C

Sac

CMa

b

a

ARRA

KHCAT

1

pcL

eCcPsfi

i

0d

Resuscitation 82 (2011) 859–862

Contents lists available at ScienceDirect

Resuscitation

journa l homepage: www.e lsev ier .com/ locate / resusc i ta t ion

linical paper

evere QTc prolongation under mild hypothermia treatment and incidence ofrrhythmias after cardiac arrest—A prospective study in 34 survivors withontinuous Holter ECG�

hristian Storma,∗ , Dietrich Haspera , Jens Neea , Achim Joerresa , Joerg C. Schefolda , Jan Kaufmannb ,attias Roserb

Department of Nephrology and Medical Intensive Care Medicine, Campus Virchow-Klinikum, Charité Universitätsmedizin Berlin, Augustenburger Platz 1, 13353 Berlin, GermanyDeutsches Herzzentrum Berlin, Department of Cardiology, Germany

r t i c l e i n f o

rticle history:eceived 20 December 2010eceived in revised form 12 February 2011ccepted 27 February 2011

eywords:ypothermiaardiac arrestrrhythmiaorsade de pointes

a b s t r a c t

Background: Mild hypothermia treatment (32–34 ◦C) in survivors after cardiac arrest (CA) is clearly rec-ommended by the current guidelines. The effects of cooling procedure towards QT interval have not beenevaluated so far outside of case series. In a prospective study 34 consecutive survivors after cardiac arrestwere continuously monitored with Holter ECG over the first 48 h.Patients and methods: A total of 34 patients were analysed and received mild therapeutic hypothermiatreatment (MTH) according to the current guidelines and irrespective of the initial rhythm. At admissionto hospital and in the field in case of OHCA, a 12-lead ECG was performed in all patients.Results: During cooling the incidence of ventricular tachycardia was low (8.8%) and in none of the patientsTorsade de pointes occurred. The QTc interval was within normal range at first patient contact with EMS inthe field (440.00 ms; IQR 424.25–476.75; n = 17) but during hypothermia treatment the QTc interval was

◦

significantly prolonged at 33 C after 24 h of cooling (564.47 ms; IQR 512.41–590.00; p = 0.0001; n = 34)and decreased after end of hypothermia to baseline levels (476.74 ms; 448.71–494.97; p = 0.15).Conclusion: The QTc interval was found to be significantly prolonged during MTH treatment, and somesevere prolongations >670 ms were observed, without a higher incidence of life-threatening arrhythmias,especially no Torsade des pointes were detected. However, routine and frequent ECG recording withrespect to the QTc interval should become part of any hypothermia standard operation protocol andshould be recommended by official guidelines.

. Introduction

Mild therapeutic hypothermia (MTH) has become a routinerocedure in survivors after cardiac arrest to improve neurologi-al outcome following the current guidelines by the Internationaliaison Committee on Resuscitation (ILCOR) and the European

Abbreviations: CA, cardiac arrest; CPC, Cerebral Performance Category; ECG,lectrocardiogram; EMS, emergency medical system; ERC, European Resuscitationouncil; ICU, intensive care unit; ILCOR, International Liaison Committee on Resus-itation; MTH, mild therapeutic hypothermia; OHCA, out-of-hospital cardiac arrest;EA, pulsless electrical activity; QTc interval, corrected QT interval; ROSC, return ofpontaneous circulation; SDNN, standard deviation of NN-interval; VF, ventricularbrillation; VT, ventricular tachycardia.� A Spanish translated version of the abstract of this article appears as Appendixn the final online version at doi:10.1016/j.resuscitation.2011.02.043.∗ Corresponding author. Tel.: +49 30 450 553232.

E-mail address: Christian.storm@charite.de (C. Storm).

300-9572/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.resuscitation.2011.02.043

© 2011 Elsevier Ireland Ltd. All rights reserved.

Resuscitation Council (ERC).1 A possible transient bradycardia iswell known under conditions of hypothermia but only a few studiesevaluated the influence of MTH towards heart rate variability andthe risk of life-threatening arrhythmias.2,3 A currently publishedcase series described in four patients undergoing MTH a prolonga-tion of the corrected QT interval (QTc).4 QTc prolongation itself canpromote the development of malignant arrhythmias, for exampleventricular tachycardia (VT).5

This prospective study was conducted to evaluate the frequencyof malignant arrhythmias and to analyse the possible effect ofhypothermia on the QTc interval using a continuous Holter ECGduring MTH treatment.

