We present a case in which LQTS induced severe prenatal and neonatal arrhythmias. LQT3 was diagnosed (mutation R1623Q). Short-acting beta-blockers were ineffective as well as sotalol and mexiletine in preventing recurrent ventricular arrhythmias. An ICD was implanted at the age of 7 months (weight and length of the infant at implantation 6 kg and 60 cm respectively). Flecainide was prescribed in addition to the ICD implantation. After an appropriate shock the flecainide plasma levels were shown to be subtherapeutic. Readjustment of the flecainide dose resulted in adequate plasma levels. No further shocks occurred during a further 17 months follow-up period. The combination of an active can with a subcutaneous patch proved feasible, and lifesaving shocks occurred at 7 months after implantation.
Ventricular arrhythmias in neonates are rare, but may be related to structural abnormalities or long QT syndrome. The long QT syndrome (LQTS) is an inherited disorder characterized by prolongation of the QT interval and by life-threatening arrhythmias. Upto now, seven different genes have been identified. The clinical features of LQTS are different between genotypes. LQT1 patients have their first cardiac event at a mean age of 9 years, the trigger being most often exercise or emotional stress. LQT2 patients have their first cardiac event at a mean age of 12 years, usually triggered by exercise, emotional stress, or loud noise. LQT3 patients are usually older at their first cardiac event, with a mean age of 16 years. In these patients cardiac events usually occur during rest or sleep . Although less frequently, some patients present perinatally [2–,15]. Mortality rate among these patients is high . Cardiac events in LQTS patients can be significantly influenced by medication. For LQT1 and LQT2 patients betablockers are most effective. Their efficacy seems to be lower in LQT3 patients . Therefore, other drugs are being investigated, of which the orally bioavailable congener of lidocaine, mexiletine, shows possible beneficial effects on ECG abnormalities associated with the LQT3 mutation. Its role in clinical practice is as yet unclear [16–,18]. In LQTS patients with syncope or aborted sudden cardiac death, mortality risk may be as high as 14% in 5 years, even after institution of betablockers . In this high risk population implantation of ICDs may further reduce risk of mortality .
Our experience with problems associated with LQT3 in a neonate, and the subsequent implantation of an ICD at the age of 7 months, is presented.
A 35-year-old woman presented to her obstetrician at 33 weeks of gestation. Her medical and family histories were unremarkable. At a gestational age of 29 weeks an irregular foetal heart rate was noted by routine ultrasound. Subsequent echocardiography revealed no structural abnormalities of the foetus, but hydrothorax and decreased ventricular contractility were present. A Caesarean section was planned. After delivery the child had to be resuscitated and was mechanically ventilated. Birth weight was 2.42 kg. There were no other congenital malformations. An electrocardiogram showed periods of polymorphic ventricular rhythm with atrio-ventricular dyssynchrony, and periods of sinus rhythm with 4:3 block and Wenckebach periodicity (Fig. 1). During sinus rhythm QTc was prolonged (594 msec). A diagnosis of LQT3 was made (mutation R1623Q). Because of periods of ventricular arrhythmias lidocaine was given in a dose of 20 μg/kg/min, which was effective in restoring normal sinus rhythm. Ventricular function improved rapidly. Reducing lidocaine resulted in re-occurrences of an accelerated ventricular rhythm. Therefore, betablockers were instituted. In the first instance esmolol, a betablocker with a short half-life was given, but it was not effective in preventing the accelerated ventricular rhythm. Afterwards sotalol was given and it proved to be effective in restoring sinus rhythm. Occasional short periods of an accelerated ventricular rhythm with an acceptable frequency (140/min) were observed. Four weeks later the child had sustained ventricular arrhythmias again, while still hospitalized. Since esmolol and sotalol had not proved effective no other beta-blockers were used. Only lidocaine was effective, so mexiletine (24 mg three times daily), a sodium channel blocker, was started with good result. Appropriate plasma levels were obtained (6 mg/l). The electrocardiogram, while on mexiletine, showed regular sinus rhythm. During a period of 4 months the child did very well and no arrhythmias were noted. However, when at