- Email: [email protected]
Memory alterations during acute high-intensity vagus nerve stimulation
Epilepsy Research 47 (2001) 37 – 42 www.elsevier.com/locate/epilepsyres
Memory alterations during acute high-intensity vagus nerve stimulation C. Helmstaedter *, Christian Hoppe, Christian E. Elger Department of Epileptology, Uni6ersity of Bonn, Sigmund-Freud-Straße 25, 53105 Bonn, Germany Received 5 May 2001; received in revised form 28 June 2001; accepted 1 July 2001
Abstract Left cervical vagus nerve stimulation (VNS) is an accepted add-on treatment for pharmacoresistant epilepsy. However, it also allows the investigation of the effects of peripheral nerve stimulation on central nervous functions. The impact of 4.5 min high intensity VNS ( \ 1 mA) on material-specific memory and decision times was evaluated in an experimental ‘box car’ design in 11 patients with pharmacoresistant epilepsy. Results indicate reversible deterioration of figural but not verbal memory and a trend of accelerated decision times during VNS. Thus, further support of cognitive effects of VNS is provided. There are indications of a major projection of VNS to activating brain structures of and the right hemisphere. Significant cognitive side effects in clinical application are unlikely because of the reversibility of the effect and differences between experimental and therapeutic stimulation conditions. However, since the effectors and the direction of the cognitive effects of VNS seem to depend strongly on stimulation conditions, we recommend future experimental research covering a larger range of stimulation conditions. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Epilepsy; Vagus nerve stimulation; Memory; Attention; Lateralized effect
The vagus nerves contribute to a series of vital behavioral systems such as digestive behavior (Rogers et al., 1996), nociception (Randich and Gebhart, 1992), reproduction (Uvnas Moberg, 1994), and emotion regulation (Porges et al., 1994). They consist of about 80% afferent fibers and do not contain any pain-associated fibers (Agostoni et al., 1957). Afferent vagal neurotransmission has widespread target regions throughout * Corresponding author. Tel.: +49-228-287-6108; fax: +49-228-287-6294. E-mail address: [email protected] (C. Helmstaedter).
the brain including brain stem (e.g. nucleus of the solitary tract), diencephalic (e.g. thalamus), and cortical regions (e.g. insular cortex; Rutecki, 1990). This obviously close integration of the vagus nerve and the CNS suggests that vagal afferent input may affect cognitive brain functions. Following early animal studies, which revealed specific changes in spontaneous electrical brain activity as well as considerable attenuation of epileptic activity due to vagus nerve stimulation (VNS; Zabara, 1992), the first epilepsy patient was provided with a completely implantable
0920-1211/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 0 - 1 2 1 1 ( 0 1 ) 0 0 2 9 1 - 1
38
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
device for continuous intermittent left cervical VNS in 1988. To date, approximately 14 000 patients have been treated with VNS. While there is an agreement that VNS is effective in seizure control (Ben Menachem et al., 1994; George et al., 1994; Handforth et al., 1998), the cognitive effects of VNS have not yet been studied extensively. In animal experiments electrical vagal stimulation affected learning performance and behavior (Clark et al., 1995, 1998). Clark et al. (1999) were the first to report word-recognition memory enhancement due to acute VNS in human subjects. According to their findings in animal studies, recognition enhancement was most expressed at moderate stimulation output currents (0.5 mA) whereas no effects were obtained with very low (0.25 mA) or higher currents (0.75– 1.50 mA). Chronic VNS application appears not to lead to changes in cognitive functioning (Dodrill, 1997; Dodrill and George, 2001; Hoppe et al., 2001). Motivated by the findings of Clark et al., we now feel encouraged to report experimental data on the cognitive effects of acute VNS phases collected from 11 epilepsy patients who participated in our first VNS series (EO3 study; Ben Menachem et al., 1994). The study was set up to evaluate whether any cognitive side-effects can be expected from long-term VNS treatment within experimental conditions. Memory and decision times were tested in a box car design and performance of a group of healthy subjects who had repeated testing without any intervention served as control.
1. Methods
1.1. Patients Inclusion criteria were, adult person; pharmakoresistant epilepsy; and at least four partialonset seizures per month for 3 consecutive months. Eleven patients agreed to participate in the experimental study and gave informed consent (age, 18–42 years; mean, 31.9 years). The evaluation was carried out 5– 17 months (mean, 10.3 months) after implantation.