2. Methods

The protocol of this prospective single center study wasapproved by the local ethics committee on human research andis conducted in accordance with the guidelines of the Declaration

8 citation 82 (2011) 859–862

ocCop3hgpcpdrwbaUIb(airadptss

ttchtHdAwraswtiQ

ifnu

1a(ufiFaodmetdi

Table 1Baseline characteristics given as number/total number and % or median andinterquartile range (IQR; 25–75).

Variable Patients (n = 34)

Baseline characteristicsAge (years) 61.5 (52.0–77.5)Female sex-no./total no.(%) 12 (35.3)APACHE score 31.5 (24.5–37.0)Location of cardiac arrest

Out-of-hospital no./total-no. (%) 28 (82.3)In-hospital no./total-no. (%) 6 (17.6)

Cause of cardiac arrestAMI-no./total no. (%) 10 (29.4)Primary arrhythmia-no./total-no. (%) 14 (41.2)Respiratory no./total-no. (%) 8 (23.5)Other no./total no.(%) 2 (5.9)

Initial rhythmAsystole no./total-no (%) 12 (35.3)PEA no./total-no (%) 7 (20.6)VF no./total-no (%) 15 (44.1)PCI-no/total no (%) 17 (50)Time to ROSC (min) 12 (8.5–17)Length of ICU stay (days) 14.5 (5.5–27.5)Time on ventilator (h) 236 (116.5–418.0)

TreatmentAmiodarone (300 mg bolus i.v.)

Pre-admission-no./total no. (%) 7 (20.6)ICU first 48 h-no./total no. (%) 3 (8.8)

AdrenalinPre-admission-no./total no. (%) 26 (76.5)Cumulative dosage (mg) 2 (1.00–2.50)

NoradrenalinICU first 48 h-no./total no. (%) 31 (91.2)Cumulative dosage (mg) 6.76 (1.80–29.87)

Dopamine

60 C. Storm et al. / Resus

f Helsinki. Written informed consent for the use of routine clini-al data is part of the standard contract between patients and theharité Universitätsmedizin Berlin and was obtained from patientsr their legal representatives, if available. During the inclusioneriod between April 2009 and December 2009 a total number of4 patients were analysed. All patients received mild therapeuticypothermia treatment after resuscitation according to the currentuidelines and regardless of the initial rhythm. Resuscitation waserformed in conformance with the current guidelines of advancedardiac life support of the American Heart Association and the Euro-ean Resuscitation Council. The hypothermia treatment startedirectly after admission to hospital as well as the continuous ECGecording. Therapeutic hypothermia was initiated after admissionith an intravenous infusion of cold saline (4 ◦C, 1000–1500 ml

olus) followed by surface cooling with a commercially avail-ble non-invasive device (ArcticSun2000® Medivance, Colorado,SA). The target temperature of 33 ◦C was maintained for 24 h.

ntravenous sedation and analgesia were induced in all patientsy a combination of midazolam (0.125 mg/kg/h) and fentanyl0.002 mg/kg/h) with dose adjustment as needed. Muscle relax-tion with repetitive administration of pancuronium (0.1 mg/kg)n order to prevent shivering was induced if necessary. Post-esuscitation care was uninfluenced by the Holter ECG resultsnd the ICU physicians followed the standard operating proce-ure for cardiac arrest patients. All patients completed the coolingrocedure and survived the first 48 h. Overcooling (central bodyemperature lower than 32 ◦C) did not occur in any case, potas-ium serum levels were closely monitored and supplemented totay within normal range and to avoid hypokalemia.

In out-of-hospital cardiac arrest (OHCA), data concerning drugreatment, initial rhythm, time to return of spontaneous circula-ion (ROSC) and the first ECG after restoration of circulation toalculate pre-clinical QTc interval were collected. At admission toospital, an ECG was performed in all patients and hypothermiareatment started immediately as well as Holter ECG recording.olter recordings were performed with a portable three-channeligital recorder continuously over 48 h in all patients (MedilogR12, Schiller MediLog, Wiesbaden, Germany). The digital tapesere independently analysed by two experienced observers. Heart

ate [beats/min] is given as mean, minimum and maximum,rrhythmias were classified as the following: salve 3–5 beats, non-ustained > 5 beats ≤ 29 s, sustained VT ≥ 30 s. Heart rate variabilityas analysed by the standard deviation of individual normal-

o-normal intervals (SDNN) as time domain variable. The QTcnterval was corrected according to the Bazett formula (QTc =T/

√RR).

Clinical outcome was assessed at the time of discharge fromntensive care unit (ICU) according to the Pittsburgh Cerebral Per-ormance Category (CPC).6 CPC 1–2 were classified as a favourableeurological outcome whereas CPC 3–5 were regarded as annfavourable outcome.