home sudden deterioration occurred and immediate resuscitation was commenced by the parents. During hospital admission, periods of torsades de pointes were seen (Fig. 2). The medication was changed to another Na+-channel blocker, flecainide 15 mg twice daily. Because of recurring ventricular arrhythmias necessitating cardiorespiratory resuscitation the implantation of an automatic internal defibrillator was planned. An ICD system was implanted under general anaesthesia at the age of 7 months. The small body size (weight 6 kg, length 60 cm) necessitated an epicardial approach. The ICD system consisted of a generator with an active can (Guidant Ventak Prizm DR HE 1853), a subcutaneous patch (Guidant Ventak standard patch model 0067, 14 cm2), and 2 bipolar epicardial pacing leads (Medtronic Capsure® epi 4968–25 cm). A median sternotomy with a short subxiphoid extension was performed. Both bipolar pace sensing electrodes were sutured to the epicardium of the right atrium and right ventricular anterior wall. A subcutaneous patch was placed under the left axilla. The active can ICD was implanted in a pocket in the right posterior rectus sheath (Fig. 3). The stimulation threshold for the atrial lead was 0.6 V (0.5 msec) and for the ventricular lead 2 V (0.5 msec). AAI pacing mode was programmed with a minimum rate of 110 to prevent bradycardia dependent ventricular tachycardia. Since it was almost impossible to fibrillate the patient, the defibrillator threshold could only be tested once, with success at 21 J, which is relatively high for the body weight of the patient. During follow-up the child remained well without any shocks until the age of 14 months, 7 months after implantation when the parents recognized a shock. This proved to be appropriate (Fig. 4). Later, in the intensive care unit 2 non-sustained VTs were recorded. With an increase of the dose of flecainide to 35 mg twice daily (5 mg/kg) these VTs disappeared. Although flecainide has been shown to provoke ST segment elevation in some LQT3 patients , this was not found in our patient. Flecainide plasma levels were below therapeutic (0.45–0.9 mg/l in our laboratory) when the shocks were delivered. After readjustment of the flecainide dose to the higher dose plasma levels were in the appropriate range and the patient has subsequently been free of any symptom for 17 months. The child shows normal psychomotor development.
This case shows several interesting features. In the first place, signs of the long QT-syndrome were already present prenatally. In the second place, an adequate dosage of mexiletine proved to be ineffective in this child with LQT3. Thirdly, placement of an ICD in this young child was uncomplicated over a follow-up period of 24 months. It was also lifesaving with appropriate shocks being given after 7 months.
The age at onset of symptoms of LQTS is dependent on the different types, and ranges from an average of 9 years in LQT1 to an average of 16 years in LQT3. Some patients with congenital LQTS, however, present prenatally. Abnormal prenatal findings are bradycardia with or without AV-block and ventricular tachycardias. We found, in the literature, 22 other cases that have been observed prenatally [2–,12]. The gestational age at the first echocardiogram ranges from 22 to 40 weeks. In 16 cases the main finding was a foetal bradycardia, while 6 foetuses had signs of ventricular tachycardia. Five children of the 22 died in the neonatal period. In 4 patients a Caesarean section was performed. In most patients, however, an uncomplicated neonatal course was noted. Bradycardia in neonatal LQTS may be due to sinus bradycardia or to atrioventricular block; the latter is associated with HERG mutations and poor prognosis . Patients in whom LQTS induced arrhythmias present in the neonatal period may also present with sudden death as a first event. If the diagnosis of LQTS has been unknown beforehand an incorrect diagnosis of sudden infant death syndrome (SIDS) could be made. An association between SIDS and LQTS has been described for both LQT1  and LQT3 . Furthermore, an association has been found between QT prolongation in the first week of life and the risk of SIDS during follow-up . Neonatal electrocardiography may identify such QT prolongation, but implementation of widespread screening is as yet controversial .
In our patient, a Caesarean section was necessary because of signs of intrauterine cardiac failure resulting in hydrops. The neonatal course was characterized by diminished ventricular contractility and dependence on inotropic support for a few days.