1.2. Computerized memory assessment Memory was assessed by two computerized memory tests, which require subsequent list learning of 15 words or nine designs in three learning trial each (Ju¨ nemann et al., 1992). Each list exposure was followed by immediate recognition of the target items out of alternatives (yes/no response). Words and designs in learning and recognition phases were presented sequentially. After completion of learning words and figures, final recognition of both materials without previous presentation was required. The wordlist of 15 nouns and the 30 distractor words were balanced for frequency of occurrence, the nine designs and 18 distractor designs (4×4 checkerboards with different patterns) were analogue to the Metric Figure test from Warrington and James (1967) and have a high demand on visual spatial memory (Warrington and James, 1967). The tests have been designed for experimental and clinical conditions which require repeated testing after short intervals and have been shown to be differentially sensitive to left/right temporal dysfunction in preand post-ictal follow-up testing in patients with lateralized temporal and frontal lobe seizures (Helmstaedter et al., 1994). For each test-session new lists were randomly generated out of a large item pool so that no patient gets the same testitems a second time and that tests of comparable difficulty were provided. Test duration is about 4 –5 min. Test analysis in healthy subjects and patients with lateralized epilepsies indicated that a summary score (sum of correct answers minus errors) was the most representative and discriminative measure of the test. Furthermore, a summary score of all decision times was calculated for control of vigilance and psychomotor speed. Parallel- or retest-reliabilities in healthy subjects and patients with left or right temporal lobe epilepsy were satisfactory with rtt-values ranging between 0.70 and 0.82 (Ju¨ nemann et al., 1992).
1.3. Experimental study design The experimental study design was a box-car design with ‘A1’ standing for baseline testing without VNS stimulation, ‘B’ standing for testing
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
during VNS stimulation and ‘A2’ standing for a third testing again without stimulation in the patient group. For the evaluation of performance courses during the three test trials normative data were available from repeated testing of 20 healthy, but unmatched subjects with average intelligence (age 19–45 years). The repeated tests in the control subjects who of course had no stimulation trial are labeled ‘A1’, ‘A2’, and ‘A3’.
1.4. Vagus ner6e stimulation VNS was performed using the implanted NeuroCybernetic Prosthesis™ System (Cyberonics, Inc.), which had been implanted for add-on therapy of phamacoresistant seizures. Vagal nerve stimulation during the experimental stimulation trial ‘B’ was delivered with the same output currents which were applied clinically at that time (output current, median 1.75 mA, range 1.00– 2.50 mA; pulse width, 500 ms; pulse frequency, 30 Hz). However, for the purpose of this study, stimulation duration was set from 30 s to a maximum length of 4.5 min to cover the time of memory testing in trial B. Thus, stimulation was active during item presentation and recognition in all learning trials. No patient complaint about prolonged stimulation during the second trial. For the baseline trials ‘A’ the stimulation was switched off using the magnet which provides external control over the device.
39
‘group’ (F= 6.6, P= 0.001) indicating that patients performed more poorly than controls across all measures (univariate tests, verbal memory F= 3.9, P= 0.057, figural memory F= 5.8, P = 0.02, verbal decision times F= 28.7, P = 0.000, figural decision times F= 0.91, P = 0.005). The main effect of the factor ‘test repetition’ was also statistically significant (F= 8.0, P= 0.02), indicating significantly decreasing decision times particularly in the verbal memory test and a reversible deterioration of figural memory performance (univariate testing, verbal decision times F= 6.2, P= 0.004, figural memory F= 2.7, P= 0.07). The significant main effect of ‘test repetition’ on figural memory could be explained fully by a highly significant interaction effect of ‘group Xtest-repetition’ on figural memory (F= 7.8, P = 0.009). Post hoc analysis confirmed that in patients only figural memory deteriorated from trial A1 to trial B (T=2.7, P= 0.02) and recovered from trial B to A2 (T= − 2.7, P= 0.02; Fig. 1). No such effect was seen for verbal memory (Fig. 2). Although the interaction effect group by test repetition did not become significant (F= 2.1, P= 0.13) for decision times, it should be noted
1.5. Statistical analysis Verbal/figural memory and decision times were analyzed by means of a repeated measurement multivariate analysis of variance with the three ‘test trials’ as the within-group factor and ‘group’ (patient/control) as the between-group factor. Post hoc analysis was carried out using T-tests for dependent samples with two-tailed significance.
2. Results Repeated measurement analysis of variance (ANOVA) showed a significant main effect of
Fig. 1. Figural recognition before, during and after VNS in patients as compared with three consecutive trials in healthy subjects (group means and S.E.M.).
40
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
Fig. 2. Verbal recognition before, during and after VNS in patients as compared with three consecutive trials in healthy subjects (group means and S.E.M.).
that without consideration of material-specificity, decision times in patients tended to be accelerated during the stimulation B as compared with both baseline tests (T-tests for dependent measures, A1 –B, t= 1.9, P =0.08; B– A2, t = −1.8, P = 0.09). In healthy subjects, decision times significantly accelerated from test A1 to A2 and remained stable from test A2 to A3 (T-tests for dependent measures, A1 – A2, t = 2.2, P = 0.04; A1 –A3, t=2.1, P= 0.05; Fig. 3).
