Letters

Original Letters

**This letter was subject to peer-review.

Vagus Nerve Stimulation in a Case of Epilepsy with CSWSS:Respiratory Side Effects during Sleep

*Claire Héberlé, †Patrick Berquin, ‡Nicole Larnicol, and *Fabrice Wallois

*Service d’Explorations Fonctionnelles du Système Nerveux and †Federation de Pediatrie, Unité Neuropédiatrique,Amiens, France, and ‡Laboratoire de Neurophysiologie, Faculte de Medecine, Amiens, France

We would like to alert the medical community to pos-sible respiratory side effects of vagal nerve stimulation(VNS) during the treatment of nonsurgical refractorypartial epilepsy. We observed hypopneic–polypneic epi-sodes during sleep in a 9-year-old girl with epilepsy withcontinuous spike–waves during slow sleep (CSWSS)syndrome. These events were time-locked to each periodof VNS stimulation and occurred without any modifica-tion of cardiac frequency. VNS may elicit a reduction of25–50% in seizure frequency in both adults and childrenwith intractable epilepsy (1). Adverse side effects andcomplications of VNS were few and mostly transient andwere reported in adult patients only. They includehoarseness, neck pain, hypersalivation, cough, and short-ness of breath (“air-missing”) sensations during physicalexercise. These sensations might be related to the in-crease in end-expiratory volume observed in some pa-tients receiving stimulation at high intensity. In contrastto the common observation that VNS modifies the respi-ratory pattern in animals, only one study has reportedalterations in the respiratory pattern in humans (2).

In the present case, the mother’s pregnancy wasmarked by fetal death of a twin sister at 24 weeks ofpostconceptional age. Birth was uneventful at 36 weeksof gestational age. Systematic transfontanellar ultrasoundanalysis revealed bilateral diffuse periventricular leuko-malacia. At age 2 years, she had right tonic partial sei-zures with loss of contact, and magnetic resonance

imaging (MRI) evidenced a left parietal extended isch-emic lesion.

At age 5 years, her seizures changed and increased infrequency: during daytime, they were characterized byabsence seizures associated with mild and inconstant ato-nia (up to one seizure every 2–3 min). Sleep EEGshowed CSWSS. Over the course of the following year,she was treated unsuccessfully with seven different an-tiepileptic drugs (AEDs). A ketogenic diet also was in-effective. She was considered an ineligible candidate forsurgery because of the extent of the left sylvian ischemiclesion.

At age 6 years, a neurocybernetic prosthesis (Cyberon-ics, Houston, Texas U.S.A.) was implanted. The stimu-lation allowed a drug-free period of 2 months, duringwhich there was an initial seizure-free period of 6 weeks.Then absences reappeared but were less numerous thanbefore implantation. Subsequently she was given la-motrigine (LTG) with VNS. Despite the use of variousAED combinations and changes in stimulation param-eters over the next year, the absences persisted.

At age 61⁄2 years, while she was being treated withclonazepam (CZP) with VNS, a brief switching off of thestimulator resulted in the frequency of absence returningto preimplantation values (>100/day), indicating the rela-tive efficacy of VNS in CSWSS.

At age 7 years, the introduction of felbamate (FBM)and clobazam (CLB) with high-intensity VNS (intensity,1.75 mA; frequency, 30 Hz; pulse width, 500 �s; andon-time/off-time 60 s/1.8 min) freed the child from sei-zures for 2 years until the present day.

At age 8 years, an overnight polysomnographic re-cording included 20 EEG derivations, an electrooculo-gram, a submental electromyogram, a respiratory

Accepted May 26, 2002.Address correspondence and reprint requests to Dr. F. Wallois at

Service d’Explorations Fonctionnelles du Système Nerveux, UnitéNeuropédiatrique, CHU Amiens Nord, Place Victor Pauchet, 80054Amiens, France. E-mail: Fabrice.Wallois@sa.u-picardie.fr

Epilepsia, 43(10):1268–1272, 2002Blackwell Publishing, Inc.© International League Against Epilepsy

1268

piezoelectric device encircling the ribcage, a nasal ther-mistance, and an electrocardiogram. The VNS signal wasmonitored by means of two electromyographic surfaceelectrodes placed over the position of the vagal elec-trodes used for stimulation. The nasal thermistancemoved during the recording and did not give reliablevalues over the whole recording period. Nevertheless,when both traces were recorded, the signal from the pi-ezoelectric device was superimposable on that of thenasal thermistance, which indicated accurate ventilatorymonitoring.

