Clinical Commentary

EEG with triphasic waves in Borreliaburgdorferi meningoencephalitis

Triphasic waves in the EEG are strongly associ-ated to metabolic encephalopathies of varioustypes, especially hepatic encephalopathy, but theyare not entirely specific. Triphasic waves havebeen described in a variety of non-metabolicencephalopathies and structural brain lesions(1–3). Here, we present a case of Borrelia burg-dorferi meningoencephalitis with triphasic wavesin the EEG.

A case report

A 72-year-old woman with hypertension, a milddiabetes mellitus without medication, hypothyroid-ism, pain in both shoulders and a right total hipreplacement due to coxarthrosis, came to the clinicof neurology for a sudden weakness of her left armand hand. At examination, she had a slight paresisin the left arm and hand but no other neurologicaldeficits. The pain in the left shoulder made theneurological evaluation somewhat difficult. Com-puted tomography of the brain showed no cerebralinfarction but leukoaraiosis. The sylvian fissureswere bilaterally a bit wider than normal. A carotidduplex was normal. ECG: sinus rhythm, inferiorlya probable pathological q-wave suggesting anearlier cardiac infarction. The condition was inter-preted as cerebral infarction, probably of lacunar

type. After 5 days, the patient could be dischargedwith aspirin 160 mg daily as prophylaxis, enhancedantihypertensive medication and oxazepam 5 mgwhen necessary against anxiety. She was alsoreferred to polyclinic rehabilitation.After only one day, the patient returned to the

emergency care unit with fatigue, anxiety andmalaise and was admitted to a neurology ward forobservation (day 1). A slight paresis remained inthe left arm.On the following day, the patient developed

stupor. Her speech was limited to a few words andshe fell asleep immediately after stimulation. Therewas suspicion of oxazepam intoxication, but theantidote flumazenil had no effect.A new computed tomography of the brain

showed no difference compared with the precedinginvestigation.Electroencephalogram on day 3 revealed gener-

alized slowing, frontal bilateral triphasic waves, abackground of delta–theta waves, and no epilep-tiform activity (Fig. 1).There were no signs of metabolic dysfunction,

neither hepatic nor azotemic. Blood ammonium,blood glucose and electrolytes were all withinnormal range.The patient recovered slowly during the follow-

ing days. A slight confusion and disorientation

Eriksson B, Wictor L. EEG with triphasic waves in Borrelia burgdorferimeningoencephalitis.Acta Neurol Scand 2007: 116: 133–136.� 2007 The Authors Journal compilation � 2007 Blackwell Munksgaard

We describe a case of encephalopathy in which the clinical picture andtriphasic waves in the EEG indicated a metabolic cause. However, theillness was caused by neuroborreliosis. The occurrence of triphasicwaves in the EEG is a strong evidence of metabolic encephalopathy,but triphasic waves are not specific for metabolic encephalopathy.Triphasic waves have been described in a number of non-metabolicencephalopaties and structural brain lesions. To our knowledge, this isthe first report of triphasic waves in Borrelia burgdorferimeningoencephalitis.

B Eriksson, L WictorDivision of Clinical Neurology, Lund University Hospital,Lund, Sweden

Key words: borrelia burgdorferi;electroencephalogram; encephalopathy;neuroborreliosis; meningoencephalitis; triphasic waves

Bengt Eriksson, Clinic of Internal Medicine, LandskronaHospital, Box 514, S-261 24 Landskrona, Sweden.Tel.: +46 418 454000Fax: +46 418 454229e-mail: Bengt.B.Eriksson@skane.se

The authors contributed equally to this work.

Accepted for publication February 15, 2007

Acta Neurol Scand 2007: 116: 133–136 DOI: 10.1111/j.1600-0404.2007.00858.x Copyright � 2007 The AuthorsJournal compilation � 2007 Blackwell Munksgaard