1977), the effects of VNS on memory performance and response latency appear to be independent from each other. Since verbal recognition performance remained unaffected, it is unlikely that unspecific distraction or irritation due to the prolonged stimulation caused the results. In fact, the trend of a reversible acceleration of decision times during VNS could rather be discussed as an indicator of positive VNS effects on attention. This would be in line with many patient reports of increased arousal and improved attention during VNS long-term treatment (George et al., 1994). As a practical consequence of the findings one may ask whether the negative effect on figural memory has any consequences for the patients’ capabilities in every day lives. We feel that this is very unlikely. The obtained effects were completely reversible, in the clinical application stimulation conditions (intensity and duration) are different, and in every day life, VNS is not timelocked to a certain cognitive process. The suggestion that there is no recognizable effect in every day functioning would be consistent with previous reports which did not find any consistent change in cognitive performance with chronic VNS treatment (Dodrill, 1997; Dodrill and George, 2001; Hoppe et al., in press). However, more experimen-
3. Discussion In contrast to previous reports, our experimental data reveals a negative effect of acute VNS on memory performance. In an experimental baseline –treatment–baseline study a fully reversible deterioration of figural recognition memory was observed when vagal stimulation was delivered during the task. No such effects could be observed on verbal recognition memory or on decision times in the respective test. In patients, yes/no-decision times tended to be accelerated during VNS and returned to baseline levels after stimulation. Since poorer memory performance normally goes along with longer decision latencies (Koppell,
Fig. 3. Decision latencies during verbal and figural memory performance before, during and after VNS in patients as compared with three consecutive trials in healthy controls (group means and S.E.M.).
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
tal research regarding the effects of different types of VNS stimulation on cognition is recommended, in order to fully exclude negative effects or even in order to show positive cognition enhancing effects. Our results appear to be in contrast to the findings of Clark et al. (1999) who demonstrated verbal memory enhancement due to acute VNS. However, the two studies completely differ with respect to the study design, i.e. stimulation conditions and output currents. In this study, VNS was delivered during both learning and immediate recognition and could be suggested to interfere with different subroutines of memory processing (e.g. perception, encoding, recognition). In the study of Clark et al., VNS was delivered only during retention intervals and thus selectively affected memory consolidation. Furthermore, in our study all patients had output currents greater than 1.00 mA, whereas the patients of Clark et al. had currents 51.00 mA. The ‘boost’ effect, which is observed with low currents, may invert to a suppression effect with higher output currents (1.25–1.75 mA). The findings could thus be complementary rather than contradictory. Surprisingly, left cervical VNS selectively affected figural learning. The figural memory task applied in this study has previously been shown to be selectively sensitive for right temporal memory dysfunction immediately after right temporal seizures (Helmstaedter et al., 1994). Thus, from a neuropsychological point of view one may speculate on a lateralized cerebral effect of VNS on right hemisphere function. Although findings from neuroimaging studies are not equivocal (Garnett et al., 1992; Henry et al., 1999), the study of Ko et al. (1999) who demonstrated decreased right temporal blood flow during VNS may give some support to this assumption. Further evidence is provided by a study of Henry et al. (1998) who found increased right thalamic activation during VNS. Interestingly, there is also some evidence that seizure response rates for VNS treatment are higher in patients with a right focus as compared to patients with a left focus (Ben Menachem et al., 1999).
41
4. Conclusion Our findings provide evidence that acute phases of VNS can affect cognitive functioning. From a neuropsychological point of view, left-hand cervical VNS may preferentially project into the contralateral hemisphere. Since direction and extent of the VNS effect on cognition appears to depend on the stimulation conditions, future experimental studies should cover the whole spectrum of output currents (0.25–2.00 mA) and should separately address different brain functions.
Acknowledgements The work of C. Hoppe was supported by a grant of Cyberonics (Webster/TX, Inc.). Additional support was provided by the Deutsche Forschungsgemeinschaft (DFG EL 122/6-2).
