Abstract
Dear Editor:
In the January 2010 issue, Reddy et al. 1 describe the effect of low-dose oral methadone on the rate corrected (QTc) interval in a palliative care population. The safety of low-dose methadone is an important question to evaluate in this population because this medication can be an excellent palliative tool but can also prolong the QTc interval, leading to torsades de pointes, ventricular fibrillation, and sudden death. As a result, one must weigh the palliative benefits of methadone with the risk of arrhythmia and its sequelae. The authors measured QTc at baseline before initiating low-dose methadone and then subsequently at 2, 4, and 8 weeks. They did not demonstrate any significant QTc prolongation, and in fact measured a mean reduction in QTc of 50 milliseconds at 4 and 8 weeks. While on the surface these results would indicate that methadone does not prolong QTc at low doses, several study design features require further scrutiny.
The authors report QTc intervals, as measured and calculated by the Bazett formula, in population aggregate as median and mean values. The fact that these data points have been aggregated for analysis may be a source of confounding given the marked decrement in evaluable participants at each subsequent time point. Although each measurement was taken in triplicate 5 minutes apart, no variation greater than 5 milliseconds was found. Harris and Steare. 2 evaluated spontaneous variation in QTc following administration of placebo and found a probability of observing a significant increase in QTc to be 3.0% for a single post dose measurement and 21.8% for eight post dose measurements, respectively. Lewis et al. 3 measured QTc variability at rest and with exercise and also demonstrated significant variability. It would certainly be expected then to find some variation in QTc over multiple serial measurements; to find no variation greater than 5 milliseconds seems implausible and calls into question the technique used to measure QTc. Additional study limitations include the lack of a control group and apparent lack of blinding. Therefore, measurement bias must be considered as a potential contributor to the lack of expected QTc-interval variability.
The authors demonstrate a mean 50-millisecond reduction in the QTc interval with low-dose methadone. Under other circumstances, this could be within the expected variability of QTc measures assessed serially and without controlling for concurrent factors known to influence cardiac repolarization. However, after extensive review of the literature, we were unable to identify any medication that shortens QTc by 50 milliseconds without also altering electrolytes or stimulating beta-adrenergic receptors. Indeed, a QTc-shortening effect of this magnitude, if real, would in itself be a cause of concern for pro-arrhythmia. 4 Methadone has never been previously reported to decrease the QTc, and this effect is inconsistent with the documented, potent blockade of the human Ether-à-go-go Related Gene (hERG)-encoded protein, the most common cause of clinically significant QTc prolongation. 5
The notion that low-dose methadone does not substantially prolong the QTc interval is not supported by the current literature. In a study done by Wedam et al., 6 methadone doses as low as 60 mg were shown to significantly prolong QTc. Though the majority of patients in the study by Reddy and colleagues were on lower doses, there were patients on doses as high as 90 mg/day. Unfortunately, the data were reported in aggregate and the breakout data were not available for analysis. As such, it remains unclear exactly what the QTc was in these patients that received daily doses as high as 90 mg.
The study by Reddy and colleagues seeks to answer a very important question regarding the cardiac safety of methadone at low doses in a palliative care population. While the study did not find evidence for QTc prolongation, the methodology utilized and the study design cannot adequately answer the question posed and necessitates further investigation before the presumption can be made that methadone in low doses does not prolong QTc in the palliative care arena. This is essential, since the study contradicts previous data among cancer patients receiving intravenous methadone that demonstrates significant (41.7 ± 7.8 milliseconds) QTc prolongation. 7
We believe the decision to perform QTc interval screening in the palliative care arena must be viewed in the broader context of risks and benefits. Indeed, our recent consensus guideline on QTc interval screening in methadone treatment notes that screening may not be applicable to patients with terminal cancer pain regardless of the propensity for methadone to prolong the QTc interval. 8 Interestingly, in our review of cohort studies in addiction, we observed an identical, 2% incidence of QTc intervals that exceeded 500 milliseconds compared with the study by Reddy and colleagues. We therefore disagree with the authors' contention that because our guideline was formulated based predominantly on addiction populations it may not be applicable to palliative care. Nonetheless, their position that routine electrocardiograms may not be warranted among patients who have transitioned to the palliative mode of care 1 is in our view quite defensible. However, from a cardiac safety perspective this conclusion is not directly supported by the findings from their study.