110

Electroencephalography and clinical Neurophy siology , 84 ( 1992) 110-114 © 1992 Elsevier Scientific Publishers Ireland, Ltd. 0168-5597/92/$05.00

EVOPOT 91040

Physiological functions of the ascending spinal tracts in HTLV-I-associated myelopathy (HAM) R. Kakigi a, y. Kuroda a, H. Takashima a, C. Endo a, R. Neshige a and H. Shibasaki b a Department of Internal Medicine, Saga Medical School, Saga (Japan), and b Department of Brain Pathophysiology, Kyoto University Faculty of Medicine, Kyoto (Japan) (Accepted for publication: 1 July 1991)

Summary Physiological functions of the spinothalamic tract (ST'F) and the posterior column (PC) were studied in 19 Japanese patients with HTLV-I-associated myelopathy (HAM). The former was evaluated by pain-related somatosensory evoked potentials (SEPs) following CO 2 laser stimulation, and the latter by conventional SEPs following electrical stimulation. Conduction velocity of STT and PC was significantly decreased in 9 and 14 patients, respectively, and more than half of them showed no clinical impairment of sensations (subclinical abnormality).

Key words: HTLV-I-associated myelopathy (HAM); Somatosensory evoked potentials (SEPs); Pain SEPs; Spinothalamic tract; Posterior column

One of the biggest recent topics in neurological science is an establishment of a new clinical entity termed HTLV-I-associated myelopathy or tropical spastic paraparesis (HAM/TSP), which is a unique chronic myelopathy caused by human T-lymphotropic virus type-I (HTLV-I) (Gessain et al. 1985; Osame et al. 1987; Shibasaki et al. 1988; Cruickshank et al. 1989). Its main symptoms are spastic paraparesis and neurogenic bladder. Though an impairment of pain-temperature sensation is absent in most HAM cases, autopsy findings indicated mild demyelination in the spinothalamic tract (STT) (Iwasaki 1990). It is now established that somatosensory evoked potentials (SEPs) induced by a low power and long wave length CO 2 laser beam (pain SEPs) are useful to investigate physiological functions of small myelinated fibers (A~) of the peripheral nerve and ST-F (Carmon et al. 1976; Treede et al. 1988; Kakigi et al. 1989, 1990, 1991a, b, c; Arendt-Nielsen 1990; Kakigi and Shibasaki 1991). The object of this paper is, therefore, to report the results of pain SEPs and to evaluate physiological function of ascending tracts of the spinal cord in a number of cases with HAM. Materials and methods

Diagnostic criteria of HAM proposed by Osame et al. (1987) were adopted. It is briefly summarized as Currespondence to: Ryusuke Kakigi, M.D., Department of Internal Medicine, Saga Medical School, Nabeshima, Saga City 849 (Japan). Tel.: 952-31-6511 (ext. 2360, 2371, 2372); Fax: 952-30-6629.

follows: (1) chronic progressive myelopathy of adult onset, (2) high antibody titers to HTLV-I in serum and cerebrospinal fluid (CSF), (3) predominantly symmetric upper motor neuron disorder and bladder dysfunctions, with mild sensory impairments, and (4) other neurological disorders, particularly compressive myelopathy, can be ruled out. Nineteen patients, 12 females and 7 males, who were diagnosed as HAM based on the above criteria, were studied. All of them were born and lived in Kyushu island in Japan. Preliminary results in 12 patients were reported in our previous paper (Kakigi et al. 1988). Age of the patients ranged from 32 to 76 years (mean + S.D. = 52.2 + 12.5) and the duration of illness ranged from 2 to 40 years (12.9 + 11.3). Their height ranged from 138 to 172 cm (154.1 + 8.9). Antibody titer to HTLV-I was assayed by the passive particle agglutination method. Titer in serum and cerebrospinal fluid ranged from 128 to 65,536 (10,017.8 + 19,974.4) and from 2 to 2048 (244.6 + 472.2), respectively. Sensory functions of the upper limbs were normal in all cases. In the lower limbs, tactile and joint sensations were normal in all cases, but vibration sensation was impaired to various degrees in 10 out of 19 patients. Pain-temperature sensation was intact in all patients except one. Pain SEPs were recorded by the special CO 2 laser stimulator (Nippon Infrared Industries Co. Ltd., Kawasaki, Japan). Its maximum power was 23 W, but an attenuator limited the output to approximately 12.5 W for safety reasons. Laser wave length was 10.6/xm, and the diameter of the irradiation beam was approxi-