The following reproducible markers were used in thedata analysis: beginning of inspiration, peak inspiratoryvelocity (PIV), and peak expiratory velocity (PEV).From pilot evaluations, PIV and PEV did indeed reflectinspiratory and expiratory peak flows. The followingtime intervals were measured by using these markers:T1, interval between the beginning of inspiration andPIV; T2, interval between the PIV and PEV; and T3,interval between PEV and the beginning of inspiration.Instantaneous respiratory frequency was taken as (T1 +T2 + T3)/60.

Breath-by-breath analysis was performed by usingpercentage changes, relative to control values, averagedfrom the three respiratory cycles immediately precedingthe onset of the stimulus. Mean changes in ventilatoryparameters were assessed by using values calculatedwithin 10-s intervals. Eighty periods of VNS stimulationwere examined for this purpose during sleep periods, freeof body movement. Differences between control andstimulation periods were tested by using analysis of vari-ance, followed by post hoc PLSD (Protected LeastSquare Differences) Fisher corrections (StatView+ pack-age).

Every period of VNS was tightly linked to a period ofrespiratory disturbance, which lasted for the entire periodof stimulation. These respiratory modifications werecharacterized by an immediate decrease in the amplitudeof both inspiratory and expiratory movements, as esti-mated from the piezoelectric trace and also as observedon nasal thermistance. At the same time, respiratory fre-quency increased (+16.0 ± 2.9%; p < 0.001) secondary todecreases in all the intermediate times (T1, –18.1 ±3.7%; p < 0.003; T2, –14.8 ± 2.1%; p < 0.0001; T3,–20.4 ± 3%; p < 0.007). This early period lasted for∼10 s.

After this, the amplitude of respiration progressivelyreturned to control values over the next 50 s. Neverthe-less, respiratory frequency remained at a steady, in-creased value. It was only after a few seconds at the endof VNS stimulation that respiratory frequency and am-plitude returned to prestimulation values.

It should be noted that respiratory disturbances wereobserved throughout the sleep period, but were less eas-ily distinguishable during wakefulness because of inces-

sant movement of the child. In this child, the occurrenceof CSWSS prevented any distinction between sleepstages within the EEG recordings, because interictal ac-tivity was not sufficiently suppressed to ascertain wheth-er the child was in a rapid eye movement (REM) ornon-REM (NREM) sleep period. No change was noticedin cardiac frequency (obtained from the RR interval),global EEG activity, or interspike interval that might berelated to individual periods of VNS.

Before implantation of the stimulator, daytime EEGpresented a stereotypic pattern of paroxysmal activities.It consisted of a short run of theta waves intermingledwith spikes over a 1-s period, followed by a period of2–3 s relatively free of abnormalities. Then polyspikesoccurred for 1 s, rapidly followed by spike–waves at 3Hz. Spike–waves at a more variable frequency persistedfor ∼2 s to several minutes, associated with a loss ofcontact. After implantation of VNS, in the different pe-riods free from seizures, the stereotypic paroxysmalactivities during daytime EEG were modified, beingless numerous and more or less confined to the lefthemisphere. The sequence of theta run followed by aperiod free of abnormalities persisted, but polyspikesor synchronized spike–waves at 3 Hz were no long-er triggered. Instead, a period of spike–waves at veryvaried frequencies occurred for short periods, never ex-ceeding 15 s.

