Abstract
Therapeutic hypothermia (TH) is used in neuroprotection following cardiac arrest due to ventricular tachycardia (VT) and ventricular fibrillation (VF). Accidental hypothermia is itself known to cause prolongation of the corrected QT interval (QTc). QTc prolongation can cause polymorphic VT and VF. If this also occurs in TH, it may induce refibrillation. We investigated the effect of TH on the QTc interval.
Prospective case series of all patients undergoing TH following cardiac arrest following VT/VF at our hospital between July 2008 and January 2009. We studied the effect of temperature on QTc. All electrocardiograms (ECGs) undertaken during TH were studied and compared with the ECG prior to this. Four patients underwent TH. A total of 10 ECGs were undertaken during TH. The QTc was normal prior to TH. It became prolonged (>460 ms) in all cases during TH and normalized after cessation of TH, apart from Patient 4 who did not have an ECG post-TH since she died from cardiogenic shock. There was a negative correlation between temperature and QTc (Pearson's correlation coefficient, r = −0.71).
Our series illustrates QTc prolongation during TH. This carries potential for refibrillation. Guidelines on ECG monitoring during TH are needed, especially since hypothermic myocardium is intrinsically prone to arrhythmias and commonly used antiarrythmic drugs such as amiodarone can prolong the QTc.
Introduction
Therapeutic hypothermia (TH) provides valuable neuroprotection following cardiac arrest due to ventricular tachycardia (VT) and ventricular fibrillation (VF). It has been introduced into the 2002 International Liaison Committee on Resuscitation (ILCOR) Guidelines 1 and 2005 European Resuscitation Council (ERC) Guidelines. 2 Since the introduction of TH to our hospital guidelines, four patients have undergone this following cardiopulmonary resuscitation (CPR) after out-of-hospital cardiac arrest due to VT or VF. However, hypothermia itself is known to cause potentially arrhythmogenic effects. These have been shown in animal models of hypothermia and in human studies of accidental hypothermia, and include QTc prolongation. This itself can cause polymorphic VT. 3–19 We sought to investigate whether QTc prolongation also occurs during TH following VT or VF in order to identify areas for further study and improvement in ensuring the safe use of this technique.
Case reports
Patient 1 was a 49-year-old man with no previous medical history. His electrocardiograms (ECGs) showed sinus rhythm with T-wave inversion in the lateral leads following CPR and defibrillation for out-of-hospital cardiac arrest due to VT. Initial serum potassium was 3.3 mmol/L and promptly corrected. He was transferred to the regional tertiary centre for implantation of an implantable cardioverter-defibrillator (ICD) on Day 7 of the admission.
Patient 2 was a 51-year-old man with no previous cardiac history whose ECG after cardiac arrest due to VF revealed new left bundle branch block (LBBB). He was promptly thrombolysed for a presumed diagnosis of VF secondary to acute myocardial infarction (ST elevation myocardial infarction, STEMI). Subsequent ECGs showed persistence of LBBB. He was transferred for implantation of an ICD on Day 9 of the admission.
Patient 3 was a 65-year-old man with angina. His ECG after cardiac arrest revealed anterior wall Q-waves. He was commenced on an intravenous amiodarone infusion for the first 24 h. Echocardiography revealed anterior regional wall motion abnormalities. It was unknown whether these were longstanding or recent. It was deemed that these were most likely new since he had no previous history of myocardial infarction or heart failure, and that his VF arrest was thus likely secondary to non-STEMI. He failed to improve and died from cardiogenic shock on Day 7 of his inpatient stay.
Patient 4 was a 59-year-old lady with no previous cardiac history. Ventricular fibrillation followed chest pain. Her post-arrest ECG revealed new LBBB. Despite thrombolysis for presumed acute STEMI, TH, and inotropic support, she failed to regain consciousness and developed rapidly deteriorating pump failure and cardiogenic shock. It was decided at this point to withdraw active treatment (TH had been completed 8 h previously). She developed intractable VF ∼12 h later and died soon after this, on Day 2.
Patients 1 and 2 had an ICD implanted following inpatient coronary angiography and appropriate revascularization, and recovered well. All four cardiac arrests were thought to be due to ischaemic coronary events.
Methods
The patients underwent TH at 32–34°C for 24 h on the intensive care unit following return of spontaneous circulation (ROSC). This was achieved using the ‘Laerdal Medicool Kit’ (Laerdal UK™, consisting of cap, blanket, groin, and axillary pads containing a water-absorbent cooled co-polymer) throughout hypothermia, and cooled intravenous saline (at 4°C) for its first 30 min. The QTc was calculated on the ECG undertaken prior to commencement of hypothermia, on all ECGs during hypothermia and on the first ECG after its completion. Bazett's formula was used [QTc (ms) = QT interval (ms)/√R-R interval (s)]. Serum potassium and calcium levels were analysed during hypothermia, and within 24 h before and after hypothermia, since abnormalities in these electrolytes can affect the QTc. 5 QTc was compared against temperature for each patient.
Statistical analysis
Regression analysis was undertaken on the overall analysis of QTc vs. temperature using Pearson's correlation coefficient (PCC value r , where −1 and +1 indicate perfect negative and positive correlations, respectively, between temperature and QTc).
