Journal of the Neurological Sciences 292 (2010) 63–71

Contents lists available at ScienceDirect

Journal of the Neurological Sciences

journal homepage: www.elsevier.com/locate/jns

Nerve excitability changes after intravenous immunoglobulin infusions in multifocal motor neuropathy and chronic inflammatory demyelinating neuropathy

Delphine Boërio a,b,c, Alain Créange b,d, Jean-Yves Hogrel c, Antoine Guéguen b,d, Dominique Bertrand e, Jean-Pascal Lefaucheur a,b,⁎ a Service de Physiologie, Explorations Fonctionnelles, Hôpital Henri Mondor, Assistance Publique, Hôpitaux de Paris, Créteil, France b EA 4391, Excitabilité Nerveuse et Thérapeutique, Université Paris 12, Créteil, France c Institut de Myologie, Hôpital Pitié-Salpêtrière, Paris, France d Service de Neurologie, Hôpital Henri Mondor, Assistance Publique, Hôpitaux de Paris, Créteil, France e Unité de Sensométrie et de Chimiométrie, ENITIAA-INRA, Nantes, France article info abstract

Article history: Intravenous immunoglobulin (IVIg) infusions may provide clinical benefits in multifocal motor neuropathy Received 18 July 2009 (MMN) and chronic inflammatory demyelinating polyneuropathy (CIDP). The short delay in the clinical Received in revised form 22 December 2009 response to IVIg therapy is not consistent with a process of remyelination or axonal regeneration. We Accepted 2 February 2010 assessed whether or not the efficacy of IVIg infusions in MMN and CIDP could reflect changes in axonal Available online 10 March 2010 membrane properties and nerve excitability. Ulnar motor nerve excitability was studied before and after three to five consecutive days of IVIg infusions (0.4 g/kg/day) in 10 patients with MMN, 10 patients with Keywords: – – Chronic inflammatory demyelinating CIDP, and 10 neurological controls (CTRLs). Excitability recovery cycle, stimulus response and strength polyneuropathy duration properties were investigated. The recovery cycle parameters (absolute and relative refractory Conduction block period durations, refractoriness and supernormality) were similar in all groups and did not change after IVIg Intravenous immunoglobulin infusions. At baseline, patients with CIDP, but not with MMN, showed a reduced strength–duration time Multifocal motor neuropathy constant (chronaxie) and increased rheobase when compared to CTRLs. After IVIg infusions, strength– Rheobase duration time constant remained stable in CTRLs, but decreased in patients with MMN or CIDP. Rheobase Sodium conductance increased in the three groups after treatment. The decreased strength–duration time constant after IVIg – Strength duration time constant infusions in patients with MMN or CIDP could reflect a reduction of persistent Na+ current, able to limit Supernormality intraaxonal Na+ accumulation and then to produce neuroprotective effects. However, this could also reflect compensatory mechanisms that did not directly underlie the therapeutic effect. Whatever the underlying process, this result revealed that IVIgs were able to produce early nerve excitability changes. © 2010 Elsevier B.V. All rights reserved.

1. Introduction lasts a few weeks, leading to perform periodic infusions to maintain therapeutic benefit [6–11]. Multifocal motor neuropathy (MMN) and chronic inflammatory Various mechanisms of action have been put forward to explain demyelinating polyneuropathy (CIDP) are acquired demyelinating the efficacy of IVIgs, including neutralization of auto-antibodies, peripheral neuropathies characterized by dysimmune pathogenesis, down-regulation of inflammatory mediators (tumor necrosis factor various clinical presentations including relapses and remissions, and alpha, interleukin 1, matrix metalloproteinases 2 and 9, complement secondary axonal loss [1,2]. Treatment by intravenous immunoglobu- fractions, and chemokines), inhibition of lymphocyte proliferation, lins, (IVIgs) can provide significant clinical benefits in more than 60% of activity and antigenic recognition, and blockade or modulation of the patients with CIDP [3–5] and almost 80% of patients with MMN [6–8]. Fcγ RII/Fcγ RIII ratio on macrophages [12–14]. Whatever the involved Clinical improvement occurs a few days after IVIg infusions, but usually mechanisms, the rapid clinical improvement induced by IVIgs is not consistent with a process of remyelination or axonal regeneration that should take several weeks. Conversely, this might reflect functional changes in axonal membrane properties and nerve excitability. Specific electrophysiological methods can appraise such changes better than conventional nerve conduction studies [15–17].These ⁎ Corresponding author. Service Physiologie, Explorations Fonctionnelles, Hôpital methods have been used to characterize nerve excitability changes in a Henri Mondor, 51 avenue de Lattre de Tassigny, 94010 Créteil cedex, France. Tel.: +33 1 4981 2694; fax: +33 1 4981 4660. few studies on dysimmune neuropathies (review: [18]). In particular, E-mail address: [email protected] (J.-P. Lefaucheur). the influence of IVIg infusions was studied only in rare cases of MMN

0022-510X/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2010.02.002 64 D. Boërio et al. / Journal of the Neurological Sciences 292 (2010) 63–71

