| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Departments of *Pediatric Anesthesia and Pediatric Cardiology, British Columbia Children's Hospital; Department of Anesthesiology, Pharmacology and Therapeutics, and Department of Pediatrics, University of British Columbia, Vancouver, British Columbia, Canada; ¶Jackson-Rees Department of Anesthesia, Royal Liverpool Children's Hospital and #Department of Anesthesia, Liverpool University, Liverpool, United Kingdom.
Address correspondence and reprint requests to Simon D. Whyte, MBBS, FRCA, Department of Pediatric Anesthesia, Room 1L7, British Columbia Children's Hospital, 4480 Oak St., Vancouver, BC, V6H 3V4, Canada. Address e-mail to swhyte{at}cw.bc.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: We investigated the effects of sevoflurane on corrected QT (QTc) and Tp-e intervals in 54 unpremedicated ASA I-II children, aged 3–10 yr, who were randomized to receive sevoflurane 1, 1.25, or 1.5 MAC, age-adjusted. Twelve-lead electrocardiograms were recorded before and after sevoflurane exposure. QTc and Tp-e were compared within and among groups using 2-way analysis of variance. Change in Tp-e after sevoflurane exposure was the primary outcome measure.
RESULTS: Sevoflurane significantly prolonged preoperative QTc at all doses (P < 0.005), with no dose-response relationship, but had no effect on preoperative Tp-e.
CONCLUSION: Sevoflurane markedly prolongs the QTc in healthy children, but does not increase dispersion of repolarization as measured by the Tp-e interval, indicating low or no torsadogenicity, and making it unlikely to increase predisposition to TdP.
|
Introduction |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
RR. Using this formula, the accepted upper limit of normal for the duration of repolarization is 440 ms (2). Prolongation of ventricular repolarization beyond 440 ms is associated with an increased risk of torsades des pointes (TdP). However, there is convincing evidence that this risk is not solely related to the presence of QT interval prolongation (3–5). A more reliable indicator of the true risk of TdP would be extremely useful. The interval from the peak to the end of the T-wave (Tp-e) is a candidate, as it is a measure of transmural dispersion of repolarization (TDR). The electrophysiological substrate for TdP is increased TDR (6–9). TdP only occurs when TDR is increased; QT interval prolongation per se is neither necessary nor sufficient to predispose to TdP (7,9). We have previously summarized the electrophysiology of TDR in detail (10). In brief, the cellular composition of the myocardium is not homogenous, and its component cells repolarize at different rates. Repolarization, therefore, occurs asynchronously across the myocardial wall, producing a physiological TDR. The intrinsic differential time course of repolarization across the myocardial wall is responsible for the morphology of the ECG T-wave (6,11). Epicardial cells repolarize first, and the peak of the T-wave corresponds with the completion of epicardial repolarization. Midmyocardial (M) cells, which repolarize last, determine the total duration of the action potential; the end of the T-wave corresponds with the full recovery of these cells (6). It follows that the Tp-e interval may be used as a measure of TDR. The M cell is characterized by the capacity for its action potential to lengthen disproportionately, compared with those in other areas of the myocardial wall, in response to various stimuli. In this milieu, early after-depolarizations (the R-on-T phenomenon) can initiate re-entrant circuits between areas of myocardium in variable states of refractoriness, i.e., exaggeration of TDR is arrhythmogenic (12).
QT interval prolongation is frequently drug-induced, resulting in an acquired long QT syndrome (LQTS) that is almost exclusively caused by drugs that block iKr channels, which are partly responsible for conducting the potassium efflux that effects repolarization. However, as previously mentioned, not all QTc-prolonging drugs are torsadogenic. It is logical to hypothesize that drugs actually associated with TdP are capable of increasing TDR through a preferential effect on M cell repolarization dynamics. As TDR can be measured as Tp-e, the latter may provide a readily available, noninvasive assay of drug torsadogenicity.
