Long-latency reflexes in patients with Behçet's disease

Journal of the Neurological Sciences 220 (2004) 67 – 71 www.elsevier.com/locate/jns

Long-latency reflexes in patients with Behc¸et’s disease Cengiz Tataroglu a,*, Umit Tursen b, Ozgur Unal a, Tamer Irfan Kaya b a

Department of Neurology, School of Medicine, Mersin University, Mersin 33079, Turkey b Department of Dermatology, School of Medicine, Mersin University, Mersin, Turkey

Received 25 July 2003; received in revised form 5 February 2004; accepted 6 February 2004

Abstract Objective: Recent studies demonstrate that the subclinical involvement of motor pathways is frequently observed in patients with Behc¸et’s disease (BD). Long-latency reflexes (LLR) provide information about the continuity of both ascending and descending neural pathways. Our aim was to evaluate the utility of LLR and somatosensory-evoked potentials (SEP) in demonstrating subclinical neural involvement in patients with BD. Methods: Twenty-nine patients with BD were studied by means of SEP and LLR. Bilateral median nerve SEPs and LLRs evoked by electrical stimulation of both median nerves were recorded. The latency of second component of LLR (LLR2), the duration of LLR2 – HR (Hoffmann reflex, spinal reflex component of LLR) interval, peak to peak amplitude of LLR2 and the amplitude ratio of LLR2/ HR were analyzed. The data obtained from patients were compared with those of 20 control subjects. Results: LLR2 latencies and the durations of LLR2 – HR interval were significantly prolonged in patients with BD ( p = 0.001 for both parameters). Increased duration of LLR2 – HR interval was the most frequent abnormality observed in the study (37.9%). Conclusion: Our findings suggest that LLR is a useful technique to demonstrate subclinical neural involvement in patients with BD. D 2004 Elsevier B.V. All rights reserved. Keywords: Behc¸et’s disease; Long-latency reflex; Somatosensory-evoked potentials; Neuro-Behc¸et’s disease

1. Introduction Behc¸et’s disease (BD) is a multisystem heterogeneous inflammatory disorder with unknown etiology. Vasculitis is the major pathological feature of this disorder [1]. Neurological involvement [neuro-Behc¸et’s disease (NBD)] has been reported in about 5– 49% of patients with BD [2 – 4]. When neurological involvement is present, early diagnosis and treatment is essential to reduce progression of CNS pathology [5]. Electrophysiological studies may provide functional information about neural involvement complementary to magnetic resonance imaging (MRI) in BD. The usefulness of evoked potentials in the diagnosis of neural involvement in patients with BD is controversial [5 –8]. However, it is claimed that motor-evoked potentials (MEP) evoked by transcranial magnetic stimulation is a more sensitive test

* Corresponding author. Tel.: +90-324-3374300-1113; fax: +90-3243374305. E-mail addresses: [email protected], [email protected] (C. Tataroglu). 0022-510X/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2004.02.005

in demonstrating neural involvement in symptomatic and asymptomatic patients [2,4]. Long-latency reflex response (LLR) evoked by electrical stimulation of mixed peripheral nerve is a well-documented reflex response [9]. Reflex pattern consists of the spinal reflex (Hoffmann reflex—HR), followed by up to three reflex responses (LLR1, 2, 3). HR is mediated by monosynaptic projection of group 1a sensorial afferents onto the alpha motor neurons and relayed monosynaptically. LLRs are thought to be reflex responses originating from supraspinal neural pathways. First and third responses (LLR1 and 3) are not stable responses. Central pathways of these waves remain unclear. LLR1 is considered a transcortical reflex and central course of LLR3 has probably more complex pathways. LLR1 and LLR3 can be obtained in only 20 –30% of normal subjects, whereas LLR2 can be evoked in 100% of normal subjects. The range of LLR2 onset latency was reported between 45 to 58 ms in previous studies. Afferent impulse of LLR2 is transmitted by Ia afferent fibers. It was demonstrated that LLR and somatosensory-evoked potentials (SEP) share same afferent neural pathway. On the other hand, it has been supposed that the efferent branch of the LLR is

