and average pain score while the influence of the aforementioned variables was held constant. The patients’ age at replantation had an impact on the association between the average pain score and the Euclidean distance between D1 ECDs (Table 6). The relationship between those variables also depended on whether N. medianus, N. ulnaris, and N. radialis were disrupted or whether one of those remained uninjured as was the case in some of the subtotal amputations. Holding the effect of the amputation level constant increased the correlative relationship between the average pain score and the Euclidean distance between D1 ECDs (Table 6). Given the small sample size, these results are not discussed in detail. Concerning the relationship between pain and differences in ECD strengths to D1 stimulations, partial correlation showed that stronger pain is associated with a smaller difference in ECD strength between the hemispheres (when the influence of time since replantation is controlled for).

None of the partial correlations for the relationship between the average pain score and all measures of reorganization for lip representation reached significance (Supplementary Table 1).

Discussion

The aims of the present study were to investigate whether amputation with consecutive replantation of the upper extremity is accompanied by persistent pain in the replanted body part and to assess cortical reorganization of SI with respect to the replanted limb as well as the representations neighbouring the replanted limb. Seventeen subjects with amputation and consecutive immediate replantation of one upper extremity reported persistent moderate pain in the replanted extremity with strong pain even many years after replantation. Somatosensory-evoked magnetic field recordings revealed reorganized D1 (n = 10) and lip (n = 9) representations in SI of the hemisphere contralateral to the replantation site when compared to the ipsilateral hemisphere. Stronger pain was associated with a smaller extent of cortical reorganization of D1 representation in the SI hand area. To our knowledge, only few studies have dealt with central changes after replantation. We are not aware of any study in replantation patients investigating changes in the homuncular organization.

Persistent pain in the replanted body part

Our study revealed that patients who have undergone macroreplantation of their upper extremity persistently experience moderate pain in the replanted extremity that is still present even 20 years after replantation. This is in line with former studies in replantation patients which showed that 3–83% of patients report mild to moderate pain in the replanted area (Ipsen et al., 1990; Hierner and Berger, 2005).

Reorganization of the representation of the lip

ECD locations of lip representations differed on average by 21 mm indicating a large effect size. Descriptive comparison of the mean coordinates of dipole locations in each hemispheres indicates a more lateral, posterior and inferior position of lip representation contralateral to the replantation as compared to the ipsilateral hemisphere. The corresponding distribution parameters of the dipole coordinates for lip representations are shown in Table 4.

Table 4

Dipole solutions for D1 representations in SI

Patient Hemisphere regarding replant Replanted side Time (ms) Coordinates
 
GoF (%) Q (nAm) V (mm3
x y z 
Vp02 Contralateral 61 41.8 32.9 87.8 98.2 92.5 0.1 
Ipsilateral  74 −47.7 33.3 74.2 94.9 18.5 13.3 
Vp04 Contralateral 60 35.5 22.6 84.9 93.7 19.3 9.3 
Ipsilateral  56 37.4 20.6 85.3 97.6 16.5 16.9 
Vp06 Contralateral 58 −44.8 28.4 81.8 96.7 54.9 0.5 
Ipsilateral  48 45.5 25.7 80.1 95.6 15.7 22.1 
Vp07 Contralateral 56 −39.5 23.4 91.4 98.2 68.3 0.1 
Ipsilateral  37 34.6 28.2 92.9 92.3 17.6 10.4 
Vp13 Contralateral 58 46.9 21.3 80.8 95.1 78.4 0.0 
Ipsilateral  54 −35.9 24.6 76.9 99.0 50.0 0.4 
Vp15 Contralateral 54 −38 23.9 76.4 98.3 113.9 0.0 
Ipsilateral  46 49.3 19 76.3 96.7 36.1 0.8 
Vp16 Contralateral 49 33.9 23.1 83.4 94.7 22.0 8.8 
Ipsilateral  32 −38.7 13.4 88.8 85.2 11.0 32.1 
Vp17 Contralateral 60 28.3 13.9 84.5 91.4 15.1 70.7 
Ipsilateral  62 −34.5 16.9 79 94.0 14.3 267.8 
Vp18 Contralateral 58 44.5 19.8 88.6 97.1 39.9 0.3 
Ipsilateral  57 −47.9 30.4 88.9 93.0 13.9 4.1 
Vp19 Contralateral 78 42.3 23.9 91.2 97.2 45.5 0.5 
Ipsilateral  60 −44.6 40.9 79.7 91.9 26.2 2.4 
Patient Hemisphere regarding replant Replanted side Time (ms) Coordinates
 