Table 1 score) and the specific MRC score for distal ulnar nerve territory (uln- Clinical data and muscle strength graded on Medical Research Council (MRC) score MRC score) were retained for analysis. before intravenous immunoglobulin (IVIg) infusions and increase in total MRC score after IVIg infusions. 2.2. Nerve excitability studies Patient Gender Age Disease Ulnar Total Number of Increase in number (years) duration MRC MRC days of total MRC A Keypoint EMG machine and pre-gelled disposable surface (years) score score IVIg score after IVIg infusions infusions electrodes (Medtronic Functional Diagnostics, Skovlunde, Denmark) were used for electrophysiological testing. The right ulnar nerve was Multifocal motor neuropathy 1 M 67 4 5 85 3 0 stimulated at the elbow and the wrist. None of our patients had a 2 M 58 14 3 94 3 0 conduction block on this nerve segment (forearm). The compound 3 M 53 3 4 108 4 2 muscle action potential (CMAP) elicited by ulnar nerve stimulation was 4 W 30 7 4 89 5 7 recorded from the abductor digiti minimi (ADM) muscle, with a muscle 5 M 54 8 3 99 3 1 – fi – 6 W 58 4 5 106 5 4 belly tendon montage. The signal was ltered through a 20 2000 Hz 7 M 70 10 5 98 4 7 bandpass filter. Room and skin temperatures were controlled. Skin 8 M 74 33 3 94 3 0 temperature near the stimulating site was maintained between 31 °C 9 M 84 20 5 98 3 0 and 33 °C during all the recordings. The temperature range was similar 10 M 68 10 5 107 3 0 in all groups. Chronic inflammatory demyelinating neuropathy First, the double collision method [22] was used to assess the 1 W 57 2 3 76 3 12 absolute refractory period (ARP). The minimal intensity of stimulation 2 M 81 5 3 89 3 13 required to obtain CMAP of maximal amplitude (Mmax) was 3 M 73 2 5 110 3 0 determined at each stimulation site (elbow and wrist). Then, paired 4 W 60 1 5 90 5 0 pulses were delivered at supra-maximal intensity (120–130% of 5 M 59 1 4 92 5 14 6 M 46 1 3 88 5 4 Mmax threshold intensity) simultaneously at the wrist and the elbow. 7 W 79 1 1 69 3 2 The first wrist pulse produced a Mmax in the ADM muscle, while the 8 W 81 1 5 104 5 0 descending volley from the first elbow pulse collided with the 9 W 21 1 5 108 3 0 antidromic volley from the wrist and did not induce any response. 10 W 44 1 5 95 4 3 Inter-stimuli interval (ISI) varied between 0.02 and 5 ms at the wrist and was constant (4 ms) at the elbow. When the second wrist pulse [19,20]. The present study assessed the ability of IVIg infusions to was delivered during the ARP, the descending volley from the second produce nerve excitability changes in patients with MMN or CIDP. elbow pulse was not collided and was able to produce a response. As nerve fibers recovered from ARP at the wrist, the descending volley from the second elbow pulse was reduced by collision. The minimal 2. Materials and methods ARP duration (ARPmin) was defined as the minimal ISI at the wrist providing the first amplitude reduction of the CMAP elicited by the 2.1. Patients and study design second elbow pulse. The maximal ARP duration (ARPmax) was determined as the minimal ISI at the wrist providing a total Thirty patients were studied. Ten patients (two women and eight cancellation of the CMAP elicited by the second elbow pulse. men, mean age (SD): 61.6 (14.7) years) fulfilled the required diagnostic To study excitability recovery cycle, paired stimuli were delivered criteria for MMN [2]. Ten patients (six women and four men, 60.1 (19.4) at the wrist. Paired stimuli consisted of a first pulse of supra-maximal years) fulfilled the required diagnostic criteria for CIDP [1]. Patient data intensity (120–130% of Mmax threshold) followed by a second pulse are summarized in Table 1. Ten patients (seven women and three men, of submaximal intensity (required to elicit CMAP (test response) with 49.9 (20.2) years) treated by IVIg infusions, but free from motor nerve an amplitude equal to 70±5% of Mmax). The ISI was increased from 2 disorders at the upper limbs, served as neurological controls (CTRL to 5 ms by 0.5 ms steps, and then from 5 to 7 ms by 1 ms steps. The group). They presented sensory neuronopathy (n=4), limb-girdle amplitude of the response to the second pulse (conditioned test myositis (n=4), or lower limb poliomyelitis (n=2). Conduction response) was obtained by subtracting the response produced by the studies and concentric needle electromyography did not reveal any first supra-maximal pulse given alone from the global response to motor nerve alteration in the upper limbs of these patients. Patients paired stimuli. The amplitude of the conditioned test response was gave their informed consent to this study that was approved by local normalized with respect to the amplitude of the unconditioned test ethical committee. The study also included an age-matched group of 10 response and plotted against ISI [23]. healthy subjects (three women and seven men, 49.0 (8.5) years) (NML The maximal duration of the relative refractory period (RRP) was group). defined by the ISI at which the amplitude of the conditioned test A minority of patients were taking drugs that could affect nerve response was equal to that of the unconditioned test response. In excitability, mainly serotonin reuptake inhibitors or tricyclic antide- addition, the percentages of refractoriness and supernormality were pressants (two patients in each group) and benzodiazepins (one patient calculated at ISIs of 2 and 7 ms, respectively [24]. Supernormality was in MMN and CTRL groups and three patients in CIDP group). All patients defined as the amplitude ratio of the conditioned to the unconditioned were naïve regarding antidepressants or anticonvulsant drugs. test responses. Refractoriness was expressed as 100% minus this ratio. Clinical and electrophysiological evaluations were performed Stimulus–response properties were studied by measuring the within a few hours before and after three to five consecutive days of threshold intensities required to obtain a CMAP of target amplitudes