References Agostoni, E., Chinnock, J.E., DeBurgh Daly, M., Murray, J.G., 1957. Functional and histological studies of the vagus nerve and its branches to the heart lungs and abdominal viscera in the cat. J. Physiol. 135, 182 – 205. Ben Menachem, E., Manon Espaillat, R., Ristanovic, R., et al., 1994. Vagus nerve stimulation for treatment of partial seizures: 1. A controlled study of effect on seizures. First International Vagus Nerve Stimulation Study Group. Epilepsia 35, 616 – 626. Ben Menachem, E., Hellstro¨ m, K., Waldton, C., Augustinsson, L.E., 1999. Evaluation of refractory epilepsy treated with vagus nerve stimulation for 5 years. Neurology 52, 1265 – 1267. Clark, K.B., Krahl, S.E., Smith, D.C., Jensen, R.A., 1995. Post-training unilateral vagal stimulation enhances retention performance in the rat. Neurobiol. Learn. Mem. 63, 213 – 216. Clark, K.B., Smith, D.C., Hassert, D.L., Browning, R.A., Naritoku, D.K., Jensen, R.A., 1998. Posttraining electrical stimulation of vagal afferents with concomitant vagal efferent inactivation enhances memory storage processes in the rat. Neurobiol. Learn. Mem. 70, 364 – 373. Clark, K.B., Naritoku, D., Smith, D.C., Browning, R.A., Jensen, R.A., 1999. Enhanced recognition memory following vagus nerve stimulation in human subjects. Nat. Neurosci. 2, 94 – 98. Dodrill, C.B., 1997. Cognitive and quality of life changes with VNS therapy. Results from a randomized controlled trial,
42
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
abstract. Current results and perspectives on vagus nerve stimulation for the treatment of epilepsy. Cyberonics, Houston/TX. Dodrill, C.B., George, M.L., 2001. Effects of vagal nerve stimulation on cognition and quality of life in epilepsy. Epilepsy Behav. 2, 46 –53. Garnett, E.S., Nahmias, C., Scheffel, A., Firnau, G., Upton, A.R.M., 1992. Regional cerebral blood flow in man manipulated by direct vagal stimulation. Pacing Clin. Electrophysiol. 15, 1579 – 1580. George, R., Salinsky, M., Kuzniecky, R., et al., 1994. Vagus nerve stimulation for treatment of partial seizures: 3. Longterm follow-up on first 67 patients exiting a controlled study. First International Vagus Nerve Stimulation Study Group. Epilepsia 35, 637 –643. Handforth, A., DeGiorgio, C.M., Schachter, S.C., 1998. Vagus nerve stimulation therapy for partial-onset seizures: a randomized active-control trial. Neurology 51, 48 – 55. Helmstaedter, C., Elger, C.E., Lendt, M., 1994. Postictal courses of cognitive deficits in focal epilepsies. Epilepsia 35, 1073 –1078. Henry, T.R., Bakay, R.A., Votaw, J.R., et al., 1998. Brain blood flow alterations induced by therapeutic vagus nerve stimulation in partial epilepsy: I. Acute effects at high and low levels of stimulation. Epilepsia 39, 983 –990. Henry, T.R., Votaw, J.R., Pennell, P.B., Epstein, C.M., Bakay, R.A., Faber, T.L., Grafton, S.T., Hoffman, J.M., 1999. Acute blood flow changes and efficacy of vagus nerve stimulation in partial epilepsy. Neurology 52, 1166 – 1173. Hoppe, C., Helmstaedter, C., Scherrmann, J., Elger, C.E., 2001. No evidence for cognitive side effects after 6 months of vagus nerve stimulation in epilepsy patients. Epilepsy Behav., in press. Ju¨ nemann, H., Helmstaedter, C., Elger, C.E., 1992. Die Mo¨ glichkeiten des Einsatzes computerisierter Geda¨ chtnistests in der pra¨ chirurgischen Epilepsiediagnostik [use of
computerized memory tests in presurgical evaluation of epilepsy]. In: Scheffner, D. (Ed.), Epilepsie 91. EinhornPresse Verlag, Reinbeck, pp. 449 – 452. Ko, D.Y., Grafton, S.C., Heck, C.N., Smith, T.D., DeGiorgio, C.M., 1999. Prolonged and progressive cerebral blood flow activation and deactivation with vagus nerve stimulation: more lateralization to contralateral structures. Poster presented at the 54th AES Meeting, December 6, 1999, Orlando, FL (2.223). Epilepsia 40 (Suppl. 7), 139. Koppell, S., 1977. Decision latencies in recognition memory: a signal detection theory analysis. J. Exp. Psychol. Hum. Learn. Mem. 3, 445 – 457. Porges, S.W., Doussard-Roosevelt, J.A., Maiti, A.K., 1994. Vagal tone and the physiological regulation of emotion. Monogr. Soc. Res. Child. Dev. 59, 167 – 186. Randich, A., Gebhart, G.F., 1992. Vagal afferent modulation of nociception. Brain Res. Rev. 17, 77 – 99. Rogers, R.C., McTigue, D.M., Hermann, G.E., 1996. Vagal control of digestion: modulation by central neural and peripheral endocrine factors. Neurosci. Biobehav. Rev. 20, 57 – 66. Rutecki, P., 1990. Anatomical physiological and theoretical basis for the antiepileptic effect of vagus nerve stimulation. Epilepsia 31 (Suppl. 2), S1 – S6. Uvnas Moberg, K., 1994. Role of efferent and afferent vagal nerve activity during reproduction: integrating function of oxytocin on metabolism and behaviour (Proceedings of the International Conference on Hormones Brain and Behaviour, 1993 Tours/France). Psychoneuroendocrinology 19, 687 – 695. Warrington, E.K., James, M., 1967. Disorders of visual perception in patients with localized cerebral lesions. Neuropsychologia 5, 253 – 266. Zabara, J., 1992. Inhibition of experimental seizures in canines by repetitive vagal stimulation. Epilepsia 33, 1005 – 1012.