GoF (%) Q (nAm) V (mm3
x y z 
Vp02 Contralateral 61 41.8 32.9 87.8 98.2 92.5 0.1 
Ipsilateral  74 −47.7 33.3 74.2 94.9 18.5 13.3 
Vp04 Contralateral 60 35.5 22.6 84.9 93.7 19.3 9.3 
Ipsilateral  56 37.4 20.6 85.3 97.6 16.5 16.9 
Vp06 Contralateral 58 −44.8 28.4 81.8 96.7 54.9 0.5 
Ipsilateral  48 45.5 25.7 80.1 95.6 15.7 22.1 
Vp07 Contralateral 56 −39.5 23.4 91.4 98.2 68.3 0.1 
Ipsilateral  37 34.6 28.2 92.9 92.3 17.6 10.4 
Vp13 Contralateral 58 46.9 21.3 80.8 95.1 78.4 0.0 
Ipsilateral  54 −35.9 24.6 76.9 99.0 50.0 0.4 
Vp15 Contralateral 54 −38 23.9 76.4 98.3 113.9 0.0 
Ipsilateral  46 49.3 19 76.3 96.7 36.1 0.8 
Vp16 Contralateral 49 33.9 23.1 83.4 94.7 22.0 8.8 
Ipsilateral  32 −38.7 13.4 88.8 85.2 11.0 32.1 
Vp17 Contralateral 60 28.3 13.9 84.5 91.4 15.1 70.7 
Ipsilateral  62 −34.5 16.9 79 94.0 14.3 267.8 
Vp18 Contralateral 58 44.5 19.8 88.6 97.1 39.9 0.3 
Ipsilateral  57 −47.9 30.4 88.9 93.0 13.9 4.1 
Vp19 Contralateral 78 42.3 23.9 91.2 97.2 45.5 0.5 
Ipsilateral  60 −44.6 40.9 79.7 91.9 26.2 2.4 

GoF = Goodness of Fit statistics; Q = dipole strength; V = confidence volume; R = right hemisphere; L = left hemisphere.

Regarding the extent of cortical reorganization, these results agree with previous studies in amputees. Elbert et al. (1994, 1998) found that dipole locations showed an average shift of about 15 mm in upper limb amputees. Similarly, Knecht et al. (1996, 1998) reported shifts of lip representations between 0.1 and 38.6 mm after arm amputation.

With respect to the direction of the change in lip ECD location, we found a more lateral and posterior position of the lip representation contralateral to the replantation compared to the ipsilateral hemisphere. Most studies of upper limb amputees have shown an invasion of the SI hand area by the lip area, demonstrating a shift of the lip representation towards a more superior, medial, and posterior position (Elbert et al., 1994, 1997; Yang et al., 1994a, b; Flor et al., 1995; Knecht et al., 1996, 1998; Wiech et al., 2000). Results in amputees were explained by deafferentation of the hand area, i.e. a lack of input into the hand area, which is caused by the immediate rupture of nerve fibres during the traumatic amputation. Because of the missing input into the hand area, adjacent representation areas enlarge, thereby invading the territory that formerly constituted the hand area. Such cortical alterations can also be observed immediately after deafferentation (Birbaumer et al., 1997; Weiss et al., 2000, 2004). Hence, it can be assumed that these processes also took place in our patients within the first 4 to 8 hours after amputation before they underwent replantation (Hahn et al., 1998; Battiston et al., 2007; Hirche and Germann, 2010). Obviously, this difference with regard to the lip representations between replanted and amputated patients can only be explained by the reconstruction of the arm nerves and its consequences. Several consequences might influence the representations: restoration of hand nerves commonly results in a partial reinnervation of the replanted extremity. The partial restoration of afferent and efferent connections between the replanted extremity and S1 (and M1) will possibly not only change its own representation but also the representation of the somatotopically neighbouring structures, i.e. the lip representation. Future longitudinal studies in patients with macroreplantation including recordings of somatosensory fields and pain over time might help to assess the interplay between deafferentation, reinnervation, reorganization, and pain.

Regarding the ECD strengths assessed in SI after lip stimulation, no evidence for increased excitability was found. This contradicts several studies in upper limb amputees that have reported a global increase in cortical activity following deafferentation (Karl et al., 2004a, b; Flor et al., 2006). This difference to amputees might possibly be explained by the nature of deafferentation after macroreplantation. The deafferentation caused by macroreplantation is not static but changes depending on the outcome of replantation surgery. Thus, a successful replantation might lead to new input to the hand area as already mentioned above. With successful replantation cortical activity in the hand area increases, thereby gaining cortical resources and possibly resulting in a loss of cortical resources and reduced activity for the lip representation.

Reorganization of the thumb representation

We used Euclidean distance between the ECD localization for left D1 stimulation and the ECD localization for right D1 stimulation and compared the corresponding ECD strengths to assess changes in the formerly deafferented area. Firstly, dipole strength to D1 stimulation of the replanted hand was significantly larger than to the contralateral hand. Secondly, we found a profound difference in the ECD locations for D1 stimulations between the replanted and the uninjured hand. Finally, this difference in ECD location depended on pain.

Strength of the equivalent current dipole and correlation with pain

In the hemisphere contralateral to the replantation, the mean dipole strength of SI hand representation evoked by D1 stimulation exceeded the dipole strength on the hemisphere ipsilateral to the replantation by 40% on average. Clearly, stimulation of the replanted D1 provokes considerably higher ECD strength than stimulation of the non-affected thumb in spite of the presumably poorer afferent transmission in the replanted arm. The fact that a dipole could be identified after stimulation of the D1 of the replanted arm shows that a successful reinnervation has taken place. It supports the idea that the representation of the formerly deafferented body parts has been maintained or regained.