IVIg infusions (0.4 g/kg/day, Tégéline®, LFB, France). These infusions (10% (i10)and90%(i90) of Mmax). A stimulus–response curve ratio (SRC fi were not the rst ones administered in our patients. However, in all ratio) was calculated for two different pulse durations (0.2 and 1 ms), cases, the preceding series of IVIg infusions were performed more ½i90−i10 according to the formula first introduced by Brismar [25]. than 4 weeks before inclusion in this study. Clinical assessment was i10 based on muscle strength measurement using a modified Medical The SRC ratio is the equivalent of the normalized slope of the stimulus– Research Council (MRC) score, grading force on a 6-point scale (from response curve. 0=null to 5=normal) in 22 muscle groups [21]. The global MRC Strength–duration properties were determined by measuring the score obtained by adding the scores of all muscle groups (tot-MRC threshold intensities required to obtain 70% of Mmax at 0.1 and 1.0 ms D. Boërio et al. / Journal of the Neurological Sciences 292 (2010) 63–71 65 pulse durations. Strength–duration time constant (chronaxie) and Table 2 rheobase were calculated by using formulae derived from Weiss' Law Parameters of conventional motor nerve conduction study for the right ulnar nerve. CMAP: compound muscle action potential. [26]. Patient Distal CMAP Distal motor Conduction F-wave Proximal motor ½0:1×ðÞi70; 0:1−i70; 1:0 number amplitude (mV) latency (ms) velocity at latency conduction Strength−duration time constant = ; i70; 1:0−ðÞ0:1×i70; 0:1 forearm (m/s) (ms) block (%)

½i70; 1:0−ðÞ0:1×i70; 0:1 Limits of normal values Rheobase = : N5 b3.2 N48 b35 b50 0:9 Multifocal motor neuropathy Finally, conventional motor nerve conduction parameters were 1 12 3 57 32.8 60 collected for the right ulnar nerve at baseline (before IVIg infusions). The 2 8.2 3 58 abs 70 following parameters were measured: Mmax to ulnar nerve stimulation 3 6.2 2.9 41 abs 75 4 11.5 1.8 51 abs 80 at the wrist, distal motor latency, conduction velocity at the forearm, 5 9.2 2.2 58 abs 80 minimal latency of the F waves, and conduction block measurement 6 14 2.5 50 31 70 between CMAPs elicited to supra-maximal stimulation at the wrist, 7 13.5 2.5 52 29.9 No elbow and Erb's point. At Erb's point, a monopolar technique was used to 8 4.6 2.3 39 abs 75 achieve supra-maximal stimulation, the cathode being applied by hand 9 10 2.7 50 34.6 65 10 13 2.9 52 33 50 under pressure on the supraclavicular fossa with a large anode being taped at a distance over the cervicoscapular region. Chronic inflammatory demyelinating neuropathy 1 4 5.3 18 abs 75 2.3. Statistical analyses 2 2.5 2.5 46 abs No 3 10 2.2 41 32.4 No 4 8 3.4 45 37.7 55 Nonparametric tests have been applied because not all data passed 5 10 3.5 34 abs 70 the normality test, as shown by the method of Kolmogorov–Smirnov. 6 3 7.7 25 abs 65 The level of significance was set at Pb0.05. Electrophysiological data 7 2 2.8 30 abs 80 at baseline (before IVIg infusions) were analyzed using the Mann– 8 9 2.5 52 35.8 No 9 15 2.8 46 33.7 No Whitney test, Fisher test, or repeated measures ANOVA (with Dunn's 10 10 4.2 43 35.8 No multiple comparisons post-tests). Correlations between MRC scores and nerve conduction or excitability parameters in patients with Controls MMN and CIDP were studied using the Spearman test. Excitability 1 10 3 50 33 No changes before and after IVIg infusions were studied using repeated 2 10.5 2.5 53 32.3 No 3 8 2.6 50 30.3 No “ ” measures ANOVA with Group (MMN, CIDP, and CTRL) as between 4 8.5 2.8 52 33.8 No factor. In case of significant ANOVA, post-hoc tests were applied. 5 11 2.5 57 26.8 No Finally, a principal component analysis (PCA) was performed to 6 9 2.6 55 28.8 No investigate possible dependence among the different variables and their 7 9 2.1 57 29.2 No 8 9 2.4 59 29.7 No ability to differentiate between groups of patients. The PCA is a statistical 9 12.5 2.5 65 25.7 No analysis which transforms potentially correlated variables into inde- 10 12 2.1 57 25 No pendent components, so-called principal components. The two princi- pal components, which best describe the variance and the correlation between data, are retained for analysis. The results are presented in a with CIDP; absent or delayed F waves in 5/10 patients with MMN and correlation circle and a scatter plot. In the correlation circle, the x-andy- 8/10 patients with CIDP. No patients showed motor conduction block axes represent the participation of each variable to the first and second at the forearm (defined as CMAP amplitude reduction by more than principal components, respectively. Then, a vector corresponding to 30% between elbow and wrist stimulation). Conversely, proximal each variable is projected on the plan defined by the two principal components. Two vectors are more positively correlated when they are closer, aligned in the same direction, and negatively correlated when Table 3 they are opposed. In the present study, variables were nerve excitability Mean (SD) values of nerve conduction and excitability parameters for the right ulnar parameters. For each patient, all excitability results were plotted on the nerve in patients with or without motor deficit in the investigated ulnar territory. vector of the corresponding variable in the PCA plan, and then the Significance (P value) of the statistical comparison between both groups is presented position of the patient was determined on this plan using a barycenter (Mann–Whitney test except *, Fisher test). P value is underlined when significant. method. In the resulting scatter plot, a 95%-confidence ellipse was CMAP: compound muscle action potential. ARP: absolute refractory period; RRP: relative refractory period; SRC: stimulus–response curve. drawn to represent the distribution of the patients in each group. Motor deficit No motor deficit