The PASW SPSS software (Version 18.0) and MedCalc (Version1.2.1), Mariakerke, Belgium, were used for statistical analysis. Fornalysis of treatment-induced changes in corrected QT intervalsgiven in milliseconds), data were checked for normal distributionsing the Kolmogorov–Smirnov test. Respective data were identi-ed to be normally distributed at all points in time of assessment.ollowing testing for normal distribution, a repeated measurementnalysis of variance (ANOVA) test was performed. For comparisonf the respective points in time of assessment, Students’ indepen-ent samples t-test was used. Descriptive parameters are given as

edian and interquartile range (IQR). Univariate analysis of differ-

nces between groups was performed using the Mann–Whitney–Uest for non-parametric unpaired data and Fisher’s exact test forichotomous variables. p < 0.05 was considered statistically signif-

cant.

ICU first 48 h-no./total no. (%) 14 (41.2)Cumulative dosage (mg) 0.00 (0.00–405.60)�-Blocker 0

3. Results

3.1. Baseline characteristics

Median age was 61.5 years (52.0–77.5 IQR) with a majority ofOHCA patients (82.3%). In n = 15 the initial rhythm was ventricularfibrillation (VF), n = 12 had asystole and n = 7 suffered cardiac arrestdue to pulseless electrical activity. The median number of defibril-lations in the VF group was 2 (1.26–4.0 IQR). 20.6% of the patientsreceived initial antiarrhythmic drug treatment with amiodaroneduring resuscitation (bolus dosage of 300 mg i.v. once) and threepatients received 300 mg amiodarone bolus during hypothermiatreatment. 76.5% received adrenalin treatment during cardiopul-monary resuscitation (median dosage 2 mg; 1–2.5 IQR). Detailedcharacteristics for all patients in the study are given in Table 1.

3.2. Continuous Holter ECG

Analyses of continuous Holter ECG revealed a median heartrate of 85.50 (IQR 68.50–93.25). The monitored arrhythmiaswere in detail in median VES 335 (IQR; 51.25–1379.00), cou-plet monomorph 1 (0.00–8.50), couplet polymorph 2 (0.00–14.50),bigeminus 0 (0.00–2.00), trigeminus 0 (0.00–3.00), and salve 0(0.00–0.25). The number of patients with at least one VT (ven-tricular tachycardia) was three (8.8%) during MTH and werenon-sustained without any additional treatment. Torsade depointes were not detected. Heart rate variability, given as SDNN

in ms (milliseconds) was 128 (101.50–179.47).

The QTc interval was assessed at seven prospectively definedpoints in time (Fig. 1). Data at the respective points in time ofassessment were found to be normally distributed. In 17 patientsa 12-lead ECG was available from the EMS after OHCA (t1), in 30

C. Storm et al. / Resuscitatio

Fig. 1. Duration of QTc interval [ms] at different time points of measurement, hor-izontal line at 450 ms indicates upper level of normal range of QTc interval. Timepoints t1: out-of-hospital ECG (n = 17), t2: admission to hospital (n = 34), t3: startof MTH (n = 34) and Holter monitoring, t4: after 6 h at 33 ◦C, t5: after 12 h at 33 ◦C,t6: end of MTH after 24 h at 33 ◦C, t7: end of Holter monitoring after 48 h > 36 ◦C.Ct(p

psdit4aaw((at(mca(rot(phpahtoiea

3

(sg

hanges in QTc interval are tested between different time points: t2–t6 (p = 0.0001),1–t7 (p = 0.15), t6–t7 (p < 0.007); * indicates testing between t2 and t3, t4, t5, t6all p = 0.0001); † indicates testing between t1–t3 (p = 0.008) and t1–t4, t5, t6 (all< 0.0001).

atients a 12-lead ECG was available from time of hospital admis-ion (t2) and from all patients (n = 34) five time points at initiation,uring and after hypothermia were available due to Holter record-

ngs (t3: start of cooling, t4: after 6 h at 33 ◦C, t5: after 12 h at 33 ◦C,6: end of MTH after 24 h at 33 ◦C, t7: end of Holter monitoring after8 h > 36 ◦C). The QTc interval after 48 h (t7 in Fig. 1) was measuredfter the end of the cooling procedure and when the patients hadlready reached normal body temperature. The QTc interval wasithin normal range at first patient contact with EMS in the field

