Epilepsiu, 18(3), 1977. Raven Press, New York

Phenobarbital and Phenytoin Effects on Somatosensory Evoked Potentials and Spontaneous EEG in Normal Cat Brain Bonnie J. Kaplan NPoropsytl?olofip Ltrborutory, Vetercins Administration Hospital, Wcst Hriven, Connr cticir t 065 16, (in d Dapmrtmen t of Neurology, Yu16, University School Qf Medicine, New HavcJn, Connecticut 06510

INTRODUCTION Although scalp-recorded evoked potentials are a valuable clinical tool in the detection and localization of neurological problems such as demyelinating disease (Halliday et al., 1976) and sensory impairment (Regan, 1972; Starr and Achor, 1975), their application to epilepsy has been of minimal value, partly because of the inability to separate the influence of anticonvulsant drugs from that of the underlying pathology. The present study was designed to examine the effects of chronically administered phenobarbital (PB) and phenytoin, two of the most widely used anticonvulsants, on the somatosensory evoked potential (SEP) and spontaneous E E G in normal unanesthetized feline brain. METHODS Surgery and A p p a r a t u s Adult cats (3-4.5 kg) were anesthetized and stereotactically implanted (Snider and Niemer, 1961) with electrodes as follows: left ventral posterolateral nucleus of thalamus (VPL: A9.0, L6.5, H+2.5), left centrum medianum (CM: A7.0, L2.5, H+0.5), right mesencephalic reticular formation (MRF: A2.0, L2.5, H - 1 . Q and right dorsal hippocampus (DH: A6.0, L3.0,

H+7.5). In addition, some cats had electrodes placed in right fastigial nucleus (Fast: P9.5, L1.5, H+l.O) and right dentate nucleus (Dent: P9.0, L6.5, H-0.25) of cerebellum. Stainless steel screws were threaded into the skull over primary somatosensory cortex (SI) of the left hemisphere, and in some cases over vermis (CB Med) and right lateral hemisphere (CB Lat) of the cerebellum. A screw over the right frontal sinus served as a reference electrode for all recordings. Leads were attached to a multi-pin connector, cemented t o the skull with dental acrylic. At the termination of the study, cats were sacrificed and perfused with 10% formalin. Electrode locations were confirmed histologically. SEPs were elicited with 500-psec stimuli, 0.5 or 1.0 mA above visible twitch threshold, delivered via percutaneous stimulating electrodes applied over the right ulnar nerve. A flexible recording cable connected the cat to an 8-channel Grass Model 7 polygraph, with half-amplitude low- and high-frequency cutoffs at 1 and 300 Hz, r e s p e c t i v e l y . S E P s c o n s i s t e d of 100-msec sweeps (beginning at 4.0 msec, 0.5 msec per point for the first 30 msec and 1 msec per point thereafter) averaged on a PDP-We computer. Spontaneous E E G activity was recorded on a Tandberg FM tape recorder. Design and Procedure

Two dosage levels of each drug were tested, Received February 14, 1977. Preliminary results were presented at the Annual Meeting of the American with the intention of producing serum levels apEEG Society, Dearborn, Michigan, September 29, proximating human therapeutic (PB: 10-25 1976. Key words: Phenohorhitul - Phrnytoin - Cuts - pgiml; phenytoin: 10-20 pg/ml) and toxlc (PB: > 25 pg/ml; phenytoin: > 20 pglml) ranges. For Evoked porrntiuls.