Memory alterations during acute high-intensity vagus nerve stimulation C. Helmstaedter *, Christian Hoppe, Christian E. Elger Department of Epileptology, Uni6ersity of Bonn, Sigmund-Freud-Straße 25, 53105 Bonn, Germany Received 5 May 2001; received in revised form 28 June 2001; accepted 1 July 2001
Abstract Left cervical vagus nerve stimulation (VNS) is an accepted add-on treatment for pharmacoresistant epilepsy. However, it also allows the investigation of the effects of peripheral nerve stimulation on central nervous functions. The impact of 4.5 min high intensity VNS ( \ 1 mA) on material-specific memory and decision times was evaluated in an experimental ‘box car’ design in 11 patients with pharmacoresistant epilepsy. Results indicate reversible deterioration of figural but not verbal memory and a trend of accelerated decision times during VNS. Thus, further support of cognitive effects of VNS is provided. There are indications of a major projection of VNS to activating brain structures of and the right hemisphere. Significant cognitive side effects in clinical application are unlikely because of the reversibility of the effect and differences between experimental and therapeutic stimulation conditions. However, since the effectors and the direction of the cognitive effects of VNS seem to depend strongly on stimulation conditions, we recommend future experimental research covering a larger range of stimulation conditions. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Epilepsy; Vagus nerve stimulation; Memory; Attention; Lateralized effect
The vagus nerves contribute to a series of vital behavioral systems such as digestive behavior (Rogers et al., 1996), nociception (Randich and Gebhart, 1992), reproduction (Uvnas Moberg, 1994), and emotion regulation (Porges et al., 1994). They consist of about 80% afferent fibers and do not contain any pain-associated fibers (Agostoni et al., 1957). Afferent vagal neurotransmission has widespread target regions throughout * Corresponding author. Tel.: +49-228-287-6108; fax: +49-228-287-6294. E-mail address: [email protected] (C. Helmstaedter).
the brain including brain stem (e.g. nucleus of the solitary tract), diencephalic (e.g. thalamus), and cortical regions (e.g. insular cortex; Rutecki, 1990). This obviously close integration of the vagus nerve and the CNS suggests that vagal afferent input may affect cognitive brain functions. Following early animal studies, which revealed specific changes in spontaneous electrical brain activity as well as considerable attenuation of epileptic activity due to vagus nerve stimulation (VNS; Zabara, 1992), the first epilepsy patient was provided with a completely implantable
0920-1211/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 0 - 1 2 1 1 ( 0 1 ) 0 0 2 9 1 - 1
38
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
device for continuous intermittent left cervical VNS in 1988. To date, approximately 14 000 patients have been treated with VNS. While there is an agreement that VNS is effective in seizure control (Ben Menachem et al., 1994; George et al., 1994; Handforth et al., 1998), the cognitive effects of VNS have not yet been studied extensively. In animal experiments electrical vagal stimulation affected learning performance and behavior (Clark et al., 1995, 1998). Clark et al. (1999) were the first to report word-recognition memory enhancement due to acute VNS in human subjects. According to their findings in animal studies, recognition enhancement was most expressed at moderate stimulation output currents (0.5 mA) whereas no effects were obtained with very low (0.25 mA) or higher currents (0.75– 1.50 mA). Chronic VNS application appears not to lead to changes in cognitive functioning (Dodrill, 1997; Dodrill and George, 2001; Hoppe et al., 2001). Motivated by the findings of Clark et al., we now feel encouraged to report experimental data on the cognitive effects of acute VNS phases collected from 11 epilepsy patients who participated in our first VNS series (EO3 study; Ben Menachem et al., 1994). The study was set up to evaluate whether any cognitive side-effects can be expected from long-term VNS treatment within experimental conditions. Memory and decision times were tested in a box car design and performance of a group of healthy subjects who had repeated testing without any intervention served as control.
1. Methods
1.1. Patients Inclusion criteria were, adult person; pharmakoresistant epilepsy; and at least four partialonset seizures per month for 3 consecutive months. Eleven patients agreed to participate in the experimental study and gave informed consent (age, 18–42 years; mean, 31.9 years). The evaluation was carried out 5– 17 months (mean, 10.3 months) after implantation.