3. Results Distal CMAP amplitude (mV) 6.1 (3.4) 11.5 (2.4) P=0.004 Distal motor latency (ms) 3.4 (1.8) 2.9 (0.6) P=0.880 All patients completed the study without any adverse event. Conduction velocity at forearm (m/s) 40.0 (13.5) 48.8 (4.9) P=0.140 F-wave presence 0/10 patients 10/10 patients P<0.0001* Proximal motor conduction block (%) 74.4 (5.3) 60.0 (7.9) P=0.009 3.1. Baseline evaluation (pre-IVIg infusions) Presence of proximal conduction block 9/10 patients 5/10 patients P=0.003* ARP min (ms) 1.3 (0.2) 1.1 (0.3) P=0.198 3.1.1. Nerve conduction measures ARP max (ms) 2.6 (0.8) 2.2 (0.5) P=0.199 Conventional nerve conduction studies showed the following Refractoriness (%) 72.4 (25.7) 68.0 (20.8) P=0.677 Supernormality (%) 108.8 (23.2) 132.2 (11.8) P=0.019 abnormalities concerning the investigated ulnar motor nerve RRP (ms) 3.2 (0.3) 3.2 (0.6) P=0.999 (Table 2): reduced CMAP amplitude to wrist stimulation in 1/10 SRC ratio (0.2 ms) 0.8 (0.5) 0.8 (0.2) P=0.821 patient with MMN and 4/10 patients with CIDP; prolonged distal SRC ratio (1 ms) 0.8 (0.3) 0.9 (0.2) P=0.393 motor latency in 5/10 patients with CIDP; reduced motor conduction Chronaxie (ms) 0.3 (0.2) 0.3 (0.1) P=0.473 velocity at the forearm in 2/10 patients with MMN and 9/10 patients Rheobase (mA) 13.1 (7.2) 7.6 (3.5) P=0.166 66 D. Boërio et al. / Journal of the Neurological Sciences 292 (2010) 63–71 conduction block (defined as CMAP amplitude reduction by more 3.1.2. Relationships between nerve conduction measures and muscle than 50% between Erb's point and elbow stimulation) was observed in strength 9/10 patients with MMN and 5/10 patients with CIDP. In the four At baseline, muscle strength was reduced in the investigated ulnar patients with CIDP and reduced CMAP amplitude to wrist stimulation, territory in 5/10 patients with MMN and 5/10 patients with CIDP the presence of very distal conduction block could not be ruled out, (Table 1). The MMN or CIDP patients with ulnar motor deficit (uln- especially for the two patients with concomitantly prolonged distal MRC scoreb5, 10 patients) were compared to those with normal motor latency. strength (uln-MRC score=5, 10 patients) regarding ulnar nerve

Fig. 1. Correlations between the Medical Research Council (MRC) score that was grading force in the distal ulnar nerve territory (Ulnar MRC Score) or in a sum of 22 muscle groups (Total MRC Score) and the amplitude of the compound muscle action potential (CMAP) evoked by ulnar nerve stimulation at the wrist (A,B), the percentage of proximal block (CMAP amplitude decrement greater than 50% between Erb's point and elbow stimulation) (C), the percentage of supernormality (D), the chronaxie (strength duration time constant) (E), or the rheobase (F) in patients with multifocal motor neuropathy (MMN, open triangles—dotted regression line) or chronic inflammatory demyelinating polyneuropathy (CIDP, filled circles—solid regression line). Correlation coefficients and P values are indicated when significant (Spearman test). D. Boërio et al. / Journal of the Neurological Sciences 292 (2010) 63–71 67 conduction parameters. Patients with motor deficit had reduced distal CMAP amplitude, absent F waves, and more severe and frequent proximal conduction blocks (Mann–Whitney or Fisher test, Table 3). The uln-MRC score correlated positively with distal CMAP amplitude in both MMN and CIDP groups and negatively with the percentage of proximal conduction block in patients with CIDP only (Spearman test, Fig. 1). The tot-MRC score correlated positively with distal CMAP amplitude in patients with CIDP only (Fig. 1). No other correlation was observed between parameters of conduction and MRC scores.

3.1.3. Nerve excitability measures Differences in nerve excitability measures between groups were observed in repeated measures ANOVA for the SRC ratios, strength– duration time constant (chronaxie), and rheobase (Table 4). Dunn's post-tests showed: (i) an increased 0.2 ms-SRC ratio in CIDP vs. NML; (ii) an increased 1 ms-SRC ratio in MMN vs. NML; (iii) a reduced strength–duration time constant (chronaxie) and increased rheobase in CIDP vs. CTRL or NML. In PCA, the correlation circle showed that the first principal component was mainly linked to rheobase and chronaxie (x-axis) and the second principal component to 1 ms-SRC ratio and RRP/refracto- riness (y-axis) (Fig. 2A). These two principal components represented 56% of the total variance, a sufficient score to validate the analysis. From the correlation circle, a negative correlation was found between rheobase and strength–duration time constant (chronaxie) on the one hand and between SRC ratios and RRP/refractoriness on the other hand. Conversely, there was a positive correlation between ARPmin/ ARPmax and rheobase on the one hand and between RRP and refractoriness on the other hand. In the scatter plot, the distribution of CTRL patients (95%-confidence ellipse) clearly differed from that of either MMN or CIDP patients (Fig. 2B). The position of CIDP ellipse provides further evidence of increased rheobase and reduced strength–duration time constant (chronaxie) in CIDP compared to CTRL (shift to the left along the x-axis). The position of MMN ellipse also confirmed that 1 ms-SRC ratio was increased in this group compared to CTRL (shift to the bottom along the y-axis). Fig. 2. A. Correlation circle of all nerve excitability variables in the plan of principal 3.1.4. Relationships between nerve excitability measures and muscle component analysis (PCA) before intravenous immunoglobulin infusions in the whole strength series of patients. ARP: absolute refractory period; RRP: relative refractory period; SRC: stimulus–response curve. B. Scatter plot of the distribution of the patients in the PCA plan Nerve excitability values were compared between patients with or according to their group (95%-confidence ellipses). CTRL: patients without any sign of without motor deficit in the investigated ulnar territory. A difference motor nerve involvement in the upper limbs. MMN: patients with multifocal motor was observed only for supernormality, which was lower in the presence neuropathy. CIDP: patients with chronic inflammatory demyelinating polyneuropathy. than in the absence of motor deficit (Mann–Whitney test, Table 3). The uln-MRC score correlated positively with supernormality in both the uln-MRC score correlated negatively with rheobase in patients with MMN and CIDP groups and with strength–duration time constant CIDP (Fig. 1). No other correlation was observed between parameters of (chronaxie) in patients with CIDP only (Spearman test, Fig. 1). Conversely, nerve excitability and MRC scores, including tot-MRC scores.