397

B . J . KAPLAN

3 98

PB, 3 mgikglday sufficed for the low dosage (yielding serum levels of 9-20 pg/ml), but in order to obtain toxic levels, the cats required 15 mgikgiday (yielding a range of 29-31 pgiml). For phenytoin, 5 and 8 mgikgiday resulted in serum ranges of 12-22 and 20-30 pgiml, respectively. The times of oral drug administration, data collection, and drawing blood were held constant across all experiments. Serum levels were d e t e r m i n e d by e n z y m e i m m u n o a s s a y micromethod. Cats lack many ofthe enzymes found in humans, and large amounts of the drugs were required to obtain the high-dosage serum levels. Debilitating side effects were of sufficient magnitude that only 3 cats were studied on the high dosage of each drug. Ten cats were studied on each drug: 7 at each low dosage and 3 at each high dosage. However, since not every cat had all electrodes implanted (in particular, the cerebellar placements), the number of animals contributing to each S E P analysis varied and is indicated on each figure. Data for each session consisted of six averages of 16 individual responses and 5 min of tape-recorded EEG. Five consecutive days of control data were collected, drug administration was begun, and when serum levels had reached the desired range (typically after 10 days for PB, using an initial loading dose, and after 7 days for phenytoin), 5 consecutive days of drug data were collected.

t ' s ( p < .05) may be expected for any given SEP analysis. This value may be compared with the number of t-tests which were significant in each experimental result (Figs. 1-3, 5). Tape-recorded E E G samples were submitted to power spectral analysis (Cooper et al., 1971). Taped EEG was prefiltered (3 d b points at 1.4 H z and 42 H z ) to prevent aliasing, fed to a PDP-12 computer, and eight 4-sec epochs were sampled to produce a single power density spectrum (0.25 to 3 1.75 H z with 0.25 H z resolution). Eight spectra from each electrode for each session were stored on magnetic tape, yielding approximately 80 spectra per experiment, per electrode location. T w o f r e q u e n c y b a n d s w e r e selected for analysis: ( 1 ) low-frequency activity and (2) the sensorimotor rhythm (SMR), which is a very clear normal rhythm in feline E E G (Brazier, 1963; Roth et al., 1967). Although there were some individual differences determined from PHENOBARBITAL. 3 rnqlkg CONTRCLDRUG

----

DH (0) N.7

Data Analysis

An overall S E P average was obtained for each group of control sessions (Control Average) and each group of drug sessions (Drug Average) in a single experiment. Point-to-point t-tests for each of the 127 time points of the 100-msec sweep were calculated on these matched pairs of Control and Drug Averages. When 127 comparisons are made on each SEP, a certain number will achieve significance by chance alone (e.g., R y a n , 1959). To obtain a n estimate of this number, t-tests were performed o n matched pairs of control SEPs: even-numbered sessions were compared to odd-numbered sessions, using the point-to-point procedures described above. A total of 95 sets of S E P analyses yielded a fre= 4.3, S D = 1.3) of the quency distribution number of significant f ' s when the only differences between SEPs were those due to chance factors. Therefore, as many as 5 false-positive

(x

CM (2)

N=7

0

I

nmec too

0

I00

InWC

FIG. 1. Effects of low-dosage PB on SEPs. Abbreviations: SI, primary somatosensory receiving cortex; VPL, ventral posterolateral nucleus; CM, centrum medianum; MRF, mesencephalic reticular formation; DH, dorsal hippocampus; CB Med, cerebellar vermis; CB Lat, lateral hemisphere of cerebell u m ; Dent, dentate nucleus of cerebellum. N , number of experiments for each electrode location. Number in parentheses is number of time points for which the difference between control and drug traces was significant. Shading indicates area of significant differences.

PHENOBARBITAL A N D PHEN YTOIN IN CATS PHENOBARBITAL. 15 mg/kg

3 99

PHENYTOIN,

5 mg/kg CONTROLDRUG----

DH N =(9) 3

%fi

- ___

CB Med(l5) N-4

C B Lot(0) N-7

Fost (I1 N: 3

Dent ( 3 ) N=4

CW9) U=7

0

100

W

C

0

100

mSeC

FIG. 2. Effects of high-dosage PB o n SEPs. Abbreviations as in Fig. 1. xx, Number of experiments was insufficient to do statistical analysis.