440.00 ms; IQR 424.25–476.75; n = 17) and on admission to ICU460.97 ms; IQR 440.00–474.00; n = 30). Repeated measurementnalysis of variance (ANOVA) were used as the same parame-er was assessed under different conditions in the same subjectsrepeated measurement ANOVA p = 0.007). Therapeutic hypother-

ia (applied from t3 to t6), induced a significant increase in theorrected QT interval. In detail, when tested against baseline levelst t1, significant changes were identified at t3 (p = 0.008) and t4–t6all p < 0.0001). Interestingly, after treatment (at t7), QTc intervalseached the levels of normal (p = 0.15 vs. baseline levels). More-ver, a significant difference between QTc levels after hypothermiareatment (at t7) versus during hypothermia (t3–t6) was observedall p < 0.007). Thus, hypothermia treatment induced an off-on-offhenomenon with regard to QTc intervals (Fig. 1). Furthermore, asypothermia was applied after point in time of assessment t2, thisoint in time was tested for significant differences in QTc intervalslso. We observed significant changes at all points in time (t3–t6) ofypothermia application (all p < 0.0001 compared to t2). This fur-hermore underlines a significant impact of hypothermia treatmentn QTc intervals. In one patient a maximum of 673.52 ms of QTcnterval prolongation was reached after 12 h of 33 ◦C as the high-st recorded value without any additional drug treatment causingQTc prolongation.

.3. Potentially QTc interval effecting drug treatment

32.4% of the OHCA patients received an intravenous bolus300 mg) of amiodarone during resuscitation; on ICU after admis-ion 8.8% received amiodarone bolus according to the currentuidelines of the European Resuscitation Council (ERC). 44.1%

n 82 (2011) 859–862 861

patients received antibiotic treatment in the early post resuscita-tion phase antibiotic treatment due to the directive of the physicianon duty. From the antimicrobial drugs given (cephalosporins, chi-nolone, ampicillin and carbapeneme) no typical influence towardsQT interval is known, but for makrolid antibiotics case reports witha QTc prolongation are published.

3.4. Electrolytes

Serum potassium was monitored very closely and was heldwithin the normal range of 3.5–5.5 mmol/L during cooling proce-dure.

3.5. Neurological outcome

17 patients (50%) were discharged with a favourable neurologi-cal outcome (CPC 1–2) whereas 50% had an unfavourable outcome(CPC 3–5). The overall mortality rate was 38.2%, 13 patients diedrespectively.

4. Discussion

This is the first study revealing a profound prolongation ofthe QTc interval during MTH without a severe incidence of life-threatening arrhythmias. Furthermore no Torsade de pointes weredetected during 48 h Holter ECG monitoring, although a tight cor-relation was observed of prolonged QTc interval and decreasingtemperature level in patients undergoing MTH and only a low inci-dence of VT has been recorded.

In general, if central body temperature decreases, a decreaseof heart rate, spontaneous depolarization of pace-maker cells,myocardial impulse conduction and increase of the action potentialis the normal response. The observations, based on these changesat the cell membrane level, are prolonged PR interval, expandedQRS and increased QT interval.7,8 Following these conditions, themonitored findings of prolonged QTc interval in this study couldbe expected, although the median prolongation after 24 h at 33 ◦Cwith a QTc interval of 564.47 is clearly higher than the normal rangein the literature. Surprisingly, the incidence of malignant or life-threatening arrhythmias such as Torsade de pointes or VT werelow compared to the QTc prolongation. Experimental data showeda stabilization of cell membranes during hypothermia and a higherlikelihood of successful defibrillation with a better ROSC rate ina swine model due to hypothermic conditions.9,10 This indicatesthat MTH lowers the incidence of arrhythmias rather than rais-ing it. One should keep in mind that more profound hypothermia<30◦ will increase the risk and therefore temperature should beclosely monitored in patients undergoing MTH. This positive effectof cooling towards reduction of potential arrhythmias is in additiondependent on a close monitoring of electrolytes, especially potas-sium, because a coincidence of potential harmful conditions couldreverse the positive effects. Additional drug treatment with sub-stances that are known to cause potential QTc prolongation needsto be kept in mind.11 Amiodarone is recommended in refractory VFor pulseless VT under resuscitation, acting at the potassium chan-nel, and is known to cause lengthening of the action potential.12

Following our data there is no increased risk for arrhythmias ifamiodarone is given during or before MTH treatment. Again, it isemphasized that it is important to hold potassium within normalrange and to avoid profound hypothermia. In case series a potential

influence on QTc interval for antibiotics and other drug treatment isdescribed.4,11,13 However, a QTc prolongation in elderly emergencypatients (QTc ≥450 ms) has been observed in almost 544/1558patients (35%) in a study by Seftchick et al. The most commoncomorbidities in this study were structural heart disease, renal

8 citatio

fgbcn

pepapmssdapw

5

ootifHQol

C

A

t

1

1

1

13. Khan JN, Prasad N, Glancy J. Amiodarone use in therapeutic hypothermia fol-lowing cardiac arrest due to ventricular tachycardia and ventricular fibrillation.Europace 2009;11:1566–7.