The higher ECD strength on the hemisphere contralateral to the replantation might be explained by several factors. One possibility is that the ECD strength is mediated by pain. This interpretation is supported by our finding that more pain correlates with smaller interhemispherical differences in ECD strengths (Table 6). In chronic patients with unilateral traumatic peripheral nerve injury, Theuvenet et al. (2011) found higher mean global field power to stimulation of residual nerves over the hemisphere contralateral to the injury. Moreover, these authors showed reduced mean global field power over the hemisphere ipsilateral to the injury when comparing mean global field power with a healthy control group. These results suggest that the drastically elevated ECD strengths in our patients might be explained by an increase of ECD strength contralateral to the replanted arm or a decrease of ECD over the hemisphere ipsilateral to the replantation or both, as a result of chronic pain. However, a direct relationship between ECD strength and pain intensity experienced in the replanted arm was not found, as in our sample of replantation patients the correlation between these two variables was not significant. Another possible interpretation might be that the activation of SI in the hemisphere contralateral to the replantation might still be changed. The afferent input to SI in this hemisphere is probably not completely restored. A proportion of SI neurons which still do not receive their normal input due to disturbed peripheral nerve growth might react to the already available input. This would result in higher ECD strength. In line with this interpretation, Sens et al. (2012) recently showed that temporary functional deafferentation by an anaesthetic cream leads to an increase of dipole strength of neighbouring representations. The elevated ECD strength can be interpreted as a gain in cortical resources for the processing of remaining input.

Location of the equivalent current dipole and correlation with pain

ECD locations of D1 representations showed a mean deviation of >10 mm. Descriptive analysis of the mean coordinates of D1 representations revealed that the dipole representing D1 in the hemisphere contralateral to the replantation was situated more medial, posterior, and superior than the D1 representation ipsilateral to the replantation (Table 5). Thus, our results point to a cortical reorganization within the hand area in SI. In that, we were able to show that reorganization does not only occur in the areas surrounding the cortex of the replanted representations, i.e. lip, in our patients, but also in the area representing the replanted arm itself or its homonym contralateral representation. One possible explanation for the differences in ECD location of D1 representations is the (only) partial reinnervation of the replanted extremity. This partial reinnervation might result in changes of the representations of single fingers in SI. Relatively more (or less) afferent input to the representation of a single finger compared to the neighbouring finger results in changes in the representation of fingers in SI (Weiss et al., 1998, 2000). Probably more important, the replantation itself does not allow a one-by-one single nerve fibre surgery leading to significant disturbances of the pattern of afferent input after replantation surgery. This might lead to a blurred representation of D1 within SI resulting in a difference of ECD location to D1 stimulation between hemispheres. Another interpretation of this difference in ECD locations is persistent pain. Persistent pain in patients with unilateral traumatic nerve injury has been shown to influence mean global field power and the distribution of magnetic fields probably resulting from different dipole locations (Theuvenet et al., 2011).

Table 5

Distribution parameters of the dipole coordinates for D1 representations (n = 10)

 Mean Median SD Minimum Maximum 
xrs 39.55 40.65 5.73 28.30 46.90 
xnrs 41.61 41.65 5.95 34.50 49.30 
yrs 23.32 23.25 4.99 13.90 32.90 
ynrs 25.30 25.15 8.28 13.40 40.90 
zrs 85.08 84.80 4.77 76.40 91.40 
znrs 82.21 82.06 6.33 74.20 92.90 
 Mean Median SD Minimum Maximum 
xrs 39.55 40.65 5.73 28.30 46.90 
xnrs 41.61 41.65 5.95 34.50 49.30 
yrs 23.32 23.25 4.99 13.90 32.90 
ynrs 25.30 25.15 8.28 13.40 40.90 
zrs 85.08 84.80 4.77 76.40 91.40 
znrs 82.21 82.06 6.33 74.20 92.90 

rs = side contralateral to replantation; nrs = side ipsilateral to replantation (not replanted).

Table 6

Partial correlations between average pain score and measures of cortical reorganization in the thumb area controlling for a third factor

Factor ED × Factor
 
ΔQ(F) × Factor
 
APS × Factor
 
ED × APS | Factor
 
ΔQ(F) × APS | Factor
 
P R2 P R2 P R2 P R2 P R2 
Δ Time −0.079 0.414 0.00 0.560 0.092 0.31 0.346 0.181 0.12 −0.625* 0.049 0.39 −0.808* 0.015 0.65 
Age −0.215 0.276 0.05 0.057 0.876 0.00 0.556 0.060 0.31 −0.607 0.055 0.37 −0.618 0.102 0.38 
Replantation level 0.076 0.417 0.00 −0.030 0.935 0.00 0.596* 0.045 0.36 −0.822* 0.006 0.68 −0.572 0.139 0.33 
Kind of replantation 0.105 0.387 0.01 −0.238 0.429 0.06 −0.456 0.109 0.21 −0.638* 0.044 0.41 −0.660 0.075 0.44 
Disrupted arm nerves 0.364 0.151 0.13 0.686* 0.029 0.47 −0.463 0.105 0.21 −0.538 0.085 0.29 −0.183 0.665 0.03 
Factor ED × Factor
 