Our experience underscores the dual influence of VNSin a case of CSWSS syndrome. This is the first report ofrespiratory disorders tightly linked to VNS in a childwith epilepsy. The changes in both respiratory timingand amplitude were very similar to those previously de-scribed in animals in response to electrical stimulation ofsmall myelinated and unmyelinated vagal afferents (3).As these responses were elicited from the central cut endof the vagus nerve, they were assumed to result fromdirect interactions with respiratory control centers. Theywere marked by an earlier onset of both inspiration andexpiration, resulting in shortening of both inspiratory andexpiratory times, as we also observed. Moreover, theincrease in respiratory frequency was associated with areduction in the amplitude of the pneumogram, an alter-ation that appeared to be quite similar to that we ob-served at the beginning of VNS. Animal studies alsoraised the possibility that the reductions in respiratoryflow and the velocity of thoracoabdominal distentionmight be due to a reflex restrictive respiratory episode.Indeed, the stimulation of the central end of vagus nerveincreased the tonic inspiratory activity of the diaphragm,the intercostal, and the laryngeal muscles, which couldresult in stiffening the respiratory tract and the chest.Finally, the possibility of a direct activation of vagalefferents should not be definitely ruled out in the presentcase.

Conversely, the present preliminary observation high-

LETTERS 1269

Epilepsia, Vol. 43, No. 10, 2002

lights the potential benefit of VNS in cases of epilepsywith CSWSS. Both the reduction in seizure frequencyand the changes in EEG activities indicated that VNSimproved the clinical status of the girl under study. Asan additional argument, her symptoms immediatelyworsened during a temporary arrest of the stimulator.Nevertheless, it remains to be determined whether theimprovements may be ascribed to VNS itself or to thecombination of VNS and FBM. This deserves larger-scale evaluations to confirm the therapeutic efficacyof VNS in CSWSS syndrome. In addition, this casereport appears sufficiently demonstrative to propose asystematic monitoring of respiratory parameters dur-ing overnight polygraphic recordings in patients treatedwith VNS. This should allow a detailed assessment ofthe risks of ventilatory disturbances, and their ex-tent, within the range of the clinical applications ofVNS.

Extrapolating previous animal data to the present andother findings in humans raises the question whether theoccurrence of discrete alterations of ventilatory param-eters during VNS might provide a functional indicator ofthe activation of small myelinated and/or unmyelinatedvagal fibers, which has been regarded as a contributingfactor to the improvement of epileptic disorders.

REFERENCES

1. Murphy JV. Left vagal nerve stimulation in children with medi-cally refractory epilepsy: the Pediatric VNS Study Group. J Pedi-atr 1999;134:563–5.

2. Malow BA, Edwards J, Marzec M, et al. Effects of vagus nervestimulation on respiration during sleep: a pilot study. Neurology2000;55:1450–4.

3. Massion J, Colle J. Influence de la stimulation du nerf vague sur lediaphragme et sur les muscles intercostaux. Arch Int Physiol Bio-chem 1960;68:656–68.

**We are greatly indebted to Dr. Plouin for her helpfulcomments and criticisms on this manuscript.

Panayiotopoulos Syndrome or Early-onset BenignChildhood Occipital Epilepsy

To the Editor:The recent “proposed diagnostic scheme for people

with epileptic seizures and with epilepsy of the Interna-tional League Against Epilepsy (ILAE) Task Force onclassification” (1) makes a significant contribution byrecognizing Panayiotopoulos syndrome (PS) (2) amongthe idiopathic focal epilepsies of childhood in addition tothe “benign childhood epilepsy with centrotemporalspikes” and the “late-onset childhood occipital epilepsy(Gastaut type).” They proposed a descriptive name“early-onset benign childhood occipital epilepsy” at-tached to an eponym “Panayiotopoulos type (syndrome)” (1).

I wish to draw attention that the descriptive nomen-

clature of PS as “occipital epilepsy,” also previously de-scribed as “with occipital paroxysms,” may be misleading.

1. Occipital paroxysms in their classic form withfixation-off sensitivity is a rare finding in PS andcertainly nonspecific (3). Interictal EEG in PSmainly manifests with multifocal spikes at variouslocations, although occipital spikes often (70%)predominate (2,4,5). EEGs may be normal or with-out occipital spikes (30%) (2,4,5). In the originalstudy of Panayiotopoulos (4) of 21 otherwise nor-mal children with ictal vomiting; occipital spikesoccurred in 12 (57%); the others had extraoccipitalspikes (five), infrequent brief generalized dis-charges (one), or consistently normal EEG (three)(4). Subsequent attention was focused on the pre-dominant group of occipital spikes and occipitalparoxysms, but this is now corrected to include thegroup of “extraoccipital spikes or normal EEG.”The clinical manifestations of PS are the same ir-respective of EEG localizations.