Efficacy of mexiletine
In the 22 patients with LQTS diagnosed prenatally, therapies used included beta-adrenergic blocking agents, mexiletine and/or insertion of a pacemaker. Our patient had a rare mutation (R1623Q), which has been characterized as an essentially severe LQT3 subtype [25,,26]. One patient with this LQT3 subtype, described before, developed a potentially lethal ventricular arrhythmia, torsades de pointes at 35 days after birth, and was successfully treated with the Na+ channel blocker mexiletine .
In a few LQT3 patients the sodium channel blocker mexiletine was shown to reduce the QT interval and normalize the morphology of the T wave. In an animal model mexiletine was shown to be very effective in abbreviating the QT interval and transmural dispersion of repolarization, thereby reducing the incidence of arrhythmogenesis in LQT3 [17,,18]. These observations, however, concern the acute effects of mexiletine, while long-term follow-up studies are so far lacking.
In our patient we succeeded for a short period, but the therapy then failed, despite adequate plasma levels. In one other case report Trippel et al. used mexiletine to treat a neonate with LQTS but the child died suddenly shortly after its initiation . We did not find further evidence of the efficacy or failure of mexiletine in the neonatal period. Whether the mexiletine failure in our patient is related to the R1623Q-mutation remains speculative. Flecainide is a stronger sodium channel blocker, and with adequate plasma levels, our patient is now free of symptoms for 17 months.
In addition to medical treatment with Na+-channel blockers cardiac pacing to prevent bradycardia and ICD-implantation have been advised in patients with LQTS-associated ventricular arrhythmias. Recently, cardiac pacing has been described to be successful over a period of 38 years . Implantable cardioverter-defibrillators are used when the combination of the beta-adrenergic blocker therapy and the pacing fails to prevent presyncopal or syncopal episodes or when the initial presenting event is a resuscitated cardiac arrest. Since the shocks of a cardioverter-defibrillator can produce emotional distress and thereby induce new arrhythmias it is essential to reduce these shocks to a minimum by medical treatment. Technological advances have increased the applicability of ICDs to the paediatric population. Several series have now evaluated the use of these devices in children [28–,30]. Kron et al. reported their experience in 40 children, of whom only 3 were 10 years or less . In this group the most prevalent diagnosis was dilated or hypertrophic cardiomyopathy, while congenital long QT syndrome accounted for 7/40 implantations. In this series the 33 month survival was 82%. A more recent series of Chatrath et al. described 16 patients with an age of 10–21 years . The main diagnoses in this series were cardiomyopathy in 6 patients and congenital long QT syndrome in another 6 patients. In this series no mortality occurred during a follow-up period of 36 months, while 7 patients received appropriate ICD shocks. Groh et al. described their experience with ICD implantation in 35 high risk LQTS patients . During a follow-up period of 31 months they found 60% appropriate discharges and no mortality. Despite these successful series of ICD placement in young patients the use of an ICD in infants is still very limited [31–,35]. Reasons for ICD implantation in these patients were rhythm disturbances induced by cardiac tumors [31,,33], or primary electrical diseases such as Brugada syndrome , or LQTS [34,,35]. In one of the LQTS patients frequent appropriate shocks probably led to myocardial damage and eventual death .
Both an epicardial patch or a subcutaneous patch can be used in preventing the necessity of transvenous access. Revision of the system during growth is to be expected. In our experience, no technical difficulties were encountered with a subcutaneous patch and an active can.
In LQTS patients with an ICD and multiple discharges, left sympathectomy may significantly reduce the number of shocks . In high-risk LQTS patients without an ICD this therapy is associated with a significant reduction in the incidence of aborted cardiac arrest and syncope .
In this heavily symptomatic LQT3 patient (6 kg, 60 cm), in which beta-blockers and mexiletine treatment proved ineffective ICD implantation was a feasible and safe treatment.
We thank M. Alders, clinical genetics, Academic Medical Center, Amsterdam for genotyping our patient.