Long QT syndrome (LQTS) is characterized by the electrocardiographic appearance of a prolonged QT interval, an increased risk of developing an atypical polymorphic ventricular tachycardia (VT) known as torsades de pointes (TdP), and an increased risk for sudden cardiac death. Reduced levels of net repolarizing current secondary to a loss of function of outward ion channel currents, or a gain of function of inward currents, underlies the prolongation of the cardiac action potential and QT interval in both congenital and acquired forms of LQTS.
Amplification of spatial dispersion of repolarization, secondary to an increase of transmural and transseptal dispersion of repolarization, is believed to generate the principal arrhythmogenic substrate. Early after depolarization-induced triggered activity also contributes to the development of the substrate and provides the extrasystoles that trigger TdP arrhythmias observed in LQTS.1 Lengthening of the QT interval or Tp-e interval (from the peak to the end of the T wave), changes in the T wave morphology, and dispersion of QT interval measurements may identify and stratify patients at risk of ventricular arrhythmia. The condition is secondary to mutations in genes that affect the structure or function of ion channels responsible for cardiac depolarization or repolarization. Patients can suffer severe cardiac events resulting in syncope, seizures, and sudden cardiac death during times of physical and emotional stress and when exposed to certain drugs.
A 2.4-kg male infant presented with bradycardia, ventricular bigeminy, and poor peripheral perfusion at delivery by emergent cesarean delivery for fetal bradycardia. He was the product of a 36-week gestation to a 28-year-old African-American mother. The family history was notable for a 6-year-old male half sibling with borderline LQTS. The infant was tracheally intubated, and umbilical vascular access was secured in the neonatal intensive care unit. Subsequently, the infant developed episodic VT and TdP. Lidocaine boluses (1 mg/kg) were administered twice and an esmolol infusion started at 66 μg/kg/min. An echocardiogram at this time was notable for a structurally normal heart with severely diminished left ventricular function. An arterial blood gas was significant for a metabolic acidemia that was buffered with bicarbonate. The child was emergently transferred to the cardiac intensive care unit at the Children's Hospital for emergent electrophysiology consult and management.
The infant had electrocardiographic evidence of prolongation of the corrected QT interval of 775 to 800 ms, sinus bradycardia at 50 to 70 beats per minute (bpm) with 2:1 atrioventricular (AV) block, and episodic VT and TdP with evidence of low cardiac output. The TdP was treated with an amiodarone bolus (1 mg/kg/dose) and magnesium boluses. Calcium gluconate was used to enhance inotropy and the esmolol infusion discontinued. The next 24 hours were characterized by refractory VT and episodic TdP. The infant received a total of 11 doses of amiodarone at 3 mg/dose, 2 doses of lidocaine 1 mg/kg, magnesium 25 mg/kg per dose, calcium 10 mg/kg per dose and potassium supplementation at 1 mEq/kg per dose, and 3 episodes of direct current cardioversion. Serum calcium, magnesium, and potassium were maintained at normal levels. Mechanical circulatory support in the form of venoarterial extracorporeal membrane oxygenation (VA-ECMO) was considered for refractory arrhythmias; however, as the patient responded to medications and electrical therapy, the risk of VA-ECMO was deemed to outweigh the benefits. Dopamine was used to improve myocardial function.
Subsequently, lidocaine and esmolol infusions were commenced at 10 μg/kg/min and 30 μg/kg/min and increased gradually to 40 μg/kg/min and 200 μg/kg/min, respectively. External transcutaneous pacing was unsuccessful in improving the rhythm due to inadequate ventricular capture. His hemodynamics improved, with good urine output after addition of a milrinone infusion at 0.2 μg/kg/min. The child was neurologically intact by clinical examination. The head ultrasound was unremarkable. The underlying rhythm at this point was a slow sinus rhythm with 2:1 AV block. An isoproterenol infusion was commenced at 0.05 μg/kg/min to increase the heart rate and reduce the likelihood of pause-dependent arrhythmias. However, additional ventricular arrhythmias required further chemical (3 boluses of amiodarone 1 mg/kg, 2 boluses of lidocaine 1 mg/kg) and electrical cardioversion (3 shocks at 2 J/kg). A phenytoin bolus of 8 mg/kg was given, and the antiarrhythmic infusions were increased to 50 μg/kg/min of lidocaine and 250 μg/kg/min of esmolol. Serum lidocaine and phenytoin levels were therapeutic at 2.3 μg/mL (1.5 to 5.0 μg/mL) and 10.1 μg/mL (10 to 20 μg/mL), respectively.