PHYSIOLOGICALSPINAL FUNCTIONS IN HAM mately 2 mm. Stimulus duration was 10 msec, and stimulus interval was 3 sec. Stimulus intensity was approximately 18 m J / m m 2 which elicited sharp pain that normal subjects described as tolerable 'pin-pricklike.' Laser beam was applied to the dorsum of the right hand and foot in all patients. Eyes were protected by putting a swimming-goggle over the face. As peripheral and spinal responses of pain SEPs could not be clearly recorded (Kakigi et al. 1989), only scalp responses were investigated. As pain SEPs were most prominent around the vertex (Kakigi et al. 1989), exploring electrodes were placed at Cz (international 10-20 system), midway between Fz and Cz (FCz), and between Cz and Pz (CPz) where Fz and Pz were sites of the international 10-20 system. Linked earlobes (A1A2) were used as the reference. Amplifier frequency response was 0.5-30 Hz ( - 3 dB). Ten to 20 responses were averaged in 1 session, and at least 4 sessions were done in each subject. After confirming a consistency of wave forms, all averages were added together and the group-averaged wave forms recorded at Cz were analyzed. As the positive components, P340 in hand-stimulated SEPs and P400 in foot-stimulated SEPs, were largest and more stable compared with the preceding negative peaks among normal individuals (Kakigi et al. 1989; Kakigi and Shibasaki 1991), these positive components were mainly analyzed in this paper. When the potential showed multiple peaks, the mean latency of them was adopted. Analysis time was 1.5 sec after stimulation. The sampling rate was 2.9 msec. Conventional SEPs following electrical stimulation of the right median and posterior tibial nerve were also recorded in all patients. Recording methods were principally based on the recommended guidelines reported by American Electroencephalographic Society (1986). The frequency response of the amplifier was 15-3000 Hz, and the sampling rate was 0.08 msec and 0.16 msec for median and posterior tibial nerve SEPs, respectively. Relative positivity at grid 1 resulted in a downward deflection in all records. Amplitude was measured from the preceding peak of the opposite polarity. Approximate conduction velocity (CV) of STT and the posterior column (PC) was measured in all patients. Its detailed methods for calculation were described in our previous paper (Kakigi and Shibasaki 1991). Fifty volunteers, aged 17-75, served as normal controls. Their heights ranged from 143 to 178 cm (mean + S.D. = 156.3 + 13.8). We have previously reported parts of their SEP results (Kakigi 1987, 1989; Kakigi et al. 1989, 1990, 1991a, b, c; Kakigi and Shibasaki 1991). Mean plus 2.5 S.D. was adopted as the upper limit of the normal range. As for the treatment, all patients except one were given corticosteroid a n d / o r interferon a n d / o r a large

111 dose of y-globulin (Kuroda et al. 1991). Pain SEPs and conventional SEPs following electrical stimulation were examined before treatment in all patients and were done after treatment in 13 and 16 patients, respectively. All patients and normal volunteers gave informed consent. Results

(1) Pain SEPs The latency of P340 potential following hand stimulation was within normal range in all patients. The P400 potential following foot stimulation was significantly prolonged in 4 patients. CV of STI" was significantly decreased in 9 patients, and 8 among them showed no impairment of pain-temperature sensation (subclinical abnormality; Fig. 1). Mean CV of STT (6.4 + 3.8 m / s e c ) was significantly smaller than normal controls (9.6 + 2.0 m / s e c ) by Student's t test (P < 0.01). (2) SEPs following electrical stimulation As for SEPs following median nerve stimulation, N9 recorded at Erb's point was of normal latency and amplitude in all patients. Both N13 recorded at the cervical vertebra and N20 recorded at the scalp were significantly prolonged in 3 patients. The interpeak latency between N9 and N13 was significantly increased in 9 patients, but the interpeak latency between N13 and N20 (so-called central conduction time) was within normal range in all patients. There was no significant correlation between clinical profiles and results of median nerve SEPs. As for SEPs following posterior tibial nerve stimulation, N19 recorded at the 12th thoracic vertebra and N27 at the 2nd cervical vertebra were recorded in 17 and 11 patients, respectively. In those patients, peak latency of N19 was normal in all, but N27 was significantly prolonged in 9 patients. As both N19 and N27 could not be recorded even in some normal subjects, it is difficult to judge the abnormality in patients whose N19 a n d / o r N27 were not consistently identified. The first cortical positive potential, P37, recorded at Cz', was significantly prolonged in 14 out of all 19 patients. Indirect measurement of CV of PC was significantly decreased in 14 out of 19 patients, and vibration sensation was judged to be normal in 7 out of them. By comparing with CV of PC in normal controls (53.1 + 7.0 m/sec), the mean value of HAM patients (29.9 + 8.7 m / s e c ) was significantly smaller by Student's t test (P > 0.01). (3) Correlation among results of pain SEPs, median nerve SEPs and posterior tibial nerve SEPs Correlation coefficients (r) among results of pair~