Table 4 Mean (SD) values of nerve excitability parameters measured before (pre) and after (post) intravenous immunoglobulin (IVIg) infusions in patients with multifocal motor neuropathy (MMN), chronic inflammatory demyelinating polyneuropathy (CIDP), or in controls without any sign of motor nerve involvement in the upper limbs (CTRL). Normal values in a series of healthy subjects are also presented (NML). Statistical analyses compare baseline (pre-IVIg) characteristics and IVIg-induced changes between groups by means of repeated measures ANOVA. P value is underlined when significant. ARP: absolute refractory period; RRP: relative refractory period; SRC: stimulus–response curve.

MMN group CIDP group CTRL group NML group Differences at baseline Difference in (pre-IVIgs) between IVIg-induced changes Pre Post Pre Post Pre Post groups (P) between groups (P)

ARP min (ms) 1.20 (0.29) 1.15 (0.28) 1.23 (0.26) 1.10 (0.15) 1.01 (0.12) 1.05 (0.21) 1.04 (0.19) 0.158 0.128 ARP max (ms) 2.21 (0.72) 2.37 (0.74) 2.54 (0.69) 2.45 (0.65) 2.02 (0.37) 1.91 (0.32) 2.07 (0.30) 0.373 0.881 Refractoriness (%) 68.9 (24.5) 75.0 (22.0) 71.5 (22.3) 72.3 (24.9) 81.1 (13.5) 84.1 (12.3) 86.0 (13.9) 0.271 0.344 Supernormality (%) 128.4 (12.4) 126.0 (25.9) 112.6 (26.3) 116.7 (17.1) 127.0 (15.2) 127.7 (15.5) 135.0 (11.3) 0.174 0.847 RRP (ms) 3.06 (0.44) 3.45 (0.49) 3.27 (0.51) 3.31 (0.73) 3.48 (0.72) 3.68 (0.57) 3.43 (0.57) 0.412 0.114 SRC ratio (0.2 ms) 0.68 (0.21) 0.64 (0.23) 0.94 (0.43) 1.18 (0.54) 0.55 (0.18) 0.51 (0.15) 0.52 (0.11) 0.024 0.408 SRC ratio (1 ms) 0.89 (0.22) 0.72 (0.34) 0.85 (0.34) 0.99 (0.37) 0.67 (0.16) 0.61 (0.22) 0.58 (0.09) 0.017 0.611 Chronaxie (ms) 0.35 (0.17) 0.21 (0.13) 0.26 (0.16) 0.23 (0.13) 0.42 (0.08) 0.43 (0.09) 0.43 (0.02) 0.045 0.009 Rheobase (mA) 8.34 (4.70) 9.95 (4.81) 12.36 (7.01) 14.49 (8.58) 4.77 (2.17) 5.89 (3.44) 4.88 (0.66) 0.005 0.005 68 D. Boërio et al. / Journal of the Neurological Sciences 292 (2010) 63–71

3.2. IVIg-induced changes positive correlation between SRC ratios. In the scatter plot, the distribution of CTRL patients (95%-confidence ellipse) still clearly 3.2.1. Nerve excitability measures differed from that of either MMN or CIDP patients (Fig. 5B). Again, the Individual variations in nerve excitability parameters are pre- position of CIDP ellipse mainly reflected changes in rheobase/ sented in Figs. 3 and 4. Changes between pre and post IVIg infusions chronaxie compared to CTRL (shift to the left along the x-axis) and were observed in repeated measures ANOVA for strength–duration that of MMN ellipse also reflected changes in SRC ratio/ARP- time constant (chronaxie) and rheobase (Table 4), but time×group refractoriness (shift to the bottom along the y-axis). interaction was only significant for strength–duration time constant (P=0.005). Dunn's post-hoc tests showed that rheobase increased 3.2.2. Relationships between changes in nerve excitability measures and after IVIg infusions in all groups, while strength–duration time muscle strength constant decreased in CIDP and MMN groups, but not in CTRL At the end of IVIg infusions, the uln-MRC score increased only in group. Nevertheless, the negative correlation between rheobase and two patients with CIDP, while the tot-MRC score increased in 5/10 strength–duration time constant did not change in any group before patients with MMN and 6/10 patients with CIDP (Table 1). No and after IVIg infusions (data not shown, Spearman test). correlation was found between changes in MRC scores and nerve In PCA, the principal components of the correlation circle were excitability measures after IVIg infusions (data not shown, Spearman again mainly linked to rheobase and chronaxie (x-axis) and to SRC test). ratio and refractoriness (y-axis), like before IVIg infusions (Fig. 5A). Patients in whom the tot-MRC score improved after IVIg infusions The two principal components represented 60% of the total variance, had a greater percentage of proximal conduction block on ulnar nerve confirming the validity of PCA. From the correlation circle, a negative at baseline than those in whom the tot-MRC score remains stable correlation was found between rheobase and strength–duration time (mean (SD): 74.4 (5.7) % vs. 62.5 (9.4) %, P=0.028, Mann–Whitney constant (chronaxie) on the one hand and between SRC ratios and test). No other baseline nerve conduction or excitability parameters refractoriness on the other hand. Conversely, there was a strong were associated with post IVIg MRC score changes.

