Case Report

Should every patient who presents with a seizure have an electrocardiogram?

Abstract

syncopal episodes of any cause that generate long lasting brain hypoperfusion can cause seizures and could be interpreted mistakenly as primary epilepsy. In some cases, syncopes are due to ventricular arrhythmias in the setting of genetic diseases of cardiac ion channels (channelopathies) such as long QT syndrome (LQTS), short QT syndrome, and Brugada syndrome. These affect ventricular repolariza- tion, with high risk of sudden death, and are usually recognized in the standard electrocardiogram (ECG). We present the case of a 23-year-old woman who was admitted for seizures in a specialized neurology center and was initially managed as primary epilepsy. Finally, she was diagnosed as having Polymorphic ventricular tachycardia due to LQTS. We present a review of the literature and discuss the clues of the differential diagnosis between seizures and syncope. We describe the LQTS and its prognosis and discuss why every patient admitted for seizures should have an ECG done.

A 23-year-old woman was admitted in the emergency department at a neurology-specialized center because of a seizure episode. This started when the patient woke up with the loud noise of her alarm clock going off. The seizure lasted about 5 minutes and was accompanied by urinary incontinence and partial amnesia. After that, the patient presented palpitations and chest pain, which finished with vomiting.

Her medical history was significant for an episode of loss of consciousness when she was 12 years old. Her family history showed that her father had died suddenly at the age of 50.

During her stay at the emergency department, she had another seizure, preceded by palpitations. The case was interpreted by the neurologist as primary epilepsy, and the patient was treated with intravenous diphenylhydantoin (20 mg/kg). A brain magnetic resonance imaging and an electroencephalogram were performed, and both were normal. The chemistry and cerebrospinal fluid were normal. She was admitted in the neurology section, and during the first 24 hours, she presented a third seizure episode. A few hours later, she complained of palpitations, near syncope,

and nausea. Blood pressure was 70/40 mm Hg, and pulse rate was weak and fast. An ECG was performed. It showed sustained ventricular tachycardia (VT) that was successfully cardioverted (Fig. 1). After cardioversion, a new ECG showed a severely prolonged QTc interval (710 milliseconds) (Fig. 2). The patient was admitted to the coronary care unit where she presented recurrent VT that required multiple cardioversions. The review of the ECG performed at admission showed an extremely prolonged QT interval (Fig. 3). She was started on intravenous magnesium sulfate and intravenous esmolol, but VT episodes recurred many times. A temporal pacemaker lead was inserted, and the patient was paced at a rate of 110 beats per minute. This abolished the VT. The second day she started oral atenolol. The QTc interval remained prolonged, but there was no VT. As the QTc interval remained severely prolonged, oral mexiletine was started, with a partial shortening of the QTc interval, which remained above 500 milliseconds. In view of the patient’s high risk of Sudden Cardiac Death besides treatment with ?-blockers, an implantable cardiac defibrilla- tor was implanted without complications.

In the follow-up, the patient had no further episodes without antiepileptic medication. She had neither VT episodes nor implantable cardiac defibrillator therapies. Screening performed on her family detected a long QTc interval in her mother and in a maternal uncle. genetic testing was not available.

Secondary seizures after a syncope episode, the so-called convulsive syncope, have been recognized for a long time and may account for up to 12% of syncope episodes [1]. The differential diagnosis between syncope and seizures is a common problem in clinical practice, which became harder when syncope generate a true seizure (convulsive syncope). There have been described convulsive syncope as the presentation form of vasovagal syncope, syncope due to bradycardia, and syncope due to ventricular tachycardia of different etiologies [2-3]. The common finding of these episodes is Brain hypoxia due to global hypoperfusion. Particularly in LQTS and other channelopathies (short QT syndrome and Brugada syndrome), the presence of seizures in a family member move us toward the diagnosis of primary epilepsy. The problem is more complex if we consider that there are some forms of epilepsy that cause secondary syncope (ictal bradycardia, temporal lobe com- plex partial seizure), in which the firing of the temporal

0735-6757/$ – see front matter (C) 2009

Fig. 1 Electrocardiogram precardioversion.

lobe generate an intense vagal reflex with prolonged asystole and secondary syncope [4]. Finally, there are true epilepsy cases that present sudden death, which mechanism is not fully understood [5]. So, the differential diagnosis between seizure and syncope is not always easy.