Volatile anesthetics are iKr channel blockers; the QTc is prolonged in healthy patients undergoing anesthesia with halothane, enflurane, isoflurane, or sevoflurane (10,13–20). Sevoflurane is widely used in pediatric anesthesia and most patients with symptomatic QT prolongation are children and young adults with an inherited mutation in one of the genes encoding the constituent proteins of channels involved in conducting repolarizing currents. There is little evidence to guide the rational selection of anesthetic drugs in these patients, where the aim is clearly to avoid anything that may increase the already elevated risk of TdP. However, only one study has investigated the effect of these anesthetic drugs on TDR, which is more relevant to the attendant risk of TdP (10). In that preliminary study, sevoflurane was investigated at only one dose, which was not adjusted for age, across an age range of 1–16 yr. Data from studies published since have better defined the change in Tp-e that would be clinically relevant. The aims of this study were to examine in greater depth the effects of sevoflurane on repolarization dynamics in a pediatric population by adjusting MAC values for age, by investigating the concentration-response relationship between sevoflurane and indices of myocardial repolarization, and by seeking any true differences in Tp-e with maximum power.
|
METHODS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
where k is the MAC multiple required and MAC40 is the ETsevo value in oxygen at 40 yr of age that is equivalent to 1 MAC (=1.8).
On arrival in the anesthetic room, and before induction of anesthesia, we placed ECG electrodes at standardized locations for acquisition of a preoperative 12-lead ECG. An intraoperative ECG was taken 15 min after induction of anesthesia, using the same electrode positions. The subject's involvement in the study was then complete, and conduct of anesthesia continued at the discretion of the supervising anesthesiologist. All ECGs were recorded in duplicate on a Hewlett Packard Pagewriter Xli Model 1700A (Philips Medical Systems, Bothell, WA), at a paper speed of 50 mm/s. No identifying data or automated analysis were printed on the recorded traces. Each ECG was given a random number three-figure code to allow identification of paired pre- and intraoperative traces after analysis.
Anesthesia was induced and maintained for 15 min with the drug allocated by randomization. Each child received sevoflurane 8% in oxygen for induction. After induction, we diluted the inspired oxygen concentration to 40% with air and titrated sevoflurane to the end-tidal concentration determined by the age-adjusted MAC value to which the child had been randomized. We maintained the airway by facemask or laryngeal mask airway. In an attempt to minimize sympathetic stimulation, laryngoscopy was not permitted during the study period. No other drugs were administered and no local anesthetic blocks were conducted during the study period. Throughout the study period, all children breathed spontaneously and received routine monitoring, including capnography. End-tidal carbon dioxide values ranged between 35–43 mm Hg.
Two authors (SS and PDB) independently analyzed all the ECG traces in accordance with predetermined criteria. Both were blinded to the concentration of anesthetic used and to the status of the ECG recording (pre- or intraoperative). Neither was involved in recruitment or randomization of patients, or in conduct of anesthesia or acquisition of ECG recordings, all of which were performed by the other two authors (SDW and JL).
We measured the QT, RR, and Tp-e intervals in leads II and V5. We measured the QT interval from the start of the QRS complex to the end of the T-wave and the Tp-e interval from the peak of the T-wave to the end of the T-wave, defined as the point of return to the T-P baseline (22). If U-waves were present, we defined the end of the T-wave as the nadir of the curve between the T and U-waves. We calculated QT and Tp-e intervals for all complete P-QRS-T cycles in each lead and averaged them to give a mean QT interval and Tp-e interval for that lead. QT intervals were corrected according to the formula of Bazett (1), where QTc = QT/RR. We used Bland-Altman plots to compare the ECG data from the two independent reviewers. Where we found an inter-observer difference of >10 ms in an RR interval or of >20 ms in a QT or Tp-e interval, we reanalyzed the recordings, still coded, and reached a consensus if possible. Thus, for leads II and V5 in each trace, we eventually obtained two values for the mean RR interval, the mean QTc interval and the mean Tp-e interval, one from each independent reviewer. Each pair of values was then averaged to give an overall mean value for use in further statistical analysis.
We performed within-group and between-group comparisons of pre- and intraoperative ECG indices using two-way analysis of variance with a priori Bonferroni correction for multiple pairwise testing. Statistical analysis was conducted with Analyse-It® software (Analyze-It Software, Leeds, England). Using published data (10,23,24), interpretation of an effect in either direction, and the criterion for significance (
) set at <0.003 (<0.05 before a priori Bonferroni correction), we had calculated that a sample size of 14 per group would detect a difference of 25 ms in Tp-e between the intraoperative means of the three groups with a power of >99%.
|
RESULTS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
We found very close agreement between the reviewers in measured RR intervals in both leads (lead II mean bias 7 ms; mean error –1.9% to +1.7%; lead V5 mean bias 3 ms; mean error –2.2% to +1.7%). The interobserver bias (95% limits of agreement) for the measurement of QTc was 5 ms (–11 to +22 ms) both in lead II and V5. For the measurement of Tp-e, the values were –4.5 ms (–18.7 to +9.7 ms) in lead II and –3.8 ms (–16.7 to +9.1 ms) in lead V5. These figures mean that, between the reviewers, the difference in measured QTc and Tp-e intervals averaged less than one small ECG square either way.