Results
Relationship between temperature and QTc
An ECG was undertaken twice for Patients 2 and 3, and three times for Patients 1 and 4 during hypothermia. A significant increase in the QTc was seen in each of these 10 ECGs ( Table 1 ). All but one were now in the range for prolonged QTc (>460 ms). All four patients had normal QTc values prior to hypothermia and the QTc returned back into the normal range after it was discontinued. This was apart from Patient 4 who did not have an ECG post-hypothermia, due to withdrawal of active treatment and thus 12-lead ECG monitoring following rapidly worsening cardiogenic shock, after completion of hypothermia ( Figures 1 and 2 ).
Pre-TH ECG undertaken for Patient 1 (QTc 430 ms).
ECG 2 undertaken during TH for Patient 1 (QTc now increased to 565 ms).
QTc values before, during, and after TH, with percentage increase in QTc during TH compared with ‘pre-TH’ (in brackets)
| Patient | Pre-TH QTc 1 | QTc 1 during TH | QTc 2 during TH | QTc 3 during TH | Post-TH QTc 1 |
|---|---|---|---|---|---|
| 1 | 430 ms | 518 ms (+24.5%) | 565 ms (+35.8%) | 478 ms (+14.9%) | 430 ms (20 h) |
| 2 | 438 ms | 442 ms (+3.1%) | 463 ms (+5.7%) | 410 ms (28 h) | |
| 3 | 434 ms | 503 ms (+15.9%) | 491 ms (+13.1%) | 387 ms (69 h) | |
| 4 | 427 ms | 503 ms (+17.8%) | 517 ms (+21.1%) | 505 ms (+18.2%) | N/A |
| Patient | Pre-TH QTc 1 | QTc 1 during TH | QTc 2 during TH | QTc 3 during TH | Post-TH QTc 1 |
|---|---|---|---|---|---|
| 1 | 430 ms | 518 ms (+24.5%) | 565 ms (+35.8%) | 478 ms (+14.9%) | 430 ms (20 h) |
| 2 | 438 ms | 442 ms (+3.1%) | 463 ms (+5.7%) | 410 ms (28 h) | |
| 3 | 434 ms | 503 ms (+15.9%) | 491 ms (+13.1%) | 387 ms (69 h) | |
| 4 | 427 ms | 503 ms (+17.8%) | 517 ms (+21.1%) | 505 ms (+18.2%) | N/A |
The post-TH QTc was obtained from the first ECG taken after TH was discontinued, with the time after discontinuation in brackets. Normal QTc is <460 ms.
QTc values before, during, and after TH, with percentage increase in QTc during TH compared with ‘pre-TH’ (in brackets)
| Patient | Pre-TH QTc 1 | QTc 1 during TH | QTc 2 during TH | QTc 3 during TH | Post-TH QTc 1 |
|---|---|---|---|---|---|
| 1 | 430 ms | 518 ms (+24.5%) | 565 ms (+35.8%) | 478 ms (+14.9%) | 430 ms (20 h) |
| 2 | 438 ms | 442 ms (+3.1%) | 463 ms (+5.7%) | 410 ms (28 h) | |
| 3 | 434 ms | 503 ms (+15.9%) | 491 ms (+13.1%) | 387 ms (69 h) | |
| 4 | 427 ms | 503 ms (+17.8%) | 517 ms (+21.1%) | 505 ms (+18.2%) | N/A |
| Patient | Pre-TH QTc 1 | QTc 1 during TH | QTc 2 during TH | QTc 3 during TH | Post-TH QTc 1 |
|---|---|---|---|---|---|
| 1 | 430 ms | 518 ms (+24.5%) | 565 ms (+35.8%) | 478 ms (+14.9%) | 430 ms (20 h) |
| 2 | 438 ms | 442 ms (+3.1%) | 463 ms (+5.7%) | 410 ms (28 h) | |
| 3 | 434 ms | 503 ms (+15.9%) | 491 ms (+13.1%) | 387 ms (69 h) | |
| 4 | 427 ms | 503 ms (+17.8%) | 517 ms (+21.1%) | 505 ms (+18.2%) | N/A |
The post-TH QTc was obtained from the first ECG taken after TH was discontinued, with the time after discontinuation in brackets. Normal QTc is <460 ms.
As temperature decreased, the QTc became more prolonged in each patient ( Figure 3 ). There was a clear negative correlation between these variables, illustrated by a PCC value ( r ) of −0.71 ( R2 = 0.507) ( Figure 4 ). Indeed in each patient, the longest QTc was seen at the lowest temperature during hypothermia.
QTc vs. temperature for each patients. As temperature decreased, QTc became more prolonged. The longest QTc occurred at the lowest temperature during hypothermia.
QTc vs. temperature for the four patients combined illustrating the inverse relationship between temperature and resulting QTc and corresponding PCC ( r ) of −0.71. Individual points taken from data in Table 1 .