In the case we presented, the patient was admitted in a neurology-specialized center, with an epilepsy clinic, and

she got a complete and expensive workup. Besides that, the final diagnosis was wrong, and the patient’s life was in serious danger.

The differential diagnosis between syncope and epilepsy start with some clues of the history. The presence of diaphoresis and pallor previous to the loss of conscious- ness or some prodromic symptoms and typical triggers

Fig. 2 Electrocardiogram postcardioversion.

Fig. 3 ECG on admission.

(long-standing position, blood extraction, and others) favor the diagnosis of syncope. In the same way, palpitations or chest pain preceding the episode alert us of the possibility of syncope due to cardiac causes. Beginning at young age, previous syncope or presyncope, the family history of syncope or sudden cardiac death, the presence of specific triggers such as stress, exercise, loud noises, swimming or diving, and others are common findings in genetic arrhythmogenic diseases. At physical examination during the seizure, the absence of pulse gives us the clue of a hemodynamic failure. In these cases, the EEG is typically normal and the ECG can give us the final diagnosis. On the other side, the presence of an aura preceding the episode and confusion after spell are typical of seizure episodes. The EEG has an acceptable sensibility for the diagnosis of epilepsy if it is done into the first 24 hours [6].

The diagnosis of genetic disease of cardiac ionic channels is relevant because they have a high risk of sudden death and they have specific treatments that can prolong patients’ life.

Long QT syndrome is a genetic disease characterized by lengthening of the QT interval and a high risk of sudden death due to polymorphic ventricular tachycardia (the so- called Torsades de pointes). The cause of the syndrome is a mutation in the gene that encodes ion channels that participate in the repolarization process of the ventricular cells. There is an autosomic dominant with variable penetrance form (Romano-Ward syndrome) and a less frequent autosomic recessive form with congenital deafness (Jervell and Lange-Nielsen syndrome). There are also some

sporadic cases. There are 7 subtypes of the disease that have been described. The 3 more frequent subtypes are as follows: LQT1 and LQT2 present mutations in the genes that encode potassium channels that carries the repolariza- tion currents IKs and IKr, producing a loss of function of these current, slowing the repolarization process and LQT3 presents a mutation of the gene that encodes the sodium channel SCN5A, with an increase of the sodium current and persistence of the plateau of the ventricular action potential. These different forms of the disease have differential characteristics in their specific triggers, electro- cardiographic characteristics, response to exercise, drug response, and prognosis that have a great clinical relevance (Table 1) [7-8].

The diagnosis of LQTS is based on the association of a prolonged QTc interval with some clinical characteristics. Some secondary causes of QT prolongation should be ruled out (hypocalcemia, hypothyroidism, drugs). A complete list of drugs that cause QT prolongation can be found at www. torsades.org. Schwartz et al [9] have developed diagnostic criteria for LQTS in the form of a score that group the patients as at low probability of LQTS (score <= 1), intermediate (score 2 or 3), and high probability or definite LQTS (score >= 4) (Table 2).

Besides the technique of QT measurement is well defined

[10] (Table 3), it is frequently not properly done. In a challenging study, Viskin et al [11] asked 902 doctors to measure the QT and the QTc interval of 4 ECG tracings (2 with LQTS and 2 normal). Of LQTS experts, 96% could classify the 4 patients properly. However, it was achieved only by 62% of arrhythmia expert and 25% of cardiologist

LQT1		LQT2	LQT3
Gene mutation	KCNQ1	KCNH2	SCN5A
Ionic current affected	?Iks	?Ikr	?Plateau INa
Typical ECG finding	Broad T wave	Low amplitude notched T wave	Long isoelectric ST segment
Triggers	Emotional or physical stress;	Emotional or physical stress;	Rest, sleep
	swimming, diving	sudden loud noise
ECG at onset of VT	No pause	Pause	Not defined
QT change with exercise	No shortening	Normal shortening	Supranormal shortening
Clinical response to ?-blockers	Yes	Less than LQT1	Uncertain
Response to mexiletine	No	No	Yes
Prognosis	More events	More events	More lethal
IKs: slow component of the potasium current. IKr: rapid component of the potasium current. Plateau INa: Sodium current during the action potential plateau.

and noncardiologist. In a similar way, Taggart et al [12], analyzed ECG tracings of 176 patients sent to Mayo Clinic’s LQTS clinic with a diagnosis of LQTS and reclassify the patients as definite LQTS, possible LQTS, and no LQTS, based on Schwartz criteria. Only 27% were definite LQTS, with 41% with no LQTS. Most of the wrong diagnosis were due a wrong measure of the QT interval or because the diagnosis was made based on borderline QTc prolongation. The real prognosis of LQTS is in some way uncertain because most of the information comes from retrospective series of a few reference centers, with a probable bias to the

Table 1 Differential characteristics of the 3 major LQTS subtypes

more severe cases.