Fig. 3. Individual changes in nerve excitability values before (pre) and after (post) intravenous immunoglobulin infusions in patients with multifocal motor neuropathy (MMN), chronic inflammatory demyelinating polyneuropathy (CIDP), or in controls without any sign of motor nerve involvement in the upper limbs (CTRL). ARP: absolute refractory period; RRP: relative refractory period. D. Boërio et al. / Journal of the Neurological Sciences 292 (2010) 63–71 69

Fig. 4. Individual changes in nerve excitability values before (pre) and after (post) intravenous immunoglobulin infusions in patients with multifocal motor neuropathy (MMN), chronic inflammatory demyelinating polyneuropathy (CIDP), or in controls without any sign of motor nerve involvement in the upper limbs (CTRL). SRC: stimulus–response curve.

4. Discussion conductance [31,32]. Persistent Na+ currents represent only 1% of the total Na+ conductance for normal motor axons [31], but are thought 4.1. Baseline evaluation to be increased in demyelinated axons, as recently shown in multiple sclerosis via the diffuse expression of Nav1.2 and Nav1.6 voltage- Excitability parameters of neurological controls (CTRL group) did gated sodium channels [33]. By restoring low levels of persistent Na+ not differ from those of healthy subjects (NML group). Patients with currents through “immune” or “non-immune” effect in the periaxonal MMN also did not differ from normals, except for 1 ms-SRC ratio. environment, IVIgs could reduce intraaxonal Na+ accumulation and These results confirmed previous studies showing nearly (but not then could diminish Na+/K+ ATPase pump function requirements. “strictly”) normal excitability profiles in territories unaffected by This could also support protective effects of IVIgs in the long-term, as conduction blocks in MMN [27,28]. intraaxonal Na+ accumulation associated with energetic failure is Conversely, patients with CIDP showed significant abnormalities known to lead to axonal degeneration in inflammatory demyelinating compared to CTRL or NML values: 0.2 ms-SRC ratio and rheobase were processes [34,35]. The neuroprotective value of IVIg infusions was increased, whereas strength–duration time constant was reduced, as suggested by the finding of long-term functional stability in patients previously reported [29,30]. Because one might expect an increased treated by high and frequent doses of IVIgs [36,37]. The respective strength–duration time constant in demyelinating diseases, its contribution of changes in passive membrane properties and reduction in patients with CIDP was attributed to a short-circuit due decreased persistent Na+ currentcouldbemorespecifically to subperineural edema or to a decrease in persistent Na+ channel addressed by the technique of latent addition [31] that has been density [29]. A reduced supernormality has also been reported in recently applied to investigate peripheral neuropathies in humans patients with CIDP [27,29,30]. This was not confirmed in the present [38,39]. The latent addition technique relies on the combination of a work. Nevertheless, supernormality was lower in case of motor prolonged conditioning subthreshold hyperpolarizing current and a deficit, regardless CIDP or MMN. Motor deficit was also associated brief test suprathreshold stimulus delivered at various intervals from with reduced CMAP amplitude, F-wave absence, and proximal the conditioning current. Two types of changes can be observed: fast conduction block. In demyelinating neuropathies, weakness may changes related to passive membrane properties and slow changes relate to various mechanisms, including alteration in myelin sheath related to persistent Na+ currents [31]. properties, internodal distances, and secondary axonal loss. Investigations were performed in motor nerve segments that were unaffected by conduction block. Therefore, the modification of axonal 4.2. IVIg-induced changes membrane properties that we observed could also reflect compensatory mechanisms occurring outside a zone of segmental demyelination, as Rheobase increased after IVIg infusions in all groups, whereas previously suggested [19]. An additional negative effect on nerve strength–duration time constant (chronaxie) decreased in CIDP and excitability, like strength–duration time constant decrease, located MMN, but not in CTRL group. This result was surprising because outside the zone of conduction block is not necessarily inconsistent with strength–duration time constant was already reduced in patients with clinical improvement following IVIg infusions. The present results show CIDP at baseline. This result can be explained by various mechanisms, that IVIgs are able to rapidly modify nerve excitability rather they one being related to the modulation of Na+ currents at axonal level. explain the mechanisms underlying the clinical benefit. Excitability Actually, strength–duration time constant depends on persistent Na+ measures across the sites of block would provide more fruitful 70 D. Boërio et al. / Journal of the Neurological Sciences 292 (2010) 63–71

improvement. It should be interesting to study nerve excitability changes in correlation with clinical evolution over time following IVIg infusions, but this time course is largely variable between patients.

5. Conclusions

The mechanism of action of IVIgs in inflammatory demyelinating neuropathies has yet to be clarified. The decrease in strength–duration time constant after IVIg infusions is the main positive finding of this study. Whatever the underlying processes, this demonstrates the occurrence of significant changes in axonal excitability, early after IVIg infusions, even when motor performance not yet improved. Excitability changes may precede clinical changes that develop a few days after the end of IVIg infusions in patients with MMN or CIDP [43].Nerve excitability studies could be fruitfully applied to better understand the mechanisms of action of various therapies in peripheral neuropathies.