68

C. Tataroglu et al. / Journal of the Neurological Sciences 220 (2004) 67–71

identical with MEP [10 –12]. The analysis of LLR2 may give valuable information about involvement of both afferent and efferent neural pathways. LLR have been studied in patients with upper motor neuron disorders. Delayed or absent LLR2 component were the most common abnormalities observed in patients with multiple sclerosis [11,13]. In patients with spasticity due to various etiologies, enlarged amplitudes of HR and absence or attenuation of LLR2 were observed [14]. To our knowledge, LLR has not been studied in patients with BD previously. The aims of this study were to evaluate the utility of LLR in BD and to compare it with somatosensory-evoked potentials (SEP) in demonstrating subclinical neural involvement in patients with BD.

2. Materials and methods 2.1. Subjects Twenty-nine consecutive patients diagnosed as BD according to the criteria of the international study group for BD were included in this study (oral ulceration at least three times within 12 months combined with at least two of the following: recurrent genital ulceration, eye lesions, skin lesions and positive pathergy test) [15]. The patients were 16 males and 13 females whose ages ranged from 23 to 60 years (mean: 37.8 F 9.1). All of the subjects were neurologically asymptomatic. No patient showed any evident neurological finding during the electrophysiological investigation. Cranial MRI investigations of 15 patients who had nonspecific neurological symptoms, such as headache, dizziness and subjective sensorial complaints, did not demonstrate any CNS involvement. Other patients did not need detailed neuroradiological investigation. The disease duration ranged from 3 months to 15 years (mean: 6.1 F 4.1 years). The patients who had any history of neurological involvement or any neurological finding were excluded from the study. No patient had any other systemic disorder. The data obtained from patients were compared with those of 20 control subjects (12 males, 8 females). Age range of controls was 23 to 52 (mean: 37.1 F 9.5). The written informed consent for the study was obtained from each patient and the study was approved by the local ethical committee.

cy was 3 Hz. Amplificatory filters were set between 2 Hz – 2 kHz. Oscilloscope analysis time was 100 ms. Subjects were requested to have sustained slight voluntary contraction of their thenar muscle during investigation (it was about 10% of maximal power). The EMG was recorded with conventional surface electrodes. The degree of EMG activity was observed visually from the oscilloscope during investigation. EMG signals were recorded and averaged (400 sweeps for each side). Medelec Synergy EMG equipment was used for averaging and analyzing. We evaluated the onset latencies of spinal reflex response (HR) and second late response (LLR2), the peak to peak amplitudes of LLR2 and HR, the amplitude ratio of LLR2/HR and the duration of LLR2 –HR interval. The latencies and the peak to peak amplitudes (absolute amplitudes) of the LLR2 and HR components were visually determined with cursors at their onsets and positive –negative peaks. LLR2 component were identified as the most stable response recorded after from HR. The onset latency of LLR2 was analyzed at the beginning of the stable EMG burst after HR. If the precise onset remains in doubt, onset latency was measured at the point where EMG responses and baseline cross. Additionally, interside differences of the latency of LLR2 and the duration of LLR2 – HR interval were analyzed. An absence of LLR2 potentials or a delay in LLR2 latency, a reduction of LLR2 amplitude and an increase in LLR2 – HR interval were considered abnormal. Both median nerve was stimulated at the wrist with a square wave current pulse of 0.1 ms duration at 3 Hz for upper extremity somatosensory-evoked potentials (SEP). Recordings were made from Ag/AgCl surface electrodes attached to the skin. Erb– Fz, Cv7 –Fz, Cz –Fz and Cc – Ci (contralateral –ipsilateral somatosensory cortex) recordings were used. A thousand responses were averaged for median SEP. All recordings were performed twice. The stimulus intensity was adjusted to produce a minimal movement at the thumb. Ossiloscope analysis time was 30 ms. Amplificator filters were between 2 Hz – 2 kHz. The peak latencies of N9, N13 and N20 potentials were measured. Central somatosensorial conduction time (CSCT) was analyzed by the subtraction of peak latency of N13 from peak latency of N20 potential. The prolongation of N20 peak latency and CSCT were considered as abnormal. SEPs recorded from Cc – Ci (contrlateral –ipsilateral) electrodes were used for the data analysis. 2.3. Statistical analysis