The International LQTS Registry started in 1979, and in 1991, Moss et al [13] reported the course of 328 LQTS families, with 3343 patients (1008 patients affected). Of the affected members, 147 died before the age of 50. Mean age at death was 21 years. After 10 years of follow-up, 37% of

Table 2 Diagnostic criteria for LQTS

Family history

1. Family members with definite LQTS 1
2. Unexplained SCD b30 among immediate family 0.5

members

Points

Criteria

the probands, 5% of the affected family members, and less than 1% of the unaffected family members had experienced a cardiac event (syncope or SCD). The factors associated with cardiac events in the total population were the QTc interval, the history of cardiac event, and the heart rate. In a later analysis, they show that among LQTS patients, the risk of cardiac events was higher in males until puberty and higher in females during adulthood [14]. In 1998, they analyzed the cumulative incidence of cardiac events (syncope, aborted sudden death, or death) from birth to age 40, according to genotype. Total cardiac events were more frequent in LQT1 (63%) and LQT2 (46%) than in LQT3 (18%), but the lethality (% of cardiac events that were lethal) was superior in LQT3 (20%) than in LQT1 and LQT2 (4% both) [15]. Finally, the same group analyzed the evolution of patients that reached adulthood (N18 years). They found female sex, QTc interval, LQT2 genotype, and frequency of cardiac events before 18 years old were predictive of any cardiac events. When they analyzed only

ECG finding A. QTc >=480 ms	3
460-479 ms	2
450 ms (males)	2
B. Torsades de pointes	2
C. T-wave alternans	1
D. Notched T waves in 3 leads	1
E. Low heart rate for age	0.5
Clinical history A. Syncope With stress	2
Without stress	1
B. Congenital deafness	0.5

Table 3 How to measure QTc How to measure the QT interval

Use a standard 12-lead ECG at 25 mm/s and 10 mm/mV Measure the QT interval from the beginning of the earliest onset

of the QRS to the end of the T wave

The end of the T wave is considered when it crosses the baseline (TP segment)

When you cannot see T wave’s end, prolong its descending limb until it crosses the baseline

When there are broad U waves, they should not be considered as part of QT interval

The QT interval should be determined as the mean value between 3 and 5 cardiac cycles

Measure the QT interval in leads II and V₅ or V₆. Use the longest of these values

Correct the QT interval for the heart rate; the most used formula is the Bazett’s formula (QTc = QT/RR^1/2)

aborted cardiac arrest or death, the predictors were female sex, QTc of more than 500 milliseconds, and syncope after the age of 18. ?-Blocker use was associated with a risk reduction (hazard ratio, 0.6) [16].

In a different set of genotyped patients, Priori et al [17] found a global 13% rate of cardiac arrest or SCD before 40 years old. These rates were 20% for LQT2, 16.4% for LQT3, and 10% in LQT1. They found the worst prognosis in female LQT2 patients and male LQT3 patients. Sex had no influence in LQT1 patients. QTc interval was another predictor of events in LQT1 and LQT2 but not in LQT3. In conclusion, they consider high-risk patients (N50% of cardiac events) those with QTc of more than 500 milli- seconds and LQT1 or LQT2, or males with LQT3.

There are a lot of drugs that can produce QTc prolongation, and the measurement of QTc interval is part of the routine analysis in the development of a new drug. Fortunately, usual Antiepileptic drugs are not associated with QTc prolongation. Moreover, in this case, seizures were clearly associated with VT episodes and were present on admission, when the patient was free of drugs.