2. Occipital epilepsy also is incorrect for the follow-ing good reasons: (a) onset of seizures is mainlywith autonomic symptoms and particularly emesis(80%) (5). Of occipital symptoms, only deviationof the eyes may originate from the occipital re-gions, but this rarely occurs at onset. Visual symp-toms are exceptional and not consistent in recurrentseizures; (b) interictal occipital spikes may neveroccur, (c) even ictal EEG has documented anterioronset (6).

3. Characterizing PS as “epilepsy” also is controver-sial and, in my opinion, unsatisfactory. One third ofchildren with PS have a single seizure, which bythe operational definition of epilepsy (more thantwo seizures) is not epilepsy. Further, PS is not “achronic neurological condition characterized by re-current epileptic seizures” (the current definition ofthe ILAE glossary) (7).

All these point out that PS should be classified among“conditions with epileptic seizures that do not require adiagnosis of epilepsy,” which is a new concept of theILAE proposal to incorporate febrile, benign neonatal,single seizures, isolated clusters of seizures, and rarelyrepeated seizures (oligoepilepsy) (1). PS is a commonclinical phenotype of the benign childhood seizure sus-ceptibility syndrome (2,5). It manifests with autonomicseizures and autonomic status epilepticus that in onethird are singular events. Prognosis is excellent even forthose (∼10%) who may initially have many seizures. In-terictal EEG shows significant variability, even for thesame child. Predominantly PS affects children aged 3–7years, probably as the result of susceptible emetic andautonomic centers of this age group (5). PS is a signifi-cant missing land in pediatric epileptology with immense

LETTERS1270

Epilepsia, Vol. 43, No. 10, 2002

clinical and management implications. Prospective stud-ies are needed to clarify the many typical and atypicalclinical presentations of PS, their autonomic compo-nents, and the EEG variations.

Zarko MartinovicInstitute for Mental Health, Department of Epilepsy

and Clinical NeurosciencesBelgrade, Yugoslavia

REFERENCES

1. Engel J Jr. A proposed diagnostic scheme for people with epilepticseizures and with epilepsy: report of the ILAE Task Force onClassification and Terminology. Epilepsia 2001;42:796–803.

2. Ferrie CD, Grunewald RA. Panayiotopoulos syndrome: a commonand benign childhood epilepsy [Commentary]. Lancet 2001;357:821–3.

3. Martinovic Z. Panayiotopoulos syndrome. Lancet 2001;358:69.4. Panayiotopoulos CP. Vomiting as an ictal manifestation of epilep-

tic seizures and syndromes. J Neurol Neurosurg Psychiatry 1988;51:1448–51.

5. Panayiotopoulos CP. Panayiotopoulos syndrome: a common andbenign childhood epileptic syndrome. London: John Libbey, 2002.

6. Oguni H, Hayashi K, Imai K, et al. Study on the early-onset variantof benign childhood epilepsy with occipital paroxysms otherwisedescribed as early-onset benign occipital seizure susceptibility syn-drome. Epilepsia 1999;40:1020–30.

7. Blume WT, Luders HO, Mizrahi EM, et al. ILAE Commissionreport: glossary of descriptive terminology for ictal semiology:report of the ILAE Task Force on classification and terminology.Epilepsia 2001;42:1212–8.

Prevention of Refractory Epilepsy

To the Editor:I was pleased to see Arroyo et al. (1) revisiting the

fundamental questions that my colleagues and I raised 20years ago (2) i.e., Why does epilepsy become intrac-table? and Can chronic epilepsy be prevented?

We asked these questions then on the basis of ourearliest prospective studies of the treatment of newlydiagnosed epilepsy in the preceding 10 years, which hadrevealed (a) a much better prognosis than had been an-ticipated on the basis of previous studies in chronic pa-tients, and (b) most patients with intractability could beidentified within the first 2 years of treatment (3).