In an effort to increase the ventricular rate and prevent TdP, epicardial pacing wire placement was performed via a limited sternotomy at the bedside at 4 days of age. The anesthetic involved fentanyl 5 μg/kg and vecuronium 0.1 mg/kg. The child received an IV bolus of lidocaine 1 mg/kg, phenytoin 5 mg/kg, and 5% albumin, 5 cc/kg, for VT associated with reduced perfusion during the procedure. The esmolol infusion was transiently increased to 500 μg/kg/min. Shortly after the procedure, the pupils were noted to be equal and reactive. Ventricular pacing was commenced at 76 bpm and gradually increased to 90 bpm with acceptable arterial blood pressure and perfusion. A follow-up neurologic examination in 4 hours was notable for fixed pupillary dilation and absence of any movement or gag reflexes. There was no evidence of intracerebral hemorrhage on an emergent brain ultrasound and computed tomography scan or seizure activity on an electroencephalogram. A neurology consult was obtained and a diagnosis of lidocaine toxicity compounded by drug–drug interactions was entertained despite normal serum lidocaine levels (3.5 μg/mL). The esmolol and lidocaine infusions were weaned over the course of 4 days; phenytoin dosing and magnesium supplementation were stopped. The patient was transitioned to oral propranolol (4 mg every 6 hours) and mexilitene (4 mg/kg every 8 hours). Abnormal neurological signs gradually, but completely resolved over the following 48 hours. Cardiac function by echocardiogram did recover after initiation of ventricular pacing and gradual weaning of antiarrhythmic therapy. Despite the small size and relative prematurity of the infant, a permanent pacemaker, set in a unipolar VVI mode at 110 bpm, was placed via a subcostal incision to control the heart rate and decrease the lethal ventricular arrhythmias. The anesthetic technique included pancuronium and isoflurane in oxygen, and was not associated with arrhythmias or hemodynamic compromise.
The infant was discharged home at 6 weeks of age, with a custom automatic external defibrillator (AED) with pediatric pads to deliver 25 J, and parents trained in cardiopulmonary bypass and AED use. Several weeks later, a blood sample drawn during the admission, sent to a specialist laboratory, revealed a compound mutation involving 2 LQT genes for LQT2 and LQT1. Ventricular function was low normal by echocardiogram with a shortening fraction of 36%. Frequent pacemaker interrogations in the first year of life revealed a few episodes of monomorphic VT, and 3 transient episodes of TdP not requiring AED use. These were typically associated with febrile illnesses, for which he is now routinely hospitalized and aggressively treated with ibuprofen. The pacemaker settings have been adjusted after these episodes. He is currently 21 months old with normal growth (he was and has remained on the fifth percentile for weight) and normal neurological development apart from febrile seizures. He remains in a ventricular-paced rhythm 60% of the time, the remainder being sinus rhythm. The doses of mexilitene and propranolol have been adjusted for growth, and he receives potassium supplementation. His serum electrolytes are maintained within normal limits.
The most frequent arrhythmias observed among neonatal patients with LQTS who are clinically symptomatic include sinus bradycardia, VT/Tdp, and AV block, particularly 2:1 block.2 In a nationwide survey regarding fetuses, neonates, and infants with LQTS in Japan, the majority of early onset VT/TdP patients were those with LQT2 or LQT3 genotype, whereas LQT1 patients typically manifested sinus bradycardia. Aggressive multimodal therapy with mexiletine and β blockers is recommended in the management of LQTS presenting with severe arrhythmias in the neonatal period.2 Early initiation of pacemaker placement is often necessary to increase the slow heart rates that make TdP more likely to occur. Because of technical constraints, implantable cardioverter defibrillator placement is rare in small infants, but has been attempted.