+ ]muv

0

visual inspection of each cat’s spectra, the typical frequency ranges were 3-7 cisec for lowfrequency activity and 12-16 cisec for SMR. Two characteristics of each of these two frequency bands were analyzed: (1) the amount of power in the band, integrated over the specified interval (Integrated Power), and (2) the frequency at which the largest amount of power in the specified range occurred (Peak Frequency). Matched pair t-tests were used to test for the significance of changes in all four variables: Integrated Power and Peak Frequency of each of the frequency ranges. F o r these two-tailed t-tests, as with the ones computed on the SEPs, 95% confidence limits were adopted.

RESULTS SEPs I . Eifects o j P B . Low dosage of PB had little effect on SEPs (Fig. l), with the exception of a slight change visible at MRF (exhibited in 5 of 7 cats). High dosage of PB had its most consistent effect at VPL, where amplitude attenuation was significant in all experiments (Fig. 2 ) .

msec 100

0

mwc 100

FIG. 3. Effects of low-dosage phenytoin on SEPs. “Fast,” fastigial nucleus of cerebellum. Other abbreviations as in Fig. 1. Note the change in the negative component occurring at approximately 55 msec.

2 . Effects of plienytoin. L o w dosage of phenytoin resulted in some general amplitude decrements across most electrodes (Fig. 3 ) . A specific effect was an amplitude change of a negative component with a latency of about 5 5 msec in DH, significant in 6 out of 7 animals, and visible at V P L and CM, although in only half the experiments. I n no cat was this change visible at SI . The amplitude change of the 55-msec component is most easily interpreted in a single cat (Fig. 4), where inter-animal latency differences do not mask the drug effect on a single component. I n intra-cat analyses, as seen in the example in Fig. 4, the SS-msec change was clearly an enhancement of a component which in control data was most prominent at DH. High dosage resulted in the enhancement at 55 msec at DH in all 3 animals, but the general amplitude decrements and latency increases

400

B. J . KAPLAN TABLE 1 . EEG power and frequency changes due to phenobarbital and phenytoin 3-7 cisec Slow activity

No. of experiments

Dosage

Phen

PB

7

10

7 10

7 10

10

12- 16 cisec SMR

Integrated Power

Peak Frequency

Integrated Power

$

f

+

4

4 4

f

4

Peak Frequency

SI

Low Low and high VPL Low Low and high CM Low Low and high

7

7 10

;1

MRF

Low Low and high DH Low Low and high CB Med Low Low and high CB Lat Low Low and high Fast Low Low and high Dent Low Low and high

6 9

7 10

7

7

10

10

4 5

4 6

7 10

7 10

3 5

0 0

4 6

3 4

c

4

6

++

c c t

t t

t

t

t

t t

$ t

$

--D: Phenobarbital change, in direction indicated, p G 0.05; +:phenytoin change, in direction indicated. p G 0.05; absence of any arrow indicates no significant change; t: there were no fastigial electrodes in any phenobarbital experiments. Abbreviations as in Fig. l .

were the most prominent features (Fig. 5). The sensory-specific late response of the SI potential, a positive component occurring between 75 and 100 msec in most cats (Goff e t al., 1966), exhibited the largest amplitude change produced by high-dosage phenytoin: it was significantly attenuated in all 3 cats. EEG 1. Effect3 of PB (Table I). There were no significant effects of low- o r high-dosage PB on slow acti\,ity Integrated Power at any electrode location. In four experiments with electrodes in Dent, that location exhibited a significant decrease (i.e., downward shift) in slo~suctivity Peak Frequency decrease = 0.3 cisec).