Table 2 shows the results of the within-group analyses of pre- and intraoperative QTc measurements in groups 1, 1.25, and 1.5 MAC. The QTc interval was normal preoperatively in all three groups, with no significant differences among groups (ANOVA, P = 0.07 and 0.22 for leads II and V5, respectively). Sevoflurane markedly prolonged the QTc in both leads II and V5, in all three groups, by 28–55 ms. After 15 min of sevoflurane anesthesia, the absolute QTc exceeded the upper limit of normal in both leads in all three groups. These differences were statistically significant in all groups. Between-group analysis of the QTc values after sevoflurane confirmed the absence of a dose-response relationship (ANOVA, P = 0.69 and 0.95 for leads II and V5, respectively).
|
Table 3 shows the results of the within-group analysis of pre- and intraoperative Tp-e measurements in Groups 1, 1.25, and 1.5 MAC. The Tp-e interval was normal preoperatively in all three groups, with no significant differences among the groups (ANOVA, P = 0.68 and 0.85 for leads II and V5, respectively). There were no significant increases in Tp-e in either lead II or V5 after sevoflurane exposure and no dose-response relationship between sevoflurane concentration and Tp-e (ANOVA, P = 0.66 and 0.5 for leads II and V5, respectively).
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our results indicate that clinically relevant doses of sevoflurane do not exaggerate Tp-e, despite prolonging the QTc. This leads us to suggest that sevoflurane has extremely low torsadogenicity or none at all. This assertion is based on the assumptions that the risk of TdP is directly attributable to increases in TDR, and that Tp-e is a valid and reliable surrogate marker of TDR.
QTc prolongation is a poor predictor of the risk of TdP actually occurring: up to 40% of patients with congenital QT prolongation are asymptomatic at the time of diagnosis (3); not all drugs that are capable of prolonging the QT interval are torsadogenic (4); 6% of patients with symptomatic LQTS, i.e., having episodes of TdP, have a baseline QTc interval that is not absolutely prolonged (5). In arterially perfused canine heart wedge models, it has been demonstrated that QTc prolongation per se does not predispose to TdP, but exaggeration of physiological TDR does (7,9,11). Preclinical drug testing for torsadogenicity includes assessment of the effect of drugs on TDR, although considerable debate continues about the best indices for measuring it (25).
The observation, in the arterially perfused wedge model, that conclusion of epicardial repolarization coincides with the peak of the T-wave and completion of M cell repolarization coincides with the end of the T-wave, led to the proposition of the Tp-e interval as a surface ECG marker of TDR (6). Evidence for the validity of this parameter in clinical practice continues to accumulate. In patients with acquired LQTS, Tp-e was found by logistic regression analysis to be the best predictor of which patients developed TdP (26). Tp-e is prolonged in symptomatic congenital LQTS patients (24). In assessing the risk of arrhythmias in LQT1 and LQT2, Tp-e appears to be a useful index of TDR (27,28). Tp-e is increased in premature neonates receiving the torsadogenic drug cisapride (23).
Our finding that sevoflurane does not prolong Tp-e is consistent with the apparent absence of TdP associated with extensive sevoflurane usage around the world, despite the drug's ability to dramatically prolong QTc. Although a recent case report attributed episodes of intraoperative TdP to the use of sevoflurane (29), several more probable causative factors were present (Whyte SD, Sanatani S, Booker PD. Torsades de points with sevoflurane. Pediatr Anaesth 2006;16:1199–1200). We are unaware of any reports that convincingly attribute an intraoperative or postoperative complication to cardiac repolarization abnormalities after sevoflurane anesthesia in pediatric patients. Interestingly, thiopental, which also prolongs QTc but appears not to be torsadogenic, reduced TDR in an in vivo animal model (30). Conversely, although there are contradictory reports on the ability of halothane to prolong QTc (13–15,31,32), it exaggerates TDR in dogs (33), and was the anesthetic in use in three case reports of perioperative TdP in patients with previously undiagnosed LQTS (34–36). All these observations are compatible with the evolving hypothesis that it is possible to have a prolonged QT interval without exaggerated TDR; the risk of TdP in these instances is very low, whereas increased TDR increases the risk of arrhythmias, even if the absolute QT interval is normal.