ΔQ(F) × Factor
 
APS × Factor
 
ED × APS | Factor
 
ΔQ(F) × APS | Factor
 
P R2 P R2 P R2 P R2 P R2 
Δ Time −0.079 0.414 0.00 0.560 0.092 0.31 0.346 0.181 0.12 −0.625* 0.049 0.39 −0.808* 0.015 0.65 
Age −0.215 0.276 0.05 0.057 0.876 0.00 0.556 0.060 0.31 −0.607 0.055 0.37 −0.618 0.102 0.38 
Replantation level 0.076 0.417 0.00 −0.030 0.935 0.00 0.596* 0.045 0.36 −0.822* 0.006 0.68 −0.572 0.139 0.33 
Kind of replantation 0.105 0.387 0.01 −0.238 0.429 0.06 −0.456 0.109 0.21 −0.638* 0.044 0.41 −0.660 0.075 0.44 
Disrupted arm nerves 0.364 0.151 0.13 0.686* 0.029 0.47 −0.463 0.105 0.21 −0.538 0.085 0.29 −0.183 0.665 0.03 

All the P-values presented in the table are two-tailed correlation values. APS = average pain score; ED = Euclidean distance between the ECD localization for left D1 stimulation and the ECD localization for right D1 stimulation; ΔQ(F) = difference in ECD strength between the hemispheres; ED × Factor: correlation between ED and Factor; ΔQ(F) × Factor: correlation between ΔQ(F) and Factor; APS × Factor: correlation between APS and Factor; ED × APS | Factor: first-order partial correlations between ED and APS in dependence of Factor; ΔQ(F) × APS | Factor: first-order partial correlations between ΔQ(F) and APS in dependence of Factor; Δ Time = time since replantation; Replantation level: forearm (= 1), elbow (= 2), upper arm (= 3); Kind of replantation: Subtotal (= 1), total (= 2); Disrupted arm nerves: N. medianus, N. radialis and N. ulnaris were disrupted (= 1), two of those three arm nerves disrupted (= 2), one of those three arm nerves disrupted (= 3).

Importantly, the difference in ECD locations to D1 stimulation significantly depends on pain intensity, as shown by the significant correlation between these parameters. The difference in ECD location was smaller the stronger the persistent pain in the replanted extremity was experienced. This result might be explained in the following way: there coexist at least two complementary processes; processes driven by the partial deinnervation (or disturbed reinnervation) and processes driven by persistent pain. Deafferentation is known to cause shrinkage of the deafferented representation that is accompanied by an invasion of neighbouring representations. If deafferentation would be the leading factor for changes in ECD locations and for pain, then one should expect a positive correlation between pain experience and differences in ECD locations. Persistent pain experience seems to maintain the representation, according to recent data by Makin et al. (2013). If persistent pain maintained the representation, then the stimulation of D1 in the replanted hand would result in activation at the same location where the representation was located before the event of amputation, i.e. persistent pain experience might maintain the representation at the location where it used to be represented before amputation and replantation. Thus, the changes in ECD to D1 stimulation of the replanted hand are smaller the stronger pain is experienced. Thus, this more stable representation can be interpreted as a kind of ‘use-(pain) dependent’ reorganization of the representation in SI in line with the persistent representation model (Makin et al., 2013).

Another possible explanation of the negative association between the changes in hand representation and experienced pain in the replanted extremity might be the coexistence of somatosensory and nociceptive maps as suggested by Mancini et al. (2012). These maps might be highly collocated and aligned with each other, making it difficult to distinguish between them with the methods used in our study. According to this explanation, the negative association between pain and hand reorganization would therefore reflect changes, not in the somatosensory map, but in the hand area of the nociceptive map.

Additional factors influencing the representation in the primary somatosensory cortex

A striking difference between amputees and patients with replanted upper extremities is the rehabilitation of motor skills of the previously amputated body part as reinnervation gradually occurs. Even though fine motor actions cannot be completely restored, physiotherapeutic interventions as well as the increasing use of the replanted limb in everyday life might also lead to a re-reorganization within the somatosensory cortices. The observed difference between D1 representations might indicate use-dependent plasticity during which D1 gained further cortical resources resulting in an enlarged D1 representation area of SI. The reduced sensibility of the replanted hand thereby recovers over time accompanied by improved motor skills. The idea of a gain in cortical resources of the D1 representation area is further supported by the fact that descriptive analyses of the coordinates showed that the ECD location to lip stimulation on the replanted body side was situated more inferior relative to the other side. This is a possible indication that the D1 representation area might have enlarged not only towards the remaining hand area but also towards the lip area. Map expansion has been shown to be associated with a gain in performance in animals (Jenkins et al., 1990; Recanzone et al., 1992a, b, c) and in humans (Elbert et al., 1995; Sterr et al., 1998a, b; Hashimoto et al., 2004). The possible enlargement of the D1 area in SI thus might reflect compensation and an improved functionality of the hand (Kalisch et al., 2009; Reuter et al., 2012).

Limitations and future directions

Several limitations of the present study warrant discussion. One important limitation is the lack of a healthy control group. The hemisphere ipsilateral to the replantation served as comparison as it is considered to be not directly changed by the deafferentation and because variation within subjects is considered minimal. We referred to findings from previous studies in healthy subjects that have reported variations of 3 to 6 mm in the corresponding representation in the SI of each hemisphere (Gallen et al., 1993). However, there are hints that point to an involvement of somatosensory representations in SI both ispi- and contralateral to chronic pain after unilateral nerve injury (Theuvenet et al., 2011). Therefore, future studies should consider a control group to analyse changes of ECD location and strength in each of the hemispheres separately and to assess not only relative changes as reported here.