ACTA NEUROLOGICASCANDINAVICA

133

remained. A new EEG on day 5 showed a markedimprovement. There were no triphasic waves,improved background activity but still a general-ized slowing around 7 Hz, and no epileptiformactivity (Fig. 2). MRI of the brain with diffusion-weighted images showed no lesions but there wasan enlargement of the ventricles suggesting anormal pressure hydrocephalus. (On request, thepatient complained of a very mild memorydisturbance for some years but she had no othersymptoms or signs of dementia or normal pres-sure hydrocephalus. Later a tap test and aninfusion test according to Katzmann turned outnormal.)On day 6, liquor examination revealed a mono-

cytic pleocytosis with 300 lymphocytes per mm3,no polymorphs, a highly elevated protein concen-tration (1.78 g ⁄ l), elevated IgG index andoligoclonal bands as signs of breakdown of theblood–brain barrier and intrathecal synthesis ofimmunoglobulins.The liquor findings primarily aroused suspicion

of neuroborreliosis. The patient was treated withoral doxycykline and recovered promptly. She wasalso treated with acyclovir for possible herpesencephalitis, but herpes simplex PCR in liquor wasnegative and acyclovir was withdrawn.A week later the patient was discharged from the

hospital in good health. Serology for Borreliaburgdorferi in serum and liquor confirmed thediagnosis neuroborreliosis. Serology for tick-borne

encephalitis was negative. Direct microscopy foracid-fast bacilli and CSF culture showed no signs oftuberculosis.The patient had no history of tick bite or

erythema chronicum migrans, but during thesummer she had often been in a tick infested area.

Discussion

In 1950, Foley, Watson and Adams describedbilateral synchronous slow activity with frontaldominance in EEG in hepatic coma, �blunt spikeand wave� (4). The term triphasic waves wasintroduced in 1955 by Bickford and Butt in astudy which only concerned patients with hepaticdisease. Triphasic waves were defined as waveswith three phases, a dominant positive turn of thescale, with a slow negative turn before and after thepositive wave. The triphasic waves occurred syn-chronously and bilaterally with frontal dominance,sometimes only frontally, sometimes over thewhole cortical surface, with latency, �anterior–posterior lag�. The presence of triphasic wavescould be linked to the patient�s degree of unre-sponsiveness to external stimulation. Triphasicwaves dominated in stupor (�semicoma�), whenpatient could be awaked only by strong stimuli,theta waves dominated in slight confusion, anddelta waves dominated in coma. In a control groupof patients with liver disease without encephalop-athy normal alpha rhythm was recorded (5).

100µV

1 s

Figure 1. Day 3. The patient is stuporous. EEG shows continuous bilateral delta–theta activity with bilateral frontal triphasic wavecomplexes. Examples of triphasic waves are surrounded by rings. Bipolar montage.

Eriksson & Wictor

134

Bickford and Butt�s description of triphasic wavemorphology is still valid with small modifications.The second positive phase is usually dominant inamplitude. Sometimes the first negative phase orthe third negative phase can be dominant togetherwith the second phase. Occasionally, the first phaseis preceded by a low-amplitude positive wave(�wave 0�). Just like the anterior–posterior lag, aposterior–anterior lag can be seen (6).Bickford and Butt proposed the mechanism

behind the triphasic waves to be either a travellingwave of positivity along the cortex or a subcorticaldisturbance at thalamocortical level (5). The path-ogenesis of the triphasic waves is still unknown.Bickford and Butt concluded that triphasic

waves probably were specific for hepatic encephal-opathy (5). Later triphasic waves have beendescribed in a number of other metabolic andtoxic encephalopathies, e.g. uremia, hypernatre-mia, hyponatremia, hypercalcemia, hypoglycemia,hyperthyroidism, hypothyrodism, side effects ofdrugs and anoxic encephalopathy (2) but alsostructural lesions like stroke, dementia, tumors andcarcinomatosis (3). In non-metabolic conditions,the triphasic waves can also be seen in alertpatients (3, 6).Among inflammatory and infectious encephalo-

pathies, the occurrence of triphasic waves in thediffuse encephalopathy of sepsis is well known (7).There are also single case reports of triphasic wavesinMollaret�smeningitis (8), i.e. recurrent episodes of

aseptic meningitis, now attributed to herpes simplexinfection, and carcinomatous meningitis (9).However, triphasic waves according to Bickford

and Butt�s definition are still considered to be highlyspecific for metabolic encephalopathy, with hepaticencephalopathy as the most common cause (1).No morphologic characteristics of the triphasic

waves or their background activity can reliablydistinguish between different metabolic-toxiccauses of the triphasic waves (10) nor betweenmetabolic or non-metabolic causes (6).Epileptiform discharges in non-convulsive

status epilepticus (generalized non-convulsivestatus epilepticus, also known as absence statusepilepticus, and partial non-convulsive statusepilepticus) can resemble triphasic waves. Gener-alized epileptic activity is and partial epilepticactivity can be bilateral and synchronous withfrontal predominance but never has the triphasicanterior–posterior (or posterior–anterior) lag. Theamplitude predominance of phase II is not seen.Phase I duration of epileptic activity is shorterthan phase I duration of triphasic waves, i.e. morespike than sharp transient. Epileptic activity canhave extra spikes which triphasic waves neverhave (11).Meningoencephalitis is a well-known but rather

rare manifestation of acute neuroborreliosis.Encephalitic signs can be reduced alertness andfocal neurologic deficits like hemiparesis, ataxiaand cranial nerve deficits. The meningoencephalitis