As we showed, syncope and epilepsy are a common and difficult differential diagnosis, and there are some over- lapping between them that is not always easy to clarify (convulsive syncope, ictal bradycardia, sudden death in epilepsy). The one that our patient presents (convulsive syncope) is a relative frequent form of presentation of syncope. Some of them are due to severe cardiac causes with a high risk of sudden death that are diagnosed with a simple resting ECG. Every doctor should know these ECG patterns [18]. On the other hand, ECG is a simple and cheap test, nothing compared with the usual work up of patients with seizures. So, we believe there is no reason to not perform an ECG to every patient admitted for a seizure- like episode.

Gustavo R. Iralde MD cardiology department, FLENI Buenos Aires, Argentina

Hernan Cohen Arazi MD Cardiology Department, FLENI Buenos Aires, Argentina

Silvina Waldman MD Cardiology Department, FLENI Buenos Aires, Argentina

Mauricio Abello MD Cardiology Department, FLENI Buenos Aires, Argentina

David Doiny MD Cardiology Department, FLENI Buenos Aires, Argentina

Ricardo Spampinatto MD Cardiology Department, FLENI Buenos Aires, Argentina

Mario Russo MD Cardiology Department, FLENI Buenos Aires, Argentina

Claudio Pensa MD Cardiology Department, FLENI Buenos Aires, Argentina

Hugo Grancelli MD Cardiology Department, FLENI Buenos Aires, Argentina

doi:10.1016/j.ajem.2008.08.010

References

1. Lin JT, Ziegler DK, Lai CW, et al. Convulsive syncope in blood donors. Ann Neurol 1982;11:525-8.
2. Pacia SV, Devinsky O, Luciano DJ, et al. The prolonged QT syndrome presenting as epilepsy: a report of two cases and literature review. Neurology 1994;44(8):1408-10.
3. Zarraga IG, Ware DL. Syncope, seizure, or both? An unusual case of complete heart block. J Electrocardiol 2007;40(6):493-5.
4. Locatelli E, Varghese J, Shuaib A, et al. Cardiac asystole and bradycardia as a manifestation of left temporal lobe complex partial seizure. Ann Intern Med 1999;130:581-3.
5. Monte’ C, Arends J, Tan I, et al. Sudden unexpected death in Epilepsy patients: risk factors. A systematic review. Seizure 2007;16:1-7.
6. McKeon A, Vaughan C, Delanty N. Seizure versus syncope. Lancet Neurol 2006;5:171-80.
7. Roden D. Long-QT syndrome. N Engl J Med 2008;358:169-76.
8. Schwartz P, Priori S, Spazzolini C, et al. Genotype-phenotype correlation in the long-QT syndrome. Gene-specific triggers for life- threatening arrhythmias. Circulation 2001;103:89-95.
9. Schwartz P, Moss A, Vincent M, et al. Diagnostic criteria for the long QT syndrome. An update. Circulation 1993;88(2):782-4.
10. Goldenberg I, Moss A, Zareba W, et al. QT interval: how to measure it and what is “normal”. J Cardiovasc Electrophysiol 2006;17:333-6.
11. Viskin S, Rosovski U, Sands AJ, et al. Inaccurate electrocardiographic interpretation of long QT: the majority of physicians cannot recognize a long QT when they see one. Heart Rhythm 2005;2(6):569-74.
12. Taggart NW, Haglund CM, Tester DJ, et al. Diagnostic miscues in congenital long-QT syndrome. Circulation 2007;115(20):2613-20.
13. Moss A, Schwartz P, Crampton R, et al. The long QT syndrome. Prospective longitudinal study of 328 families. Circulation 1991;84: 1136-44.
14. Locati E, Zareba W, Moss A, et al. Age- and sex-related differences in clinical manifestations in patients with congenital long-QT syndrome. Findings from the international LQTS registry. Circulation 1998;97: 2237-44.
15. Zareba W, Moss A, Schwartz P, et al. Influence of the genotype on the clinical course of the long QT syndrome. N Engl J Med 1998;339:960-5.
16. Sauer A, Moss A, Mc Nitt S, et al. Long QT syndrome in adults. J Am

Coll Cardiol 2007;49:329-37.

1. Priori S, Schwartz P, Napolitano C, et al. Risk stratification in the long- QT syndrome. N Engl J Med 2003;348:1866-74.
2. Brugada P, Geelen P. Some electrocardiographic patterns predicting sudden cardiac death that every doctor should recognize. Acta Cardiol 1997;52(6):473-84.