Our sample size is sufficiently large for the analyses; it is one of the largest studies analysed to date with respect to cortical reorganization in patients after macroreplantation. Nevertheless, the number of patients enrolled is still too small to account for the heterogeneity of injuries that can precede macroreplantations. A larger sample would be necessary to investigate differences in reorganization that depend on the level of amputation. The more proximal the amputation has occurred, the longer the reinnervation takes and the greater the risk for distorted nerve growth. A larger sample size would also allow for more reliable results concerning correlations between pain and the difference in ECD locations to D1 stimulation taking into account the borderline significance (P = 0.04–0.05). Another limitation is that the time window between replantation and examination is quite varied. Future studies would benefit from a larger cohort evaluated in a longitudinal study so that the time course of reorganization processes suggested here might be characterized. In such a study it might be possible to assess whether maladaptive plasticity triggered by deafferentation and pain-induced plasticity are independent processes. Other aspects such as training-dependent plasticity during rehabilitation programs should also be considered as additional influencing factors.

Conclusion

In summary, we found cortical reorganization within the SI of patients who underwent immediate replantation of a traumatically amputated upper limb. This reorganization seems to be mediated by at least two complementary processes: (i) reorganization of representations neighbouring the representation of the replanted extremity; and (ii) reorganization of the representation of the replanted extremity, which correlates negatively with the pain still experienced in the replanted limb.

Acknowledgements

Thanks to Dr. Sandra Preißler and Marcel Franz for their contribution in data acquisition and analyses. We also thank Dr. Jeremy Thorne for helpful discussion of the data and language advice and the reviewers for their extremely helpful comments.

Funding

This study was supported in part by German Social Accident Insurance (DGUV FR196). There are no conflicts of interest for any of the authors.

Supplementary material

Supplementary material is available at Brain online.