1.2. Computerized memory assessment Memory was assessed by two computerized memory tests, which require subsequent list learning of 15 words or nine designs in three learning trial each (Ju¨ nemann et al., 1992). Each list exposure was followed by immediate recognition of the target items out of alternatives (yes/no response). Words and designs in learning and recognition phases were presented sequentially. After completion of learning words and figures, final recognition of both materials without previous presentation was required. The wordlist of 15 nouns and the 30 distractor words were balanced for frequency of occurrence, the nine designs and 18 distractor designs (4×4 checkerboards with different patterns) were analogue to the Metric Figure test from Warrington and James (1967) and have a high demand on visual spatial memory (Warrington and James, 1967). The tests have been designed for experimental and clinical conditions which require repeated testing after short intervals and have been shown to be differentially sensitive to left/right temporal dysfunction in preand post-ictal follow-up testing in patients with lateralized temporal and frontal lobe seizures (Helmstaedter et al., 1994). For each test-session new lists were randomly generated out of a large item pool so that no patient gets the same testitems a second time and that tests of comparable difficulty were provided. Test duration is about 4 –5 min. Test analysis in healthy subjects and patients with lateralized epilepsies indicated that a summary score (sum of correct answers minus errors) was the most representative and discriminative measure of the test. Furthermore, a summary score of all decision times was calculated for control of vigilance and psychomotor speed. Parallel- or retest-reliabilities in healthy subjects and patients with left or right temporal lobe epilepsy were satisfactory with rtt-values ranging between 0.70 and 0.82 (Ju¨ nemann et al., 1992).
1.3. Experimental study design The experimental study design was a box-car design with ‘A1’ standing for baseline testing without VNS stimulation, ‘B’ standing for testing
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
during VNS stimulation and ‘A2’ standing for a third testing again without stimulation in the patient group. For the evaluation of performance courses during the three test trials normative data were available from repeated testing of 20 healthy, but unmatched subjects with average intelligence (age 19–45 years). The repeated tests in the control subjects who of course had no stimulation trial are labeled ‘A1’, ‘A2’, and ‘A3’.
1.4. Vagus ner6e stimulation VNS was performed using the implanted NeuroCybernetic Prosthesis™ System (Cyberonics, Inc.), which had been implanted for add-on therapy of phamacoresistant seizures. Vagal nerve stimulation during the experimental stimulation trial ‘B’ was delivered with the same output currents which were applied clinically at that time (output current, median 1.75 mA, range 1.00– 2.50 mA; pulse width, 500 ms; pulse frequency, 30 Hz). However, for the purpose of this study, stimulation duration was set from 30 s to a maximum length of 4.5 min to cover the time of memory testing in trial B. Thus, stimulation was active during item presentation and recognition in all learning trials. No patient complaint about prolonged stimulation during the second trial. For the baseline trials ‘A’ the stimulation was switched off using the magnet which provides external control over the device.
39
‘group’ (F= 6.6, P= 0.001) indicating that patients performed more poorly than controls across all measures (univariate tests, verbal memory F= 3.9, P= 0.057, figural memory F= 5.8, P = 0.02, verbal decision times F= 28.7, P = 0.000, figural decision times F= 0.91, P = 0.005). The main effect of the factor ‘test repetition’ was also statistically significant (F= 8.0, P= 0.02), indicating significantly decreasing decision times particularly in the verbal memory test and a reversible deterioration of figural memory performance (univariate testing, verbal decision times F= 6.2, P= 0.004, figural memory F= 2.7, P= 0.07). The significant main effect of ‘test repetition’ on figural memory could be explained fully by a highly significant interaction effect of ‘group Xtest-repetition’ on figural memory (F= 7.8, P = 0.009). Post hoc analysis confirmed that in patients only figural memory deteriorated from trial A1 to trial B (T=2.7, P= 0.02) and recovered from trial B to A2 (T= − 2.7, P= 0.02; Fig. 1). No such effect was seen for verbal memory (Fig. 2). Although the interaction effect group by test repetition did not become significant (F= 2.1, P= 0.13) for decision times, it should be noted
1.5. Statistical analysis Verbal/figural memory and decision times were analyzed by means of a repeated measurement multivariate analysis of variance with the three ‘test trials’ as the within-group factor and ‘group’ (patient/control) as the between-group factor. Post hoc analysis was carried out using T-tests for dependent samples with two-tailed significance.
2. Results Repeated measurement analysis of variance (ANOVA) showed a significant main effect of
Fig. 1. Figural recognition before, during and after VNS in patients as compared with three consecutive trials in healthy subjects (group means and S.E.M.).
40
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
Fig. 2. Verbal recognition before, during and after VNS in patients as compared with three consecutive trials in healthy subjects (group means and S.E.M.).
that without consideration of material-specificity, decision times in patients tended to be accelerated during the stimulation B as compared with both baseline tests (T-tests for dependent measures, A1 –B, t= 1.9, P =0.08; B– A2, t = −1.8, P = 0.09). In healthy subjects, decision times significantly accelerated from test A1 to A2 and remained stable from test A2 to A3 (T-tests for dependent measures, A1 – A2, t = 2.2, P = 0.04; A1 –A3, t=2.1, P= 0.05; Fig. 3).