2.2. Electrophysiological studies For electrophysiological investigations, subjects were instructed to lie down on a comfortable armchair. LLR were recorded from both upper extremity thenar muscles by using square wave electrical stimulation of both median nerves at the wrist level. The stimulation intensity was kept at the motor threshold for motor fibers of median nerve. Stimulus duration was 0.2 ms. Stimulus repetition frequen-

Electrophysiological results obtained from patients and controls were compared with Student’s t-test (when appropriate, Mann –Whitney U-test was used). Pearson correlation test (when appropriate, Spearman’s test) was used to show any relation between electrophysiological parameters and disease duration. Criteria for pathology were latencies of LLR2 and LLR2 –HR interval values exceeding the upper confidence limit (mean + 2.5 S.D. of normal con-

C. Tataroglu et al. / Journal of the Neurological Sciences 220 (2004) 67–71

Fig. 1. The latencies of LLR2 and LLR2 – HR interval time were significantly prolonged in patients with BD.

trols). Chi-square and Fisher’s tests (when appropriate) were used to compare the number of patients who’s exceeded + 2.5 S.D. of the normal controls.

3. Results Electrophysiological data were recorded and analyzed bilaterally in the patients and the controls. Each side was compared between patients and controls separately. LLR2 and SEP recordings were successfully registered in all subjects. Mean latency of LLR2 and the duration of LLR2 –HR interval were significantly prolonged in patients with BD ( p = 0.001; Fig. 1). The amplitudes of LLR2 and N20 potentials, the latency of N20 and CSCT did not show

69

significant differences between the patients and the controls. The mean amplitude of the LLR2 was relatively decreased in the patient group but this difference did not reach statistical significance ( p = 0.08 for left side, p = 0.4 for right side). Mean amplitude of HR and the amplitude ratio of LLR2/HR were not different between the patients and the controls (Table 1). The upper limit of the latency of LLR2 (mean + 2.5 S.D.) was 52.1 ms and that of duration of LLR2 –HR interval (mean + 2.5 S.D.) was 24 ms. Prolongation of LLR2 was found in eight patients (27.6%), and 11 out of 58 limbs were tested. The patients with LLR2 prolongation was significantly higher than the controls (v2 = 6.22, p = 0.014, Fisher’s p = 0.015). Increased duration of LLR2 –HR interval was found in 11 patients (37.9%), and 16 out of 58 limbs were tested. The number of patients with LLR2 – HR interval prolongation was significantly higher than controls (v2 = 9.23, p = 0.002, Fisher’s p = 0.003). Three patients had LLR2 – HR interval prolongation without LLR2 prolongation. Upper limits of interside differences of the LLR2 latency and LLR2 – HR interval were 2.5 and 5.3 ms, respectively. Increased interside differences were observed in 7 patients (24.1%). Significantly, more patients showed increased interside differences compared to controls (v2 = 5.31, p = 0.02, Fisher’s p = 0.03). The upper limits of N20 latencies and CSCT (mean + 2.5 S.D.) were 21.4 and 7.8 ms, respectively. Only three patients (10.7%) showed prolongation of peak latency of N20 and/or prolonged CSCT (one bilateral, two unilateral). These patients had already LLR2 abnormalities. The patients and the controls did not show significant differences with respect to SEP abnormalities (Fisher’s p = 0.8; Fig. 2). There were poor correlations between the latencies of N20 and LLR2 (r = 0.2, p = 0.12) and between CSCT and LLR2 – HR time interval (r = 0.1, p = 0.8) in the patient group. There was a strong correlation between LLR2 latency and LLR2 – HR time interval (r = 0.85, p < 0.0001). The duration of disease duration did not show any correla-