Acknowledgments

D. Boërio was granted by LFB. The authors gratefully acknowledge Prof. Hugh Bostock for his helpful comments on the manuscript.

References

[1] Hughes RA, Bouche P, Cornblath DR, Evers E, Hadden RD, Hahn A, et al. European Federation of Neurological Societies/Peripheral Nerve Society guideline on management of chronic inflammatory demyelinating polyradiculoneuropathy. Eur J Neurol 2006;13:326–32. [2] van Schaik IN, Bouche P, Illa I, Léger JM, Van den Bergh P, Cornblath DR, et al. European Federation of Neurological Societies/Peripheral Nerve Society guideline on management of multifocal motor neuropathy. Eur J Neurol 2006;13:802–8. [3] Hahn AF, Bolton CF, Zochodne D, Feasby TE. Intravenous immunoglobulin treatment in chronic inflammatory demyelinating polyneuropathy. A double- blind, placebo-controlled, cross-over study. Brain 1996;119:1067–77. [4] Mendell JR, Barohn RJ, Freimer ML, Kissel JT, King W, Nagaraja HN, et al. Randomized controlled trial of IVIg in untreated chronic inflammatory demyelinating poly- radiculoneuropathy. Neurology 2001;56:445–9. [5] Vermeulen M, van Doorn PA, Brand A, Strengers PF, Jennekens FG, Busch HF. Intravenous immunoglobulin treatment in patients with chronic inflammatory demyelinating polyneuropathy: a double blind, placebo controlled study. J Neurol Neurosurg Psychiatry 1993;56:36–9. Fig. 5. A. Correlation circle of all nerve excitability variables in the plan of principal [6] Chaudhry V, Corse AM, Cornblath DR, Kuncl RW, Drachman DB, Freimer ML, et al. Multifocal motor neuropathy: response to human immune globulin. Ann Neurol component analysis (PCA) after intravenous immunoglobulin infusions in the whole series 1993;33:237–42. of patients. ARP: absolute refractory period; RRP: relative refractory period; SRC: stimulus– [7] Kaji R, Shibasaki H, Kimura J. Multifocal demyelinating motor neuropathy: cranial response curve. B. Scatter plot of the distribution of the patients in the PCA plan according nerve involvement and immunoglobulin therapy. Neurology 1992;42:506–9. fi to their group (95%-con dence ellipses). CTRL: patients without any sign of motor nerve [8] Nobile-Orazio E, Meucci N, Barbieri S, Carpo M, Scarlato G. High-dose intravenous involvement in the upper limbs. MMN: patients with multifocal motor neuropathy. CIDP: immunoglobulin therapy in multifocal motor neuropathy. Neurology 1993;43: patients with chronic inflammatory demyelinating polyneuropathy. 537–44. [9] Léger JM, Chassande B, Musset L, Meininger V, Bouche P, Baumann N. Intravenous immunoglobulin therapy in multifocal motor neuropathy: a double-blind, information. However, this approach is usually unsuccessful, as the placebo-controlled study. Brain 2001;124:145–53. nerve is, of course, poorly excitable in those segments [19]. [10] Van den Berg LH, Kerkhoff H, Oey PL, Franssen H, Mollee I, Vermeulen M, et al. Treatment of multifocal motor neuropathy with high dose intravenous immu- noglobulins: a double blind, placebo controlled study. J Neurol Neurosurg 4.3. Limitations Psychiatry 1995;59:248–52. [11] Van den Berg LH, Franssen H, Wokke JH. The long-term effect of intravenous This study has several limitations. First, literature data regarding immunoglobulin treatment in multifocal motor neuropathy. Brain 1998;121:421–8. [12] Créange A, Gregson NA, Hughes RA. Intravenous immunoglobulin modulates excitability studies in dysimmune neuropathies are based on lymphocyte CD54 and monocyte FcgammaRII expression in patients with chronic ‘threshold-tracking’ methods that differ from the present ones and inflammatory neuropathies. J Neuroimmunol 2003;135:91–5. we do not know if these two methods are comparable. The [13] Dalakas M. Mechanisms of action of IVIg and therapeutic considerations in the reproducibility of such measures could also be questioned. This treatment of acute and chronic demyelinating neuropathies. Neurology 2002;59: 13–21. issue has been previously addressed by several studies, including ones [14] Kazatchkine M, Kaveri SV. Immunomodulation of autoimmune and inflammatory performed in our laboratory, showing satisfactory test–retest repeat- diseases with intravenous immune globulin. N Engl J Med 2001;345:747–55. ability [23,40,41]. Second, the relationships between nerve excitabil- [15] Bostock H, Cikurel K, Burke D. Threshold tracking techniques in the study of human peripheral nerve. Muscle Nerve 1998;21:137–58. ity and conduction parameters were assessed before but not after IVIg [16] Kiernan MC, Burke D, Andersen KV, Bostock H. Multiple measures of axonal infusions, because motor nerve conduction studies were not repeated excitability: a new approach in clinical testing. Muscle Nerve 2000;23:399–409. after IVIgs. Actually, conduction studies do not provide parameters [17] Krishnan AV, Lin CS, Park SB, Kiernan MC. Assessment of nerve excitability in toxic and metabolic neuropathies. J Peripher Nerv Syst 2008;13:7–26. sensitive enough to identify nerve function changes related to this [18] Lefaucheur JP, Boërio D, Hogrel JY, Créange A. Etude de l'excitabilité nerveuse dans treatment in the short term. In particular, literature results are l'exploration des neuropathies dysimmunitaires. Rev Neurol (Paris) 2006;162: conflicting regarding the value of conduction block estimates as reliable 3S19-28. fi [19] Kiernan MC, Guglielmi JM, Kaji R, Murray NM, Bostock H. Evidence for axonal indicators of IVIg ef cacy [9,10,42]. Third, axonal excitability was tested membrane hyperpolarization in multifocal motor neuropathy with conduction soon after the end of IVIg infusions, prior to the peak of clinical block. Brain 2002;125:664–75. D. Boërio et al. / Journal of the Neurological Sciences 292 (2010) 63–71 71