Abbreviations

    Abbreviations
  • D1

    thumb

  • ECD

    equivalent current dipole

  • MEG

    magnetoencephalography

  • SI

    primary somatosensory cortex

References

Battiston
B
Tos
P
Clemente
A
Pontini
I
Actualities in big segments replantation surgery
J Plast Reconstr Aesthet Surg
 , 
2007
, vol. 
60
 (pg. 
849
-
55
)
Birbaumer
N
Lutzenberger
W
Montoya
P
Larbig
W
Unertl
K
Töpfner
S
, et al.  . 
Effects of regional anesthesia on phantom limb pain are mirrored in changes in cortical reorganization
J Neurosci
 , 
1997
, vol. 
17
 (pg. 
5503
-
8
)
Björkman
A
Waites
A
Rosén
B
Larsson
EM
Lundborg
G
Cortical reintegration of a replanted hand and an osseointegrated thumb prosthesis
Acta Neurochir Suppl
 , 
2007a
, vol. 
100
 (pg. 
109
-
12
)
Björkman
A
Waites
A
Rosén
B
Lundborg
G
Larsson
EM
Cortical sensory and motor response in a patient whose hand has been replanted: one-year follow up with functional magnetic resonance imaging
Scand J Plast Reconstr Surg Hand Surg
 , 
2007b
, vol. 
41
 (pg. 
70
-
6
)
Brenneis
C
Löscher
WN
Egger
KE
Benke
T
Schocke
M
Gabl
MF
, et al.  . 
Cortical motor activation patterns following hand transplantation and replantation
J Hand Surg Br
 , 
2005
, vol. 
30
 (pg. 
530
-
3
)
Buchner
H
Reinartz
U
Waberski
TD
Gobbelé
R
Noppeney
U
Scherg
M
Sustained attention modulates the immediate effect of de-afferentation on the cortical representation of the digits: source localization of somatosensory evoked potentials in humans
Neurosci Lett
 , 
1999
, vol. 
260
 (pg. 
57
-
60
)
Candia
V
Wienbruch
C
Elbert
T
Rockstroh
B
Ray
W
Effective behavioral treatment of focal hand dystonia in musicians alters somatosensory cortical organization
Proc Natl Acad Sci USA
 , 
2003
, vol. 
100
 (pg. 
7942
-
6
)
Darnall
BD
Ephraim
P
Wegener
ST
Dillingham
T
Pezzin
L
Rossbach
P
, et al.  . 
Depressive symptoms and mental health service utilization among persons with limb loss: results of a national survey
Arch Phys Med Rehab
 , 
2005
, vol. 
86
 (pg. 
650
-
8
)
Desmond
DM
Coping, affective distress, and psychosocial adjustment among people with traumatic upper limb amputations
J Psychosom Res
 , 
2007
, vol. 
62
 (pg. 
15
-
21
)
Eickhoff
SB
Dafotakis
M
Grefkes
C
Shah
NJ
Zilles
K
Piza-Katzer
H
Central adaptation following heterotopic hand replantation probed by fMRI and effective connectivity analysis
Exp Neurol
 , 
2008
, vol. 
212
 (pg. 
132
-
44
)
Elbert
T
Candia
V
Altenmüller
E
Rau
H
Sterr
A
Rockstroh
B
, et al.  . 
Alteration of digital representations in somatosensory cortex in focal hand dystonia
Neuroreport
 , 
1998
, vol. 
9
 (pg. 
3571
-
5
)
Elbert
T
Flor
H
Birbaumer
N
Knecht
S
Hampson
S
Larbig
W
, et al.  . 
Extensive reorganization of the somatosensory cortex in adult humans after nervous system injury
Neuroreport
 , 
1994
, vol. 
5
 (pg. 
2593
-
7
)
Elbert
T
Pantev
C
Wienbruch
C
Rockstroh
B
Taub
E
Increased cortical representation of the fingers of the left hand in string players
Science
 , 
1995
, vol. 
270
 (pg. 
305
-
95
)
Elbert
T
Sterr
A
Flor
H
Rockstroh
B
Knecht
S
Pantev
C
, et al.  . 
Input-increase and input-decrease types of cortical reorganization after upper extremity amputation in humans
Exp Brain Res
 , 
1997
, vol. 
117
 (pg. 
161
-
4
)
Faggin
BM
Nguyen
KT
Nicolelis
MA
Immediate and simultaneous sensory reorganization at cortical and subcortical levels of the somatosensory system
Proc Natl Acad Sci USA
 , 
1997
, vol. 
94
 (pg. 
9428
-
33
)
Flor
H
Phantom-limb pain: characteristics, causes, and treatment
Lancet Neurol
 , 
2002
, vol. 
1
 (pg. 
182
-
9
)
Flor
H
Elbert
T
Knecht
S
Wienbruch
C
Pantev
C
Birbaumer
N
, et al.  . 
Phantom-limb pain as a perceptual correlate of cortical reorganization following arm amputation
Nature
 , 
1995
, vol. 
375
 (pg. 
482
-
4
)
Flor
H
Elbert
T
Mühlnickel
W
Pantev
C
Wienbruch
C
Taub
E
Cortical reorganization and phantom phenomena in congenital and traumatic upper-extremity amputees
Exp Brain Res
 , 
1998
, vol. 
119
 (pg. 
205
-
12
)
Flor
H
Nikolajsen
L
Jensen
TS
Phantom limb pain: a case of maladaptive CNS plasticity?
Nat Rev Neurosci
 , 
2006
, vol. 
7
 (pg. 
873
-
81
)
Florence
SL
Kaas
JH
Large-scale reorganization at multiple levels of the somatosensory pathway follows therapeutic amputation of the hand in monkeys
J Neurosci
 , 
1995
, vol. 
15
 (pg. 
8083
-
95
)
Gallen
CC
Sobel
DF
Lewine
JD
Sanders
JA
Hart
BL
Davis
LE
, et al.  . 
Neuromagnetic mapping of brain function