Table 1 The comparison of electrophysiological parameters obtained from patients and controls Patients

LLR2 latency (ms) LLR2 amplitude (AV) LLR2 – HRa (ms) N20 latency (ms) N20 amplitude (AV) CSCTb (ms) LLR2/HR ratio HR amplitude

Controls

Right

Left

Right

Left

48.5 F 3.9 (44.4 – 53.5) 130.3 F 135.9 (25 – 500) 22.1 F 2.6 (18.0 – 27.3) 20.1 F 3.5 1.5 F 0.9 6.9 F 2.7 0.48 F 0.38 (0.11 – 1.5) 274.2 F 313.5 (50 – 1200)

48.8 F 3.4 (44.6 – 56.6) 110.9 F 78.2 (25 – 260) 23.0 F 3.3 (18.1 – 31.4) 19.01 F 1.4 1.6 F 0.9 6.8 F 0.9 0.40 F 0.36 (0.10 – 1.40) 278.3 F 325.2 (50 – 1200)

45.6 F 2.6 (43.6 – 48.9) 157.2 F 103.6 (50 – 400) 19.7 F 1.7 (17.7 – 23.0) 18.7 F 1.3 1.9 F 0.6 6.1 F 0.7 0.56 F 0.2 (0.25 – 1.2) 280.6 F 223.1 (90 – 800)

45.8 F 1.5 (43.3 – 49.0) 153.9 F 82.9 (80 – 350) 20.3 F 1.7 (17.1 – 22.9) 18.6 F 0.9 1.9 F 0.5 6.1 F 0.5 0.50 F 0.20 (0.20 – 1.10) 305.3 F 158.7 (100 – 600)

p1

p2

0.001 0.4 0.0001 0.08 0.17 0.2 0.4 0.9

0.001 0.08 0.002 0.2 0.2 0.1 0.6 0.7

p1: p-value between right sides of patients and controls. p2: p-value between left sides of patients and controls. The latencies of LLR2 and LLR2 – HR interval showed prolongation in patients with BD. Additionally, maximum and minimum ranges of LLR values were shown. a The latency difference between LLR2 and HR (LLR2 – HR interval). b Central sensory conduction time (N20 – N13).

70

C. Tataroglu et al. / Journal of the Neurological Sciences 220 (2004) 67–71

Fig. 2. LLR traces obtained from a control subject (A) and a patient with BD (B). The latency of LLR2 and the duration of LLR2 – HR interval obtained from right side were showed moderately prolongation. Arrows indicate the onset latencies of LLR2 components. Broken lines show the onset latencies of HR.

tion with electrophysiological parameters, except a mild correlation with LLR2 latency (r = 0.37, p = 0.046).

4. Discussion Subclinical neural involvement in BD has been reported previously [5,8,16]. Frank onset of neurological involvement is commonly 4 – 6 years after the onset of BD. However, neurological involvement due to BD occurs prior to its characteristic oral and cutaneous lesions in some patients [1]. Additionally, subclinical neurological involvement has been clearly demonstrated in some series previously and some of these patients had developed acute attacks in a clinical follow-up period [1]. Therefore, it should not be surprising to diagnose subclinical neural involvement in patients with BD. Early detection of neural involvement may provide early establishment of appropriate treatment for neuro-Behcßet’s disease. We enrolled this study because electrophysiological assessment can be useful in evaluation of the existence of subclinical neurological involvement in patients with BD. LLR2 abnormalities (prolonged LLR2 latency and increased LLR2 –HR interval) were observed in 37.9% of BD patients without neurological involvement in present study. Our results suggest that LLR can be a useful electrophysiological method in evaluation of patients with BD. Afferent and efferent reflex arcs of LLR are identical