(x

The only effect of PB on S M R Integrated Power was a decrease at C B Med. S M R Peak Frequency decreased at both SI (3 = 0.7 cisec) and D H = 0.5 cisec), but at no other recording sites. 2. Ejyects o f p l r c n ~ r o i n(Table I). Phenytoin . produced significant increases in s l o ~ cic.tii.ity Integrated Power for both dosage levels at VPL, CM, DH, C B Lat and Dent. Interestingly, this effect was not visible at SI or at the locations along the fastigiobulbar pathway: i.e., C B Med and Fast. Although the reticular formation electrode was not ideally placed for examination of this pathway, the effect seen at MRF was consistent with that seen at C B Med and Fast. At each of two locations, VPL and DH, the Peak Fre-

(x

401

PHENOBARBITAL A N D PHENYTOIN IN CATS quency of the slon. cictivity decreased an average of 0.3 cisec. S M R Integrated Power increased at the same electrode locations as the slow frequency range. S M R Peak Frequency decreased in only three locations: SI (X decrease = 1.1 cisec), C B Lat 6 = 0.6 cisec), and Dent (X = 0.7 cisec).

DISCUSSION SEPs and EEG PB had little effect on the SEPs at low dosage, but at high dosage it produced general amplitude attenuation, particularly at VPL. E E G changes due to PB were few in number and exhibited no pattern of effects. The main effect of phenytoin on the SEPs was enhancement of the negative component which occurs at 55 msec and is most apparent at DH. The high dosage of phenytoin resulted in some S E P amplitude decrements, particularly of the SI sensory-specific late response. Phenytoin effects on the E E G power spectra tended t o be increases in Integrated Power and decreases in Peak Frequencies in both frequency ranges, and the changes occurred in a pattern: at VPL, CM, DH, C B Lat and Dent, but not at MRF, C B Med, o r Fast. The SEP effects reported here have several implications for human E P research. Both the PB and the phenytoin data suggest that modification of the S E P should be minimal at nontoxic dosages: amplitude decreases would be the predicted result of excessive dosages of both drugs. The phenytoin-enhanced 55-msec component may be related to an evoked potential component described in humans (Williamson et al., 1977).In certain epileptic patients an abnormally large negative S E P component at 55 msec has been recorded from electrodes over the contralateral somatosensory area. Correspondence between the human and feline negative components is speculative; however, if such a correspondence is valid, the present data suggest that the enhanced amplitude in humans is due to drug action. Although Williamson et al. suggested that the human negative component is cortically generated, the feline negative wave is most prominent at DH in control recordings (Fig. 4). A notable result in the measure of E E G slow activity was the lack of change following PB administration. This is in contrast to previous reports (Gehrmann and Killam, 1976; Schallek and Johnson, 1976) of increases in slow activity power in squirrel and rhesus monkeys given PB.

PHENYTOIN, 5 mg/kg CAT 301

- - - - - - - __

CONTRUDRUG

FIG. 4. Effects of low-dosage phenytoin in a single cat. Note enhancement of the negative component occurring at 55 msec. Abbreviations as in Fig. I .

The contradictory results may be due in part to differences in dosage and in methods of administration: in the monkeys the recordings were performed within hours after a single dosage of the drug, whereas the present research studied chronic oral administration. The Peak Frequency analysis of the E E G was of particular interest because of a recent report on human E E G by Dodrill and Wilkus (1976). Although phenytoin is not generally recognized as a cause of EEG slowing in humans (except as a correlate of toxicity), Dodrill and Wilkus found an inverse relationship between serum levels and Peak Frequency of occipital alpha-type activity (which they called “dominant posterior rhythm frequency”) in 90 patients whose only anticonvulsant medication was phenytoin. This report raises the question of whether E E G slowing is primarily the result of pathology or of anticonvulsant medication. If it is primarily an indicator of anticonvulsant medication, then Peak Frequency analysis might be used in conjunction with serum levels, for example, as an indicator of toxicity. I n the present study of anticonvulsant drugs in healthy cat brain, evidence is presented that the drugs themselves d o produce slowing of Peak Frequencies of normal EEG rhythms. Phenytoin affected the EEG recorded at the cortex (SI) more readily than did PB, although at high dosages both drugs did produce

B . J . KAPLAN

402 PHENYTOIN.

Anatomical Site of Action

8 rng/kg

CONT DRUG

0

, mk& '