Radiology
 , 
1993
, vol. 
87
 (pg. 
863
-
7
)
Gotz
T
Huonker
R
Miltner
WH
Witte
OW
Dettner
K
Weiss
T
Task requirements change signal strength of the primary somatosensory M50: Oddball vs. one-back tasks
Psychophysiology
 , 
2011
, vol. 
48
 (pg. 
569
-
77
)
Graham
B
Adkins
P
Tsai
TM
Firrell
J
Breidenbach
WC
Major replantation versus revision amputation and prosthetic fitting in the upper extremity: a late functional outcomes study
J Hand Surg
 , 
1998
, vol. 
23
 (pg. 
783
-
91
)
Grüsser
SM
Mühlnickel
W
Schaefer
M
Villringer
K
Christmann
C
Koeppe
C
, et al.  . 
Remote activation of referred phantom sensation and cortical reorganization in human upper extremity amputees
Exp Brain Res
 , 
2004
, vol. 
154
 (pg. 
97
-
102
)
Grüsser
SM
Winter
C
Mühlnickel
W
Denke
C
Karl
A
Villringer
K
, et al.  . 
The relationship of perceptual phenomena and cortical reorganization in upper extremity amputees
Neuroscience
 , 
2001
, vol. 
102
 (pg. 
263
-
72
)
Hahn
P
Frank
U
Genz
G
Lanz
U
Replantation. Indications and organization
Orthopade
 , 
1998
, vol. 
27
 (pg. 
414
-
21
)
Hashimoto
I
Suzuki
A
Kimura
T
Iguchi
Y
Tanosaki
M
Takino
R
, et al.  . 
Is there training-dependent reorganization of digit representations in area 3b of string players?
Clin Neurophysiol
 , 
2004
, vol. 
115
 (pg. 
435
-
47
)
Hierner
R
Berger
A
Long-term results after total and subtotal macroamputations of the upper limb
Eur J Plast Surg
 , 
2005
, vol. 
28
 (pg. 
119
-
30
)
Hirche
C
Germann
G
Makroreplantationen - eine interdisziplinäre Herausforderung
Orthopädie und Unfallchirurgie up2date
 , 
2010
, vol. 
5
 (pg. 
57
-
72
)
Ipsen
T
Lundkvist
L
Barfred
T
Pless
J
Principles of evaluation and results in microsurgical treatment of major limb amputations. A follow-up study of 26 consecutive cases 1978-1987
Scand J Plast Reconstr Surg Hand Surg
 , 
1990
, vol. 
24
 (pg. 
75
-
80
)
Jenkins
WM
Merzenich
MM
Ochs
MT
Allard
T
Guíc-Robles
E
Functional reorganization of primary somatosensory cortex in adult owl monkeys after behaviorally controlled tactile stimulation
J Neurophysiol
 , 
1990
, vol. 
63
 (pg. 
82
-
104
)
Kalisch
T
Ragert
P
Schwenkreis
P
Dinse
HR
Tegenthoff
M
Impaired tactile acuity in old age is accompanied by enlarged hand representations in somatosensory cortex
Cereb Cortex
 , 
2009
, vol. 
19
 (pg. 
1530
-
8
)
Karl
A
Birbaumer
N
Lutzenberger
W
Cohen
LG
Flor
H
Reorganization of motor and somatosensory cortex in upper extremity amputees with phantom limb pain
J Neurosci
 , 
2001
, vol. 
21
 (pg. 
3609
-
18
)
Karl
A
Diers
M
Flor
H
P300-amplitudes in upper limb amputees with and without phantom limb pain in a visual oddball paradigm
Pain
 , 
2004a
, vol. 
110
 (pg. 
40
-
8
)
Karl
A
Mühlnickel
W
Kurth
R
Flor
H
Neuroelectric source imaging of steady-state movement-related cortical potentials in human upper extremity amputees with and without phantom limb pain
Pain
 , 
2004b
, vol. 
110
 (pg. 
90
-
102
)
Knecht
S
Henningsen
H
Elbert
T
Flor
H
Hohling
C
Pantev
C
, et al.  . 
Reorganizational and perceptional changes after amputation
Brain
 , 
1996
, vol. 
119
 (pg. 
1213
-
9
)
Knecht
S
Henningsen
H
Höhling
C
Elbert
T
Flor
H
Pantev
C
, et al.  . 
Plasticity of plasticity? Changes in the pattern of perceptual correlates of reorganization after amputation
Brain
 , 
1998
, vol. 
121
 (pg. 
717
-
24
)
Kooijman
CM
Dijkstra
PU
Geertzen
JHB
Elzinga
A
van der Schans
CP
Phantom pain and phantom sensations in upper limb amputees: an epidemiological study
Pain
 , 
2000
, vol. 
87
 (pg. 
33
-
41
)
Langer
N
Hanggi
J
Muller
NA
Simmen
HP
Jancke
L
Effects of limb immobilization on brain plasticity
Neurology
 , 
2012
, vol. 
78
 (pg. 
182
-
8
)
Lissek
S
Wilimzig
C
Stude
P
Pleger
B
Kalisch
T
Maier
C
, et al.  . 
Immobilization impairs tactile perception and shrinks somatosensory cortical maps
Curr Biol
 , 
2009
, vol. 
19
 (pg. 
837
-
42
)
Lotze
M
Grodd
W
Birbaumer
N
Erb
M
Huse
E
Flor
H
Does use of a myoelectric prosthesis prevent cortical reorganization and phantom limb pain?
Nat Neurosci
 , 
1999
, vol. 
2
 (pg. 
501
-
2
)
Lundborg
G
Richard
P
Bunge memorial lecture. Nerve injury and repair—a challenge to the plastic brain
J Peripher Nerv Syst
 , 
2003
, vol. 
8
 (pg. 
209
-
26
)
Makin
TR
Scholz
J
Filippini
N
Henderson Slater
D
Tracey
I
Johansen-Berg
H
Phantom pain is associated with preserved structure and function in the former hand area
Nat Commun
 , 
2013
, vol. 
4
 pg. 
1570
 