Apart from Patients 2 and 4 whose ECGs illustrated LBBB, the ECGs of the other two patients all revealed normal QRS widths (<120 ms) before, during, and after TH. Osborn (J) waves were not evident in the ECGs performed during hypothermia in any of the patients. There was no evidence of high-degree heart block (second- or third-degree), bradyarrhythmia, or tachyarrhythmia in the ECGs recorded in the four patients.
Relationship between serum potassium and calcium and QTc
Initial serum potassium was slightly low at 3.3 mmol/L prior to cooling in Patient 1. All other measurements of serum calcium and potassium were within normal limits in all patients.
Discussion
Neurological damage is a feared and serious complication in cardiac arrest survivors. In order to improve neurological outcome, TH was introduced into the 2002 ILCOR Guidlines 1 and 2005 ERC Guidelines. 2 It is has now become a standard practice for unconscious patients to be cooled to the recommended 32–34°C for 12–24 h after ROSC following CPR for out-of-hospital cardiac arrest due to VT or VF. 1–4
However, hypothermia itself is known to cause numerous potentially arrhythmogenic cardiovascular and electrophysiological effects, both when therapeutic and accidental. Our study is the first to report a case series illustrating QTc prolongation in successive patients undergoing TH following cardiac arrest due to VT or VF. Prior to commencing hypothermia, they all had normal QTc intervals, which on cooling entered the range for prolonged QTc (>460 ms). Indeed, in 6 of the 10 ECGs during hypothermia in our series, the QTc was >500 ms. This is particularly worrying since QTc prolongation is itself a cause of polymorphic VT [Torsades de Pointes (TDP)] and VF. This is especially the case during hypokalaemia and hypomagnesemia, which are known to occur during hypothermia. 10 , 11 A case report of TDP 13 and idioventricular rhythm 14 have been described in TH and a recent case report demonstrated QTc prolongation (590 ms) with Osborn (J) waves and subsequent refibrillation during TH. 15
Atrial fibrillation with slow ventricular rate. 5
The Presence of J-waves (Osborn waves). 5
Bradycardias, including junctional and even asystole. 5
Prolongation of PR, QRS, and QTc intervals. 5
Premature ventricular beats, VT, and VF. Lower thresholds for fibrillation (72% lower) have been demonstrated in animal models of hypothermic ventricular myocardium in swines. 6 This has the potential to cause refbrillation, especially when cooling is commenced immediately after circulation is restored. 1 , 2 , 5
The effect of hypothermia on success of defibrillation is less clear. Bodicker et al.7 illustrated significantly improved defibrillation success in hypothermic swine myocardium. However, Ujhelyi et al.6 showed no differences in defibrillation energy requirement in swine myocardium once cooled. Further research is needed into optimization of electrical defibrillation and cardioversion in hypothermic states.
Electrophysiological changes in hypothermia are similar to those is ischaemia. Since most VT and VF cardiac arrests are secondary to ischaemic events, some have proposed that hypothermia and ischaemia can induce and potentiate one another. 2 , 6
Reduced efficacy of inotropic drugs and antiarrythmics including amiodarone and bretylium has been in animal studies (dogs) during hypothermic VF. 8 , 9
The most commonly used antiarrythmic drug in VT and VF is amiodarone. Its Class III effects at the potassium channel are known to cause lengthening of the action potential which is seen as a prolongation of the QT interval. 12 , 16 Indeed, although uncommon, amiodarone has been shown to cause TDP due to QTc prolongation according to data collated by the World Health Organization Drug Monitoring Centre. During the period 1983–99, 47 cases were reported, accounting for 0.34% of all amiodarone-related adverse effects. 12 , 17 However, this risk may be potentiated in a setting such as TH, where the QTc may already be prolonged due to the cooling itself. 5 , 15 , 18 , 19 Therefore, amiodarone must be used judiciously in the case of patients following cardiac arrest due to VT or VF who are undergoing TH, since it could potentially worsen the situation. Research is needed into the use of amiodarone during TH. In our case series, Patient 3 was commenced on an intravenous amiodarone infusion within 1 h of arrival at hospital (total 1.2 g over 24 h). This may have contributed to the QTc prolongation seen during hypothermia.
Despite TH being recommended in the ILCOR Guidelines 1 and ERC Guidelines 2 and becoming commonplace, it is concerning that there are no guidelines on ECG monitoring. When considered in light of the potentially arrhythmogenic effects of hypothermia, in particular the risk of further VT or VF secondary to QTc prolongation, and the reduced efficacy of antiarrythmic measures seen in studies during hypothermia, guidelines on ECG monitoring (including QTc) during TH are required.
Further research is required to determine optimal cardiac rhythm monitoring and the use of antiarrythmic medications (in particular, amiodarone) during TH following VT and VF. Temporary pacing can be performed in the treatment of resistant VT. 20 None of the patients in our case series underwent this treatment. However, had ventricular arrhythmias occurred in our patients during QTc prolongation, after removal of potential causes (i.e. hypothermia, amiodarone, and electrolyte imbalances), temporary pacing would have been considered. Research is needed on the effectiveness of temporary pacing for VT occurring in the setting of a hypothermic myocardium. Investigating the incidence of TDP and recurrence of VF during TH, and its relationship with temperature and QTc are required.
Conflict of interest: none declared.