[20] Priori A, Cinnante C, Pesenti A, Carpo M, Cappellari A, Nobile-Orazio E, et al. [33] Craner MJ, Newcombe J, Black JA, Hartle C, Cuzner ML, Waxman SG. Molecular Distinctive abnormalities of motor axonal strength–duration properties in multifocal changes in neurons in multiple sclerosis: altered axonal expression of Nav1.2 and motor neuropathy and in motor neurone disease. Brain 2002;125:2481–90. Nav1.6 sodium channels and Na+/Ca2+ exchanger. Proc Natl Acad Sci U S A [21] Kleyweg RP, van der Meché FGA, Schmitz PIM. Interobserver agreement in the 2004;101:8168–73. assessment of muscle strength and functional abilities in Guillain–Barre [34] Bechtold DA, Smith KJ. Sodium-mediated axonal degeneration in inflammatory syndrome. Muscle Nerve 1991;14:1103–9. demyelinating disease. J Neurol Sci 2005;233:27–35. [22] Ingram DA, Davis GR, Swash M. The double collision technique: a new method for [35] Stys PK, Waxman SG. Activity-dependent modulation of excitability: implication measurement of the motor nerve refractory period distribution in man. for axonal physiology and physiolpathology. Muscle Nerve 1994;17:969–74. Electroencephalogr Clin Neurophysiol 1987;66:225–34. [36] Vucic S, Black KR, Chong PS, Cros D. Multifocal motor neuropathy: decrease in [23] Boërio D, Hogrel JY, Créange A, Lefaucheur JP. A reappraisal of various methods for conduction blocks and reinnervation with long-term IVIg. Neurology 2004;63:1264–9. measuring motor nerve refractory period in humans. Clin Neurophysiol 2005;116: [37] Vucic S, Black K, Baldassari LE, Tick Chong PS, Dawson KT, Cros D. Long-term effects of 969–76. intravenous immunoglobulin in CIDP. Clin Neurophysiol 2007;118:1980–4. [24] Kiernan MC, Mogyoros I, Burke D. Differences in the recovery of excitability in [38] Misawa S, Kuwabara S, Kanai K, Tamura N, Nakata M, Ogawara K, et al. Nodal sensory and motor axons of human median nerve. Brain 1996;119:1099–105. persistent Na+ currents in human diabetic nerves estimated by the technique of [25] Brismar T. Changes in electrical threshold in human peripheral neuropathy. latent addition. Clin Neurophysiol 2006;117:815–20. J Neurol Sci 1985;68:215–23. [39] Tamura N, Kuwabara S, Misawa S, Kanai K, Nakata M, Sawai S, et al. Increased [26] Bostock H, Bergmans J. Post-tetanic excitability changes and ectopic discharges in nodal persistent Na+ currents in human neuropathy and motor neuron disease a human motor axon. Brain 1994;117:913–28. estimated by latent addition. Clin Neurophysiol 2006;117:2451–8. [27] Cappelen-Smith C, Kuwabara S, Lin CS, Mogyoros I, Burke D. Activity-dependent [40] Boërio D, Hogrel JY, Lefaucheur JP, Wang FC, Verschueren A, Pouget J, et al. hyperpolarization and conduction block in chronic inflammatory demyelinating Stimulus–response curve of human motor nerves: multicenter assessment of polyneuropathy. Ann Neurol 2000;48:826–32. various indexes. Neurophysiol Clin 2008;38:31–8. [28] Cappelen-Smith C, Kuwabara S, Lin CS, Burke D. Abnormalities of axonal [41] Mogyoros I, Lin C, Dowla S, Grosskreutz J, Burke D. Reproducibility of indices of excitability are not generalized in early multifocal motor neuropathy. Muscle axonal excitability in human subjects. Clin Neurophysiol 2000;111:23–8. Nerve 2002;26:769–76. [42] Ghosh A, Virgincar A, Kennett R, Busby M, Donaghy M. The effect of treatment [29] Cappelen-Smith C, Kuwabara S, Lin CS, Mogyoros I, Burke D. Membrane properties in upon temporal dispersion in IvIg responsive multifocal motor neuropathy. chronic inflammatory demyelinating polyneuropathy. Brain 2001;124:2439–47. J Neurol Neurosurg Psychiatry 2005;76:1269–72. [30] Cappelen-Smith C, Lin CS, Kuwabara S, Burke D. Conduction block during and after [43] Van Doorn PA, Vermeulen M, Brand A, Mulder PG, Busch HF. Intravenous ischaemia in chronic inflammatory demyelinating polyneuropathy. Brain immunoglobulin treatment in patients with chronic inflammatory demyelinating 2002;125:1850–8. polyneuropathy. Clinical and laboratory characteristics associated with improve- [31] Bostock H, Rothwell JC. Latent addition in motor and sensory fibres of human ment. Arch Neurol 1991;48:217–20. peripheral nerve. J Physiol (London) 1997;498:277–94. [32] Mogyoros I, Kiernan MC, Burke D. Strength–duration properties of human peripheral nerve. Brain 1996;119:439–47.