14. Seftchick MW, Adler PH, Hsieh M, et al. The prevalence and factors associated

62 C. Storm et al. / Resus

ailure, and stroke.14 Five percent of the patients with QTc prolon-ation died in the emergency department or during hospitalizationut none had QTc prolongation or Torsade de pointes listed as aause of death. Therefore a delay of repolarisation per se seems notecessarily torsadogenic.11

This study has some limitations that need to be discussed. Inost cardiac arrest patients a minor QTc prolongation is in gen-ral possible for multiple reasons. In our analysis the severe QTcrolongation is closely associated with the duration and temper-ture level of hypothermia treatment. In addition, when coolingrocedure was finished the QTc interval reached almost the nor-al range. Therefore the mild therapeutic hypothermia treatment

eems to have the major effect on severe QTc prolongation in thisetting. Due to ethical reasons and current guidelines a randomizedesign with a control group is not possible in cardiac arrest patientsnymore. In OHCA patients a 12-lead ECG was not available for allatients and therefore the pre-clinical data have to be interpretedith caution and only can indicate a trend of the QTc interval.

. Conclusion

According to our results, during mild hypothermia treatmentne should closely monitor the QTc interval and should be awaref potentially upcoming malignant arrhythmias. Furthermore, thereatment with drugs causing an additional prolongation of QTcnterval, especially anti-arrhythmic drugs and antibiotics that arerequently used in the ICU setting, should be made with caution.owever, routine and frequent ECG recording with respect to theTc interval should become part of any hypothermia standardperation protocol and should be recommended by official guide-ines.

onflict of interest

None declared.

cknowledgments

The authors would like to thank Astrid Caemmerer for assis-ance and support throughout the study. This work was supported

n 82 (2011) 859–862

by Schiller MediLog GmbH, Wiesbaden, Germany, with technicalequipment (Medilog AR12 recorder).

References

1. Nolan JP, Morley PT, Hoek TL, Hickey RW. Therapeutic hypothermia after car-diac arrest. An advisory statement by the Advancement Life support TaskForce of the International Liaison committee on Resuscitation. Resuscitation2003;57:231–5.

2. Lasky RE, Parikh NA, Williams AL, Padhye NS, Shankaran S. Changes in the PQRSTintervals and heart rate variability associated with rewarming in two newbornsundergoing hypothermia therapy. Neonatology 2009;96:93–5.

3. Tiainen M, Parikka HJ, Makijarvi MA, et al. Arrhythmias and heart rate variabil-ity during and after therapeutic hypothermia for cardiac arrest. Crit Care Med2009;37:403–9.

4. Khan JN, Prasad N, Glancy JM. QTc prolongation during therapeutic hypother-mia: are we giving it the attention it deserves? Europace 2010;12:266–70.

5. Johnson JN, Ackerman MJ. QTc: how long is too long? Br J Sports Med2009;43:657–62.

6. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet1975;1:480–4.

7. Aslam AF, Aslam AK, Vasavada BC, Khan IA. Hypothermia: evaluation, elec-trocardiographic manifestations, and management. Am J Med 2006;119:297–301.

8. Polderman KH, Herold I. Therapeutic hypothermia and controlled normother-mia in the intensive care unit: practical considerations, side effects, and coolingmethods. Crit Care Med 2009;37:1101–20.

9. Boddicker KA, Zhang Y, Zimmerman MB, Davies LR, Kerber RE. Hypothermiaimproves defibrillation success and resuscitation outcomes from ventricularfibrillation. Circulation 2005;111:3195–201.

0. Rhee BJ, Zhang Y, Boddicker KA, Davies LR, Kerber RE. Effect of hypother-mia on transthoracic defibrillation in a swine model. Resuscitation 2005;65:79–85.

1. Redfern WS, Carlsson L, Davis AS, et al. Relationships between preclinical car-diac electrophysiology, clinical QT interval prolongation and Torsade de pointesfor a broad range of drugs: evidence for a provisional safety margin in drugdevelopment. Cardiovasc Res 2003;58:32–45.

2. Haverkamp W, Breithardt G, Camm AJ, et al. The potential for QT prolongationand proarrhythmia by non-antiarrhythmic drugs: clinical and regulatory impli-cations, Report on a policy conference of the European Society of Cardiology.Eur Heart J 2000;21:1216–31.

with QTc prolongation among emergency department patients. Ann Emerg Med2009;54:763–8.