FIG. 5. Effects of high-dosage phenytoin on SEPs. Abbreviations as in Fig. 1. x x , Number of experiments was insufficient to do statistical analysis.

decreased SMR Peak Frequencies recorded at the cortex. Based on the report of anti-seizure properties of SMR conditioning in cats (Sterman et al., 1969), there was some reason to believe that S M R would be sensitive t o anticonvulsant drugs. In the present study, phenytoin resulted in increased SMR Integrated Power a t VPL, DH, CM, C B Lat and Dent, while SMR Peak Frequency decreased at SI. C B Lat and Dent. These results are consistent with both the Sterman et al. report of increased SMR being associated with resistance to seizures, and also the report of Dodrill and Wilkus of phenytoinrelated slowing of Peak Frequencies.

Although - the anatomical site of action of p h e n y t o i n h a s not b e e n defined, electrophysiological a n d histological data have pointed to the sensitivity of cerebellar Purkinje cells (Julien and Halpern, 1971; Ghatak et al., 1976). In one report (Utterback, 1958) greater localizing evidence has been provided: Purkinje cells of the lateral hemispheres were affected to a greater extent than those of the vermis, implying that phenytoin action is primarily by the lateral hemisphere-to-dentate nucleus pathway, thus suggesting the greater inhibitory capacity of this pathway compared to the fastigiobulbar efferents. However, data from cerebellar stimulation research do not entirely substantiate this. In some reports, fastigiobulbar stimulation appears to be most effective in inhibiting seizures (Hablitz, 1976), while in others, dentatothalamic stimulation was more effective (Babb et al., 1974). Although the present study was not specifically designed to clarify which of these two cerebellar efferent systems is more sensitive to phenytoin o r more indicative of inhibitory activity, the E E G data reported here do indicate a selective phenytoin effect o n t h e dentatothalamic outflow. The E E G changes (increases in SMR Integrated Power and decreases in SMR Peak Frequency) were significant at CB Lat, Dent (to which CB Lat projects), CM and VPL . of thalamus. The placements which most nearly approached the fastigiobulbar outflow (CB Med, Fast, MRF) were consistently unchanged by phenytoin administration. It must be noted, however, that reciprocal connections between fastigial nucleus and VPL have been reported (Heath, 1972), thereby questioning the independence of the two pathways. SUMMARY Chronic oral administration of phenobarbital (PB) and phenytoin was studied in chronically implanted cats. The effects of two dosages (PB, 3 mgikg and 15 mglkg; phenytoin, 5 mg/kg and 8 mgikg) were analyzed with two physiological measures: somatosensory evoked potentials (SEPs) and power spectral analysis of the EEG. The low dosage of PB had no effect on SEPs except those recorded in the mesencephalic reticular formation; high dosage resulted in general S E P attenuation. P B had little effect on the

P H E N O B A R B I T A L A N D P H E N Y T O f N IN C A T S

EEG. Phenytoin at both dosages produced three main effects: ( 1 ) A negative component in the SEP which appears to be dominant in the dorsal hippocampus was significantly enhanced. (2) EEG changes occurred in the dentatothalamic pathway but not at those electrode locations which recorded fastigiobulbar activity. (3) The EEG changes which did occur were Peak Frequency shifts comparable to those previously reported in humans. The results are interpreted as demonstrating (1) minimal E E G and S E P modifications a t the cerebral cortical level due to nontoxic dosages of PB and phenytoin, (2) a selective phenytoin effect on the dentatothalamic outflow of the cerebellum and on the dorsal hippocampus, and (3) a potential usefulness of Peak Frequency analysis as a measure of E E G response to anticonvulsant drugs.