Mancini
F
Haggard
P
Iannetti
GD
Longo
MR
Sereno
MI
Fine-grained nociceptive maps in primary somatosensory cortex
J Neurosci
 , 
2012
, vol. 
32
 (pg. 
17155
-
62
)
Merzenich
MM
Nelson
RJ
Stryker
MP
Cynader
MS
Schoppmann
A
Zook
JM
Somatosensory cortical map changes following digit amputation in adult monkeys
J Comp Neurol
 , 
1984
, vol. 
224
 (pg. 
591
-
605
)
Piza-Katzer
H
Brenneis
C
Loscher
WN
Benke
T
Schocke
M
Gabl
MF
, et al.  . 
Cortical motor activation patterns following hand transplantation and replantation
Acta Neurochir Suppl
 , 
2007
, vol. 
100
 (pg. 
113
-
5
)
Pons
TP
Garraghty
PE
Ommaya
AK
Kaas
JH
Taub
E
Mishkin
M
Massive cortical reorganization after sensory deafferentation in adult macaques
Science
 , 
1991
, vol. 
252
 (pg. 
1857
-
60
)
Preißler
S
Feiler
J
Dietrich
C
Hofmann
GO
Miltner
WH
Weiss
T
Gray matter changes following limb amputation with high and low intensities of phantom limb pain
Cereb Cortex
 , 
2013
, vol. 
23
 (pg. 
1038
-
48
)
Recanzone
GH
Jenkins
WM
Hradek
GT
Merzenich
MM
Progressive improvement in discriminative abilities in adult owl monkeys performing a tactile frequency discrimination task
J Neurophysiol
 , 
1992a
, vol. 
67
 (pg. 
1015
-
30
)
Recanzone
GH
Merzenich
MM
Jenkins
WM
Frequency discrimination training engaging a restricted skin surface results in an emergence of a cutaneous response zone in cortical area 3a
J Neurophysiol
 , 
1992b
, vol. 
67
 (pg. 
1057
-
70
)
Recanzone
GH
Merzenich
MM
Jenkins
WM
Grajski
KA
Dinse
HR
Topographic reorganization of the hand representation in cortical area 3b owl monkeys trained in a frequency-discrimination task
J Neurophysiol
 , 
1992c
, vol. 
67
 (pg. 
1031
-
56
)
Reuter
EM
Voelcker-Rehage
C
Vieluf
S
Godde
B
Touch perception throughout working life: effects of age and expertise
Exp Brain Res
 , 
2012
, vol. 
216
 (pg. 
287
-
97
)
Röricht
S
Machetanz
J
Irlbacher
K
Niehaus
L
Biemer
E
Meyer
BU
Reorganization of human motor cortex after hand replantation
Ann Neurol
 , 
2001
, vol. 
50
 (pg. 
240
-
9
)
Rossini
PM
Martino
G
Narici
L
Pasquarelli
A
Peresson
M
Pizzella
V
, et al.  . 
Short-term brain ‘plasticity' in humans: transient finger representation changes in sensory cortex somatotopy following ischemic anesthesia
Brain Res
 , 
1994
, vol. 
642
 (pg. 
169
-
77
)
Schaefer
M
Mühlnickel
W
Grüsser
SM
Flor
H
Reproducibility and stability of neuroelectric source imaging in primary somatosensory cortex
Brain Topogr
 , 
2002
, vol. 
14
 (pg. 
179
-
89
)
Schaefer
M
Nöennig
N
Karl
A
Heinze
HJ
Rotte
M
Reproducibility and stability of neuromagnetic source imaging in primary somatosensory cortex
Brain Topogr
 , 
2004
, vol. 
17
 (pg. 
47
-
53
)
Sens
E
Teschner
U
Meissner
W
Preul
C
Huonker
R
Witte
OW
, et al.  . 
Effects of temporary functional deafferentation on the brain, sensation, and behavior of stroke patients
J Neurosci
 , 
2012
, vol. 
32
 (pg. 
11773
-
9
)
Sterr
A
Müller
MM
Elbert
T
Rockstroh
B
Pantev
C
Taub
E
Changed perceptions in Braille readers
Nature
 , 
1998a
, vol. 
391
 (pg. 
134
-
5
)
Sterr
A
Müller
MM
Elbert
T
Rockstroh
B
Pantev
C
Taub
E
Perceptual correlates of changes in cortical representation of fingers in blind multifinger Braille readers
J Neurosci
 , 
1998b
, vol. 
18
 (pg. 
4417
-
23
)
Taulu
S
Simola
J
Kajola
M
Applications of the signal space separation method
IEEE Trans Signal Process
 , 
2005
, vol. 
53
 (pg. 
3359
-
72
)
Theuvenet
PJ
de Munck
JC
Peters
MJ
van Ree
JM
Lopes da Silva
FL
Chen
AC
Anesthetic block of pain-related cortical activity in patients with peripheral nerve injury measured by magnetoencephalography
Anesthesiology
 , 
2011
, vol. 
115
 (pg. 
375
-
86
)
Weiss
T
Miltner
WH
Adler
T
Bruckner
L
Taub
E
Decrease in phantom limb pain associated with prosthesis-induced increased use of an amputation stump in humans
Neurosci Lett
 , 
1999
, vol. 
272
 (pg. 
131
-
4
)
Weiss
T
Miltner
WH
Dillmann
J
Meissner
W
Huonker
R
Nowak
H
Reorganization of the somatosensory cortex after amputation of the index finger
Neuroreport
 , 
1998
, vol. 
9
 (pg. 
213
-
6
)
Weiss
T
Miltner
WH
Huonker
R
Friedel
R
Schmidt
I
Taub
E
Rapid functional plasticity of the somatosensory cortex after finger amputation
Exp Brain Res
 , 
2000
, vol. 
134
 (pg. 
199
-
203
)
Weiss
T
Miltner
WHR
Liepert
J
Meissner
W
Taub
E
Rapid functional plasticity in the primary somatomotor cortex and perceptual changes after nerve block
Eur J Neurosci
 , 
2004
, vol. 
20
 (pg. 
3413
-
23
)
Wiech
K
Preissl
H
Birbaumer
N
Neuroimaging of chronic pain: phantom limb and musculoskeletal pain
Scand J Rheumatol
 , 
2000
, vol. 
113
 
Suppl
(pg. 
13
-
8
)
Yang
TT
Gallen
C
Schwartz
B
Bloom
FE
Ramachandran
VS
Cobb
S
Sensory maps in the human brain
Nature
 , 
1994a
, vol. 
368
 (pg. 
592
-
3
)
Yang
TT
Gallen
CC
Ramachandran
VS
Cobb
S
Schwartz
BJ
Bloom
FE
Noninvasive detection of cerebral plasticity in adult human somatosensory cortex
Neuroreport
 , 
1994b
, vol. 
5
 (pg. 
701
-
4
)