with neural pathways for SEP and MEP [10 – 12]. Therefore, LLR abnormalities can result from any lesion of both affecting both ascending somatosensory and descending motor pathways. Among the multimodal evoked potential techniques, MEP abnormalities were more commonly observed in patients with BD [2,4,17]. Stigsby et al. [17] observed a significant prolongation of central motor conduction time (CMCT) in nine patients with BD without major impairment of central sensorial conduction. Parisi et al. [2] demonstrated that the MEP abnormalities were observed in 55% of BD patients without evident neurological involvement. They concluded that MEP was a useful tool in demonstration of presymptomatic central motor involvement in BD. Recently, Stigsby et al. [4] showed that CMCT and MEP abnormalities (89%) were slightly more common than MRI (85%) and considerably more common than SEP and BAEP (31%) abnormalities in patients with BD. Besana et al. [5] also reported that the sensitivity of SEP was very low in BD. Because BD primarily affects the motor circuits in the CNS, predominance of MEP abnormalities in BD may not be an unexpected finding [1]. The predilection of motor involvement may be explained by the more common involvement of small- and intermediatesized cerebral vessels and, in this way, lateral and ventral brain-stem in neuro-Behcß et disease [4]. Our results showed that the analysis of SEP did not provide any additional information to LLR. Therefore, it seems that the analysis of SEP is not essential in patients with BD without evident

C. Tataroglu et al. / Journal of the Neurological Sciences 220 (2004) 67–71

neurological involvement. LLRs can be compared with motor-evoked potentials in further studies. LLR abnormalities were described in some neurological disorders such as multiple sclerosis [11– 13]. Delayed latency of LLR2 and absence of LLR2 were the main abnormalities observed in patients with multiple sclerosis. The analysis of LLR2 can provide information about both afferent and efferent pathways. Additionally, the difference of LLR2 and HR latencies were analyzed in present study (LLR2 –HR interval). The duration of LLR2 – HR interval can reflect the central conduction time including central motor and sensorial pathways [12]. LLR2 –HR interval was the most useful parameter in our study (sensitivity, 37.9%; specificity, 100%). Poor correlation between LLR2 and SEP parameters can be explained by the effects of descending motor pathways on the generation of LLR. Significant amplitude differences in LLR2 were not observed in present study. The absence of amplitude differences between the patients and the controls can be explained by the absence of evident axonal involvement in our patients. The amplitude of HR and amplitude ratio of LLR2/HR also did not show any significant differences between the patients and the controls. Enhanced HR and absent or reduced LLR2 would reflect the dysfunction at suprasegmental descending projections due to any lesion along the corticospinal tractus [12]. Only a mild involvement of this pathway might not be sufficient to cause abnormalities in amplitude ratios of these potentials. The amplitude of LLR2 was also not changed in spite of prolonged latency of this component in some degenerative conditions of CNS [18]. Therefore, it seems that the reduction in the number of corticospinal fibers may cause a mild increase in LLR2 latency. The integrity and synchronization of afferent and efferent pathways of LLR can determine the amplitude of LLR2 component. The activity of basal ganglia structures and supplementary motor area can also affect this parameter [19]. Additionally, a stronger contraction of thenar muscle may enhance the amplitude of LLR2 without any alteration in its latency [14]. It seems that the latency of LLR2 is a more consistent parameter in the pathological conditions of CNS than the amplitude. Our results suggest that LLR can be considered as a useful electrophysiological investigation in evaluating subclinical neural involvement in patients with BD. Nevertheless, it can not be certainly claimed that abnormality of LLR is a symptom or an evidence of clinical neurological involvement due to Behcß et’s disease. Our patients with LLR abnormalities have not developed clinical neurological dysfunction for last 2 years. Prolonged neurological follow up of these patients may disclose the probable relationship between electrophysiological abnormalities and later development of evident neurological involvement.