ACKNOWLEDGMENTS

I thank Joyce Cramer for the determination of anticonvulsant serum levels, and Joseph G . Jasiorkowski, Paul Kasper, and Philip Bergey for technical assistance. This study was supported by the Medical Research Service of the Veterans Administration, National Institute of Mental Health grant MH05286 and U.S. Public Health Service grant 2P50 N S 06208 NSPB. REFERENCES Babb T L , Mitchell AG Jr, and Crandall PH. Fastigiobulbar and dentatothalamic influences on hippocampal cobalt epilepsy in the cat. Electroencephulogr Clin Nruropliy.sio1 36: 141- 154, 1974. Brazier MAB. The problem of periodicity in the electroencephalogram: Studies in the cat. Electroencephulogr Clit7 Nerrrophysiol 15:287-298, 1963. Cooper R , Pocock PV, and Warren WJ. OLFFTI and FETCHFFT. DECUS P r o g r a m Lihrury N o . 1 2 - 6 3 , 1971. Dodrill CB and Wilkus RJ. Neuropsychological correlates of the electroencephalogram in epileptics: 11. The waking posterior rhythm and its interaction

403

with epileptiform activity. Epilepsiu 17:lOl- 109, 1976. Gehrmann JE and Killam K F Jr. Assessment of CNS drug activity in rhesus monkeys by analysis of the EEG. Fed P r o c 35:2258-2263, 1976. Ghatak NR, Santoso RA, and McKinney WM. Cerebellar degeneration following long-term phenytoin therapy. N e u r o h g y 26:818-820, 1976. Goff WR, Sterman MB, and Allison T. Cortical midline late response during sleep in the cat. Bruin R e s 1:311-314, 1966. Hablitz JJ. Intramuscular penicillin epilepsy in the cat: Effects of chronic cerebellar stimulation. E r p Neurol 50:505-514, 1976. Halliday AM, Halliday E, Kriss A, McDonald WI, and Mushin J. The pattern-evoked potential in compression of the anterior visual pathways. Bruin 991357-374, 1976. Heath RG. Physiologic basis of emotional expression: Evoked potential and mirror focus studies in rhesus monkeys. Biol Psychialr 5:15-31, 1972. Julien RM and Halpern LM. Diphenylhydantoin: Evidence for a central action. Life Sci 10575-582. 1971. Regan D. Evoked Potenliuls in Psychology, Sensory Physiology and Clinical Medicine. Chapman and Hall, London, 1972, 328 pp. Roth SR, Sterman MB, and Clemente CD. Comparison of EEG correlates of reinforcement, internal inhibition and sleep. EIec.troencephaloKr Clin Neurophysiol 23 :509- 520, 1967. Ryan TA. Multiple comparisons in psychological research. Psycho1 Bull 56:26-47, 1959. Schallek W a n d Johnson TC. Spectral density analysis of the effects of barbiturates and benzodiazepines on the electrocorticogram of the squirrel monkey. Arch I n t Pharmcicodyn Ther 223:301-310, 1976. Snider RS and Niemer WT. A Srereotaxic Atlus o , f t h e C a t Bruin. University of Chicago Press, Chicago, 1961. Starr A and Achor LJ. Auditory brain responses in neurological disease. A r c h Nertrol 32:761-768, 1975. Sterman MB, LoPresti RW, and Fairchild MD. Electroencephalographic and behavioral studies of monomethylhydrazine toxicity in the cat. Tech Rep A M R L - T R - 6 9 - 3 , Aerospace Medical Research Laboratory, Wright-Patterson AFB, 1969. Utterback RA. Parenchymatous cerebellar degeneration with Dilantin intoxication. J Neuropathol E x p Neirrol 17516-519, 1958. Williamson PD, Allison T, Goff WR, and Mattson RH. Evoked potential abnormalities in epilepsy: The E-Wave. EIectroi,ncepholoXr Clin Nerirophy.sio1 42:729-730, 1977.

Phenobarbital and phenytoin effects on somatosensory evoked potentials and spontaneous EEG in normal cat brain.

Epilepsiu, 18(3), 1977. Raven Press, New York Phenobarbital and Phenytoin Effects on Somatosensory Evoked Potentials and Spontaneous EEG in Normal Ca...
626KB Sizes 0 Downloads 0 Views