Since I reviewed the subject in Epilepsia in 1987 (4),it is interesting that on the basis of more recent studiesArroyo et al. are able to support our view that refractoryepilepsy can be a progressive disorder, which, if con-trolled early, might never develop into a full syndromewith all its associated sequelae, and that the early initia-tion of aggressive therapy may improve outcome andoverall quality of life.

There are, however, a number of factors that Arroyo etal. overlook in their otherwise thoughtful review.

1. They do not mention the subject of compliancewith medication, although in our experience, the

commonest cause of failure of treatment associatedwith intractability was poor compliance (3,4).

2. Among many factors that Arroyo et al. (4,5) rightlyidentify as increasing the risk of intractability, thesingle most important in our own studies was thenumber of seizures before the start of treatment.

3. While referring, like ourselves, to the processes ofkindling and secondary epileptogenesis, they omit-ted what we called the Gowers hypothesis (2,4), hisview that each seizure predisposes to the next,which is an extension of the kindling phenomenon,of which Gowers was, of course, unaware.

4. Any discussion of processes of progression, re-lapse, or intractability should be balanced by a dis-cussion of processes of remission (5,6).

It is apparent, as Arroyo et al. acknowledged, thatespecially in childhood, some epilepsy syndromes appearto go into remission with or without treatment. The keyto understanding the evolution and prognosis of epilepsysyndromes lies in the relation of these two competingprocesses. To what extent our treatments can prevent therelapsing processes or enhance the remitting processes isa central question of our future understanding and man-agement of epilepsy and its various syndromes.

E. H. ReynoldsInstitute of Epileptology

London, Kings College, United Kingdom

REFERENCES

1. Arroyo S, Brodie MJ, Avanzini G, et al. Is refractory epilepsypreventable? Epilepsia 2002;43:437–44.

2. Reynolds EH, Elwes RDC, Shorvon SD. Why does epilepsy be-come intractable? Prevention of chronic epilepsy. Lancet1983;ii:952–4.

3. Elwes RDC, Johnson AL, Shorvon SD, et al. The prognosis forseizure control in newly diagnosed epilepsy. N Engl J Med 1984;311:944–7.

4. Reynolds EH. Early treatment and prognosis of epilepsy. Epilepsia1987;28:97–106.

5. Reynolds EH. The influence of antiepileptic drugs on the naturalhistory of epilepsy. In: Pisani F, Perucca E, Avanzini G, et al., eds.New antiepileptic drugs. Amsterdam: Elsevier, 1991:15–9.

6. Reynolds EH. Mechanisms of intractability. In: Wolf P, ed. Epi-leptic seizures and syndromes. London: John Libbey, 1994:599–603.

Response: Prevention of Refractory Epilepsy

To the Editor:We thank Dr. Reynolds for his support of our article,

and we acknowledge the insight and contributions thattheir group had in the area of progression of epilepsy afew years ago. We agree with Dr. Reynolds that a com-mon cause of failure of the treatment is the patient’snoncompliance. However, for the purpose of our review,we considered it understood that noncompliant patientsare not truly refractory.

LETTERS 1271

Epilepsia, Vol. 43, No. 10, 2002

The number of seizures before the start of treatmenthas been found in some but not all studies to be a riskfactor for intractability. In any case, we concur with Dr.Reynolds in that it appears to be a relevant risk factor. Ina recent article from one of us (1), it was confirmed thatthose patients with 20 or more seizures before treatmenthad a significantly higher chance of intractability.

Finally, in our text, we discussed the phenomenon ofsecondary epileptogenesis and of the possibility that sei-zures beget seizures. Unfortunately, because of numer-ous methodologic issues, the evidence that thisphenomenon occurs in humans has been difficult to as-certain. In this sense, different types of epilepsy mightbehave differently. Even in those types that we believe

are progressive (like mesial temporal lobe epilepsy), aninitial and sometimes years-long “honeymoon period” isoften observed, implying that the disease evolution andnot the seizures in themselves might be responsible forthe intractability.

Dr. Santiago ArroyoMedical College of WisconsinMilwaukee, Wisconsin, U.S.A.

REFERENCE

1. Kwan P, Brodie MJ. Early identification of refractory epilepsy. NEngl J Med 2000;342:314–9.

LETTERS1272

Epilepsia, Vol. 43, No. 10, 2002