71

References [1] Serdaroglu P. Behc¸et’s disease and the nervous system. J Neurol 1998;245:197 – 205. [2] Parisi L, Terracciano ME, Valente GO, Calandriello E, Accorinti M, Spadaro M. Pre-symptomatic neurological involvement in Behc¸et’s disease: the diagnostic role of magnetic transcranial stimulation. Electroencephalogr Clin Neurophysiol 1996;101:42 – 7. [3] Akman-Demir G, Serdaroglu P, Tasci B, Neuro-Behcet Study Group. Clinical patterns of neurological involvement in Behcet’s disease: evaluation of 200 patients. Brain 1999;122:2171 – 81. [4] Stigsby B, Bohlega S, McLean DR, Al-Kawi MZ. Transcranial magnetic sitmulation in Behcßet’s disease: a cross-sectional and longitudinal study with 44 patients comparing clinical, neuroradiological, somatosensory and brain-stem auditory evoked potential findings. Clin Neurophysiol 2000;111:1320 – 9. [5] Besana C, Comi G, Del Maschio A, et al. Electrophysiological and MRI evaluation of neurological involvement in Behc¸et’s disease. J Neurol Neurosurg Psychiatry 1989;52:749 – 54. [6] Rizzo PA, Valle E, Mollica MA, Sanarelli L, Pozzessere G. Multimodal evoked potentials in neuro-Behcß et’s disease: a longitudinal study of two cases. Acta Neurol Scand 1989;79:18 – 22. [7] Nakamura Y, Takahashi M, Ueyama K, et al. Magnetic resonance imaging and brain-stem auditory evoked potentials in neuro-Behcßet’s disease. J Neurol 1994;241:481 – 6. [8] Stigsby B, Bohlega S, al-Kawi MZ, al-Dalaan A, el-Ramahi K. Evoked potential findings in Behc¸et’s disease. Brain-stem auditory, visual and somatosensory evoked potentials in 44 patients. Electroencephalogr Clin Neurophysiol 1994;92:273 – 81. [9] Upton AR, McComas AJ, Sica RE. Potentiation of late responses evoked in muscles during effort. J Neurol Neurosurg Psychiatry 1971; 34:699 – 711. [10] Deuschl G, Ludolph A, Schenck E, Lu¨cking CH. The relations between long-latency reflexes in hand muscles, somatosensory evoked potentials and transcranial stimulation of motor tracts. Electroencephalogr Clin Neurophysiol 1989;74:425 – 30. [11] Michels R, Wessel K, Klo¨hn S, Kompf D. Long latency reflexes, somatosensory evoked potentials and transcranial magnetic stimulation: relation of the three methods in patients with multiple sclerosis. Electroencephalogr Clin Neurophysiol 1993;89:235 – 41. [12] Cruccu G, Deuschl G. The clinical use of brainstem reflexes and hand-muscle reflexes. Clin Neurophysiol 2000;111:371 – 87. [13] Deuschl G, Strahl K, Schenck E, Lu¨cking CH. The diagnostic significance of long-latency reflexes in multiple sclerosis. Electroencephalogr Clin Neurophysiol 1988;70:56 – 61. [14] Hallett M, Berardelli A, Delwaide P, et al. Central EMG and tests of motor control. Report of an IFCN committee. Electroencephalogr Clin Neurophysiol 1994;90:404 – 32. [15] International Study Group for Behcßet’s Disease. Criteria for diagnosis of Behcet’s disease. Lancet 1990;1:1078 – 80. [16] Morrissey SP, Miller DH, Hermaszewski R, Rudge P, MacManus DG, McDonald WI. Magnetic resonance imaging of the central nervous system in Behcßet’s disease. Eur Neurol 1993;33:287 – 93. [17] Stigsby B, Bohlega S, Delaan A, Al-Kawi Z. Magnetic brain stimulation: motor-evoked potential and central conduction studies in Behcß et’s disease (abstract). Acta Neurol Scand 1992;85(Suppl. 138):40. [18] Lee Y.C., Chen J.T., Liao K.K., Wu Z., Soong B.W.. Prolonged cortical relay time of long latency reflex and central motor conduction in patients with spinocerebellar ataxia type 6. Clin Neurophysiol 2003;114:458 – 62. [19] Naumann M, Reiners K. Long-latency reflexes of hand muscles in idiopathic focal dystonia and their modification by botulinum toxin. Brain 1997;120:409 – 16.