Our patient was a neonate with a compound mutation involving both LQT2 and LQT1 genes. This could explain the severity and refractory nature of malignant arrhythmias despite maximal pharmacotherapy. In this case, mechanical cardiac dysfunction is postulated to be a manifestation of the frequent episodes of TdP, compound pharmacotherapy with several drugs known to cause myocardial depression, and the frequent defibrillation applications for ventricular arrhythmias.
Lidocaine neurotoxicity may be confused with coma,3 stroke,4 and pupillary mydriasis in neonates.5 However, the symptoms and signs resolve once lidocaine is cleared from the system. There are several reasons this patient may have developed lidocaine toxicity. First, the relative prematurity of our patient would have been associated with immature end organ function of the liver, kidney, and brain, thus affecting pharmacokinetic mechanisms of drug distribution and excretion. In addition, neonates have reduced levels of α1 acid glycoprotein6 (0.2 to 0.3 g/L versus 0.7 to 1 g/L at 1 year of age) that binds lidocaine, and this can result in an increased free fraction of lidocaine. Amiodarone inhibits CYP3A4, the enzyme that metabolizes lidocaine, resulting in a reduction in lidocaine clearance.6 Phenytoin is a sodium (Na2+) channel blocker that blocks Na2+ channels in the motor cortex. The interactions between phenytoin and lidocaine could further augment neurotoxic side effects of either drug. Lastly, the infant was in a low cardiac output state during the perioperative period, which would have further compromised hepatic extraction of lidocaine.
We advise caution in the use of multimodal antiarrhythmic therapy in premature neonates and recommend constant vigilance for neurotoxicity. One major lesson to be learned from our case report is that medications that commonly treat ventricular arrhythmias, lidocaine and esmolol, may not be effective, and will just result in additive side effects if the underlying substrate (the bradycardia and extreme QT prolongation) are not improved. The most effective treatment in these babies with AV block associated with LQTS is to increase the heart rate, and this should have been introduced sooner in our patient. Although some may be reluctant to use isoproterenol in babies who are having VT/TdP, ventricular repolarization becomes more homogeneous, and pause-dependent runs of ventricular arrhythmias are decreased.
Early use of isoproterenol or pacing therapy might have avoided the need for aggressive lidocaine therapy and multimodal therapy by increasing the heart rate and avoiding pause-dependent arrhythmias. The outcome was positive owing to effective diagnosis and treatment of the arrhythmia and the neurotoxicity.
Name: Aruna T. Nathan, MBBS, FRCA.
Contribution: Performance of case, manuscript preparation.
Name: Maryam Naim, MD.
Contribution: performance of case, manuscript preparation.
Name: Lisa M. Montenegro, MD.
Contribution: Manuscript preparation.
Name: Victoria L. Vetter, MD, MPH.
Contribution: Manuscript preparation.
This manuscript was handled by: Peter J. Davis, MD.
1. Schimizu W. Long QT syndrome: therapeutic implications of a genetic diagnosis. Cardiovasc Res 2005;67:347–56
2. Horigome H, Nagashima M, Sumitomo N, Yoshinaga M, Yshinohama H, Iwamoto M, Shiono J, Ichihashi K, Hasegawa S, Yoshikawa T, Matsunaga T, Goto H, Waki K, Arima M, Takasugi H, Tanaka Y, Tauchi N, Ikoma M, Inamura N, Takahashi H, Shimizu W, Horie M. Clinical characterestics and genetic background of congenital Long-QT syndrome diagnosed in fetal, neonatal and infantile life. A nationwide questionnaire survey in Japan.. Circ Arrhythm Electrophysiol 2010;3:10–7
3. Bryant CA, Hoffman JR, Nichter LS. Pitfalls and perils of intravenous lidocaine. West J Med 1983;139:528–30
4. Bursell B, Ratzan RM, Smally AL. Lidocaine toxicity misinterpreted as a stroke. West J Emerg Med 2009;10:292–4
5. Berger I, Steinberg A, Schleisinger Y, Seelenfreund M, Schimmel MS. Neonatal mydriasis: intravenous lidocaine adverse reaction. J Child Neurol 2002;17:400–1
6. Mazziot J-X, Dalens BJ. Pharmacokinetics of local anesthetics in infants and children. Clin Pharmacokinet 2004;43:17–32