100µV

1 s

Figure 2. Day 5. The patient is fully awake but slightly confused and disoriented. EEG shows marked improvement. Continuoustheta activity around 7 Hz. No triphasic wave complexes. Bipolar montage.

EEG with triphasic waves

135

is often combined with radiculitis or myelitis. Thepatient is often afebrile (12, 13).The stuporous patient�s EEG clearly showed

triphasic waves. There were frontal triphasic waveswith phase I as a sharp transient and a predom-inant phase II, giving no indication of the differ-ential diagnosis partial generalized statusepilepticus.The clinical picture which primarily aroused

suspicion of metabolic encephalopathy was alsowell compatible with a Borrelia burgdorferi men-ingoencephalitis. The patient had a short durationof symptoms and the MRI gave no indication ofcerebrovascular neuroborreliosis which is usuallyseen in late stages of neuroborreliosis. The firstmanifestation of the illness, paresis of the left arm,could possibly have been a radiculitis, although thepain in the left shoulder was not new.Our case is a further confirmation that triphasic

waves are not entirely specific for metabolicencephalopathy. To our knowledge this is thefirst report of triphasic waves in Borrelia burgdor-feri meningoencephalitis.

Acknowledgement

We would like to thank Associate Professor Gert Andersson,Division of Clinical Neurophysiology, Lund University Hos-pital, Lund, Sweden, for help with the interpretation andpreparation of the electroencephalograms.

References

1. Niedermeyer E. Metabolic central nervous systemdisorders. In: Niedermeyer E, et al. eds. Electroencepha-

lography: Basic Principles, Clinical Applications, andRelated Fields. Philadelpha: Lippincott Williams & Wilk-ins, 5th edn, 2004;439–53.

2. Kaplan PW. The EEG in metabolic encephalopathy andcoma. J Clin Neurophysiol 2004;21:307–18.

3. Aguglia U, Gambardella A, Oliveri RL, Lavano A,Quattrone A. Nonmetabolic causes of triphasic waves: areappraisal. Clin Electroencephalogr 1990;21:120–5.

4. Foley JM, Watson CW, Adams RD. Significance of theelectroencephalographic changes in hepatic coma. TransAm Neurol Assoc 1950;75:161–5.

5. Bickford RG, Butt HR. Hepatic coma: the electroe-cephalographic pattern. J Clin Invest 1955;34:790–9.

6. Sundaram MBM, Blume WT. Triphasic waves: clinicalcorrelates and morphology. Can J Neurol Sci 1987;14:136–40.

7. Young BG, Bolton CF, Archiblad YM, Austin TW, Wells

GA. The electroencephalogram of sepsis-associatedencephalopathy. J Clin Neurophysiol 1992;9:145–52.

8. Sundaram MBM, Siemens P. Triphasic waves in Mollaret�smeningitis. J Neurol Neurosurg Psychiatr 1986;49:331.

9. Miller B, Brick J. Triphasic waves in a patient with car-cinomatous meningitis. Clin Electroencephalogr 1989;20:259–61.

10. Fisch BJ, Klass DW. The diagnostic specificity of triphasicwave patterns. Electroencephalogr Clin Neurophysiol1988;70:1–8.

11. Boulanger J-M, Deacon C, Lecuyer D, Gosselin S,Reiher J. Triphasic waves versus nonconvulsive statusepilepticus: EEG distinction. Can J Neurol Sci 2006;33:175–180.

12. Kruger H, Heim E, Schuknecht B, Scholz S. Acute andchronic neuroborreliosis with and without CNS involve-ment: a clinical, MRI, and HLA study of 27 cases. J Neurol1991;238:271–80.

13. Oschmann P, Dorndorf W, Hornig C, Schafer C, Wellensiek

HJ, PFLUGHAUPT KW. Stages and syndromes of neurobor-reliosis. J Neurol 1998;245:262–72.

Eriksson & Wictor

136