1977), the effects of VNS on memory performance and response latency appear to be independent from each other. Since verbal recognition performance remained unaffected, it is unlikely that unspecific distraction or irritation due to the prolonged stimulation caused the results. In fact, the trend of a reversible acceleration of decision times during VNS could rather be discussed as an indicator of positive VNS effects on attention. This would be in line with many patient reports of increased arousal and improved attention during VNS long-term treatment (George et al., 1994). As a practical consequence of the findings one may ask whether the negative effect on figural memory has any consequences for the patients’ capabilities in every day lives. We feel that this is very unlikely. The obtained effects were completely reversible, in the clinical application stimulation conditions (intensity and duration) are different, and in every day life, VNS is not timelocked to a certain cognitive process. The suggestion that there is no recognizable effect in every day functioning would be consistent with previous reports which did not find any consistent change in cognitive performance with chronic VNS treatment (Dodrill, 1997; Dodrill and George, 2001; Hoppe et al., in press). However, more experimen-
3. Discussion In contrast to previous reports, our experimental data reveals a negative effect of acute VNS on memory performance. In an experimental baseline –treatment–baseline study a fully reversible deterioration of figural recognition memory was observed when vagal stimulation was delivered during the task. No such effects could be observed on verbal recognition memory or on decision times in the respective test. In patients, yes/no-decision times tended to be accelerated during VNS and returned to baseline levels after stimulation. Since poorer memory performance normally goes along with longer decision latencies (Koppell,
Fig. 3. Decision latencies during verbal and figural memory performance before, during and after VNS in patients as compared with three consecutive trials in healthy controls (group means and S.E.M.).
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
tal research regarding the effects of different types of VNS stimulation on cognition is recommended, in order to fully exclude negative effects or even in order to show positive cognition enhancing effects. Our results appear to be in contrast to the findings of Clark et al. (1999) who demonstrated verbal memory enhancement due to acute VNS. However, the two studies completely differ with respect to the study design, i.e. stimulation conditions and output currents. In this study, VNS was delivered during both learning and immediate recognition and could be suggested to interfere with different subroutines of memory processing (e.g. perception, encoding, recognition). In the study of Clark et al., VNS was delivered only during retention intervals and thus selectively affected memory consolidation. Furthermore, in our study all patients had output currents greater than 1.00 mA, whereas the patients of Clark et al. had currents 51.00 mA. The ‘boost’ effect, which is observed with low currents, may invert to a suppression effect with higher output currents (1.25–1.75 mA). The findings could thus be complementary rather than contradictory. Surprisingly, left cervical VNS selectively affected figural learning. The figural memory task applied in this study has previously been shown to be selectively sensitive for right temporal memory dysfunction immediately after right temporal seizures (Helmstaedter et al., 1994). Thus, from a neuropsychological point of view one may speculate on a lateralized cerebral effect of VNS on right hemisphere function. Although findings from neuroimaging studies are not equivocal (Garnett et al., 1992; Henry et al., 1999), the study of Ko et al. (1999) who demonstrated decreased right temporal blood flow during VNS may give some support to this assumption. Further evidence is provided by a study of Henry et al. (1998) who found increased right thalamic activation during VNS. Interestingly, there is also some evidence that seizure response rates for VNS treatment are higher in patients with a right focus as compared to patients with a left focus (Ben Menachem et al., 1999).
41
4. Conclusion Our findings provide evidence that acute phases of VNS can affect cognitive functioning. From a neuropsychological point of view, left-hand cervical VNS may preferentially project into the contralateral hemisphere. Since direction and extent of the VNS effect on cognition appears to depend on the stimulation conditions, future experimental studies should cover the whole spectrum of output currents (0.25–2.00 mA) and should separately address different brain functions.
Acknowledgements The work of C. Hoppe was supported by a grant of Cyberonics (Webster/TX, Inc.). Additional support was provided by the Deutsche Forschungsgemeinschaft (DFG EL 122/6-2).