Our second finding, that sevoflurane significantly prolongs the QTc, is in keeping with previous studies in adults and children, which have consistently shown a marked propensity for sevoflurane to prolong the QTc (16–20). In addition, we have shown that there is no concentration-response phenomenon in the 1–1.5 MAC range. QT prolongation by sevoflurane is a predictable side effect. In guinea pig cardiac myocytes, sevoflurane inhibits iKr channels, which contribute significantly to the repolarizing current in phase 3 of the cardiac cycle action potential in many species (37). Mutations in iKr channel proteins cause LQTS 2 and 6 in humans (38,39) and most drug-induced LQTS is caused by iKr channel blockade (40). However, sevoflurane had no significant effect on the Tp-e interval in this study, suggesting that it does not increase the risk of TdP, despite its propensity to prolong the QTc. One explanation to account for this discrepancy is that sevoflurane has an equal effect on repolarization in epicardial, endocardial, and M cells, such that there is prolongation in the overall duration of repolarization (reflected in QTc prolongation), but no increase in TDR.
QTc is conventionally measured in lead II. The best lead for measuring Tp-e has not been defined, but it has been suggested that the left precordial lead values may best reflect true TDR (30). We have previously examined interobserver variability in all 12 leads (10) and found it to be least in leads I, II, and V5. Hence, we thought that leads II and V5 were the most appropriate from which to report our results. Xia et al. (41) have reported that maximal Tp-e correlates with TDR in pigs, while Tp-e in leads II and V5 may under-estimate TDR. Conversely, a computer simulation study suggests that Tp-e may over-estimate TDR (42). Further studies of the correlation between TDR and surface ECG T-wave morphology in humans are needed to refine the relationship between TDR and Tp-e.
We used published criteria to define the end of the T-wave (22). U-waves were commonly present in precordial leads, but did not complicate definition of the T-wave end. There are objective criteria for defining the peak of T-waves with complex morphology, but not for normal, monophasic T-waves, whose peaks are typically discrete and readily identified. Nevertheless, the subjective defining of T-wave peak introduces a potential source of inaccuracy in measuring Tp-e.
The normal range (and hence the upper limit of normal) for Tp-e has not been defined. We found a mean Tp-e of 76.0 (sd 11.6) ms in our sample of 53 preoperative ECG traces from healthy children. This is similar to the value of 72.2 (sd 10.9) ms in our pilot study, on which we based our power calculations. Preliminary studies in neonates on cisapride, and in treated LQTS patients, suggest that an increase of 20–35 ms in Tp-e is clinically significant. Our study gave us 99% power to detect a 25 ms difference.
The major limitation in this study is that we did not study children with overt long QT syndrome, which inevitably makes speculative our conclusions regarding the effects of sevoflurane on dispersion of repolarization in such patients. For this reason, we must emphasize caution in the extrapolation of our findings to these patients. Nevertheless, we defend the clinical relevance of our findings, as there is a hidden population of phenotypically normal patients, with subclinical genetic mutations in iKr, iKs, and iNa cardiac ion channels that result in reduced repolarization reserve (43,44); such patients are at risk of unmasking of their genotype when exposed to additional impairment of their repolarization mechanisms, such as pharmacological iKr blockade (45,46).
In conclusion, sevoflurane 1–1.5 MAC non-dose-dependently prolongs the QTc interval in healthy children, but does not appear to exaggerate physiological TDR, as measured by the Tp-e interval. These results suggest that sevoflurane has low or no intrinsic torsadogenicity and does not increase the risk of TdP.
|
Footnotes |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. V. Hume-Smith, S. Sanatani, J. Lim, A. Chau, and S. D. Whyte The Effect of Propofol Concentration on Dispersion of Myocardial Repolarization in Children Anesth. Analg., September 1, 2008; 107(3): 806 - 810. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Hunter, P. Sharma, and S. Rathi Long QT syndrome CEACCP, April 1, 2008; 8(2): 67 - 70. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