R. KAKIGI ET AL.

112

A

Patient 10

B

Patient 12 Pain SEPs

Pain SEPs C z - A 1 A2

C z - A 1A2 Hand stirn. ~

Hand Stim. ~

[5~JV

P340

[ 5lJV

P340

Foot stim. ~

Foot Stim. ~ P400

P400 I

I

I

I

I

I

I

I

I

I

sec

0

I

I

I

I

- -

C4'-Fz

N13 - '

~

I

I

I

1.5

N20

C 4 ' - F z ~--"

I

Median nerve SEPs

N20

'

I

sec

Median nerve SEPs

Cv2-Fz ~

I

0

1.5

-

~

P25 ' -

-

-

.

/

" ~ _ ~ d - - - ' - ' ~

N13

"

N9

12uV

r°-FzT'

_

~ P25

V

Cv2-Fz

- -""--

12uV

N9 Erb-Fz I

I

I

I

3

I

I

I

I

I

msec

I

I

I

3

I

I

I

I

I

I

I

msec

43

43 Posterior tibial nerve SEPs

Posterior tibial nerve SEPs Cz'-Fz

Cz'-Fz

~

1 P37

P37

N27

Cv2-Fz

~

N27

[ luV

I

Cv2-Fz ~ N19 Th12-Th12' \A - v

[luV ~

v I

5

I

I

~_A/~ _ v - - L._.L/-~

k/ I

I

msec

I

I

I

t

I

85

5

msec

85

Fig. 1. Pain SEPs, median nerve SEPs and posterior tibial nerve SEPs in patient 10 (A) and patient 12 (B). Th12: twelfth thoracic vertebra; Thl2': 4 cm rostral to Thl2; Cv2:second cervical vertebra; Cz': 2 cm behind Cz; Erb: Erb's point (supraclavicular site); C4': 2 cm behind C4. Fz, Cz and C4 are sites of the international 10-20 system. In patient 10, peak latency of P340 is normal but that of P400 is at the upper limit of the normal range. Conduction velocityof the spinothalamic tract (CV of STT) is significantly small (3.7 m/see). Peak latencies of both P340 and P400 and CV of STT are within the normal range in patient 12. Peak latency of N9 is within the normal range in both patients, but the interpeak latency between N9 and N13 is significantlyprolonged (5.3 msec) in patient 10. In both patients, peak latencyof N19 is within the normal range, but that of N27 as well as P37 is significantlyprolonged. Conduction velocities of the posterior column (CV of PC) are 30.6 and 25.2 m/see in patient 10 and 12, respectively.

SEPs, median nerve SEPs and posterior tibial nerve SEPs were calculated. A significant high correlation ( P < 0.01) was found between the CVs of STT and PC (r = 0.684, Fig. 2), between N20 latency and P37 latency (r = 0.789) and between P37 latency and CV of PC (r = 0.799).

lin or some of them, but sensory impairments were not clearly changed in any patient. SEP results were not clearly changed between before and after the treatment in any case.

Discussion

(4) Treatment effects Motor functions a n d / o r neurogenic bladder were improved to various degrees in 13 out of 18 patients who were given corticosteroid, interferon and 7-globu-

This is the first systematic and thorough study of pain SEPs in H A M / T S P . In the present study, pain SEP following the hand stimulation (P340) was of

PHYSIOLOGICAL SPINAL FUNCTIONS IN HAM

Y=O.354657X-4.21936 r:O.684(P