References Agostoni, E., Chinnock, J.E., DeBurgh Daly, M., Murray, J.G., 1957. Functional and histological studies of the vagus nerve and its branches to the heart lungs and abdominal viscera in the cat. J. Physiol. 135, 182 – 205. Ben Menachem, E., Manon Espaillat, R., Ristanovic, R., et al., 1994. Vagus nerve stimulation for treatment of partial seizures: 1. A controlled study of effect on seizures. First International Vagus Nerve Stimulation Study Group. Epilepsia 35, 616 – 626. Ben Menachem, E., Hellstro¨ m, K., Waldton, C., Augustinsson, L.E., 1999. Evaluation of refractory epilepsy treated with vagus nerve stimulation for 5 years. Neurology 52, 1265 – 1267. Clark, K.B., Krahl, S.E., Smith, D.C., Jensen, R.A., 1995. Post-training unilateral vagal stimulation enhances retention performance in the rat. Neurobiol. Learn. Mem. 63, 213 – 216. Clark, K.B., Smith, D.C., Hassert, D.L., Browning, R.A., Naritoku, D.K., Jensen, R.A., 1998. Posttraining electrical stimulation of vagal afferents with concomitant vagal efferent inactivation enhances memory storage processes in the rat. Neurobiol. Learn. Mem. 70, 364 – 373. Clark, K.B., Naritoku, D., Smith, D.C., Browning, R.A., Jensen, R.A., 1999. Enhanced recognition memory following vagus nerve stimulation in human subjects. Nat. Neurosci. 2, 94 – 98. Dodrill, C.B., 1997. Cognitive and quality of life changes with VNS therapy. Results from a randomized controlled trial,
42
C. Helmstaedter et al. / Epilepsy Research 47 (2001) 37–42
abstract. Current results and perspectives on vagus nerve stimulation for the treatment of epilepsy. Cyberonics, Houston/TX. Dodrill, C.B., George, M.L., 2001. Effects of vagal nerve stimulation on cognition and quality of life in epilepsy. Epilepsy Behav. 2, 46 –53. Garnett, E.S., Nahmias, C., Scheffel, A., Firnau, G., Upton, A.R.M., 1992. Regional cerebral blood flow in man manipulated by direct vagal stimulation. Pacing Clin. Electrophysiol. 15, 1579 – 1580. George, R., Salinsky, M., Kuzniecky, R., et al., 1994. Vagus nerve stimulation for treatment of partial seizures: 3. Longterm follow-up on first 67 patients exiting a controlled study. First International Vagus Nerve Stimulation Study Group. Epilepsia 35, 637 –643. Handforth, A., DeGiorgio, C.M., Schachter, S.C., 1998. Vagus nerve stimulation therapy for partial-onset seizures: a randomized active-control trial. Neurology 51, 48 – 55. Helmstaedter, C., Elger, C.E., Lendt, M., 1994. Postictal courses of cognitive deficits in focal epilepsies. Epilepsia 35, 1073 –1078. Henry, T.R., Bakay, R.A., Votaw, J.R., et al., 1998. Brain blood flow alterations induced by therapeutic vagus nerve stimulation in partial epilepsy: I. Acute effects at high and low levels of stimulation. Epilepsia 39, 983 –990. Henry, T.R., Votaw, J.R., Pennell, P.B., Epstein, C.M., Bakay, R.A., Faber, T.L., Grafton, S.T., Hoffman, J.M., 1999. Acute blood flow changes and efficacy of vagus nerve stimulation in partial epilepsy. Neurology 52, 1166 – 1173. Hoppe, C., Helmstaedter, C., Scherrmann, J., Elger, C.E., 2001. No evidence for cognitive side effects after 6 months of vagus nerve stimulation in epilepsy patients. Epilepsy Behav., in press. Ju¨ nemann, H., Helmstaedter, C., Elger, C.E., 1992. Die Mo¨ glichkeiten des Einsatzes computerisierter Geda¨ chtnistests in der pra¨ chirurgischen Epilepsiediagnostik [use of
computerized memory tests in presurgical evaluation of epilepsy]. In: Scheffner, D. (Ed.), Epilepsie 91. EinhornPresse Verlag, Reinbeck, pp. 449 – 452. Ko, D.Y., Grafton, S.C., Heck, C.N., Smith, T.D., DeGiorgio, C.M., 1999. Prolonged and progressive cerebral blood flow activation and deactivation with vagus nerve stimulation: more lateralization to contralateral structures. Poster presented at the 54th AES Meeting, December 6, 1999, Orlando, FL (2.223). Epilepsia 40 (Suppl. 7), 139. Koppell, S., 1977. Decision latencies in recognition memory: a signal detection theory analysis. J. Exp. Psychol. Hum. Learn. Mem. 3, 445 – 457. Porges, S.W., Doussard-Roosevelt, J.A., Maiti, A.K., 1994. Vagal tone and the physiological regulation of emotion. Monogr. Soc. Res. Child. Dev. 59, 167 – 186. Randich, A., Gebhart, G.F., 1992. Vagal afferent modulation of nociception. Brain Res. Rev. 17, 77 – 99. Rogers, R.C., McTigue, D.M., Hermann, G.E., 1996. Vagal control of digestion: modulation by central neural and peripheral endocrine factors. Neurosci. Biobehav. Rev. 20, 57 – 66. Rutecki, P., 1990. Anatomical physiological and theoretical basis for the antiepileptic effect of vagus nerve stimulation. Epilepsia 31 (Suppl. 2), S1 – S6. Uvnas Moberg, K., 1994. Role of efferent and afferent vagal nerve activity during reproduction: integrating function of oxytocin on metabolism and behaviour (Proceedings of the International Conference on Hormones Brain and Behaviour, 1993 Tours/France). Psychoneuroendocrinology 19, 687 – 695. Warrington, E.K., James, M., 1967. Disorders of visual perception in patients with localized cerebral lesions. Neuropsychologia 5, 253 – 266. Zabara, J., 1992. Inhibition of experimental seizures in canines by repetitive vagal stimulation. Epilepsia 33, 1005 – 1012.