ELECTROCARDIOGRAPHIC CHANGES WITH ZIPRASIDONE

ELECTROCARDIOGRAPHIC CHANGES WITH ZIPRASIDONE

LETTERS TO THE EDITOR ELECTROCARDIOGRAPHIC CHANGES WITH ZIPRASIDONE To the Editor: Blair et al. (2005) recently evaluated the electrocardiographic s...

67KB Sizes 0 Downloads 21 Views

LETTERS TO THE EDITOR

ELECTROCARDIOGRAPHIC CHANGES WITH ZIPRASIDONE

To the Editor: Blair et al. (2005) recently evaluated the electrocardiographic safety profile of ziprasidone in a nonrandomized, single-arm trial involving 20 pediatric outpatients with Tourette_s syndrome, obsessive-compulsive disorder, or pervasive developmental disorder. Based on random and uncontrolled electrocardiographic (ECG) measurements, the authors reported the mean QTc interval increase from a single baseline measurement to the peak of multiple (2Y11) random ECG measurements as 28 ms (SD = 26), with a maximal increase of 114 ms, and with 8 subjects experiencing prolongation of Q440 ms (Bazett corrected). Because QT interval effects were not determined in a manner that is consistent with current standards for the design, analysis, or interpretation of studies of the QT interval, we believe the reported results are questionable and potentially misleading. First, use of the mean peak increase from a single random, uncontrolled baseline ECG value is not regarded as an acceptable method for the assessment of QTc change. The International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) E14 guidelines on the clinical evaluation of QT/QTc interval effects acknowledge this: BThe process for choosing the highest of multiple observed values will also almost invariably cause an apparent change from any single baseline value, a phenomenon found in both drug and placebo-treated groups[ (International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use, 2005). The mean change in QTc was reported based only on these maximum values. To evaluate the robustness of these results, sensitivity analyses should have been provided, including the mean baseline to endpoint change in QTc. Second, the study only reports Bazett-corrected QTc results, despite the well-known limitations of Bazett_s method in the assessment of drug-induced QTc change

636

(Fenichel et al., 2004; Malik, 2001). ICH guidance states that BBazett_s correction overcorrects at elevated heart rates and under corrects at heart rates below 60 bpm and hence is not an ideal correction. Fridericia_s correction (QTcF) is more accurate than Bazett_s correction in subjects with such altered heart rates.[ Heart rates in this study were, in fact, generally elevated above 60 bpm, with mean baseline and on-treatment heart rates of 78 bpm (range 50Y117) and 92 bpm (range 65Y149), respectively. This large 14 bpm (SD = 11) change in mean heart rate during the study is likely to have resulted in an artifactual increase in the Bazettcorrected QT interval. The Bazett-corrected QTc interval is negatively proportional to heart rate, with a slope estimated as 0.1 to 0.2 (Malik, 2003). Note that an ideally corrected QT interval would be completely independent of heart rate (or R-R interval). Given this estimated error inherent in Bazett_s correction, the heart rate elevation (or resultant decrease in R-R interval) observed in this study leads to a 12- to 23-ms increase in QTc, which is purely related to the error of Bazett_s. This magnitude of change is almost exactly what is reported as the ziprasidone-related QTc increase, making it likely that the authors reported mostly, if not exclusively, a methodological artifact of Bazett_s correction. This is a primary reason why ICH E14 suggests examining other correction factors, such as Fridericia, particularly in the absence of individual or population-based QT interval corrections (which are generally optimal but harder to obtain). Third, the lack of a placebo (or other) control and the open-label design of this study are likely to have induced both selection and estimation biases in assessing the potential ziprasidone effect on QTc change, particularly given the unbalanced gender ratio and the wide range of ages and diagnoses included in the study population. The use of random ECG measurements provided no control for potentially important confounding factors such as the timing of these measurements in relationship to activity level, postural changes, circadian patterns, and drug and food ingestion. These issues further compromise the interpretation of the reported data.

J. AM . ACAD. CHILD ADOLESC. PSYCHIATRY, 45:6, JUNE 2006

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

LETTERS TO THE EDITOR

Fourth, the authors report that 8 subjects experienced QTc intervals of Q440 ms and 3 of those had values of Q450 ms. Apart from the fact that these values most likely largely reflect errors introduced by Bazett_s correction, ICH E14 guidance recommends that categorical analysis be based on QTc threshold values of 9450, 9480, and 9500 ms. Using these thresholds and noting that both ICH and the more recent pediatric QTc guidance by Labellarte et al. (2003) refer to 450 ms as the upper limit of normal (not 440 ms as proposed by the author), it is apparent from examination of Figure 1 that only 2 of 20 subjects experienced QTc 9450 ms and 0 experienced 9480 ms (maximum of 470 ms). The sensitivity of these categorical incidence rates to small variations in the choice of threshold values suggests that the reported QTc results are not robust and are difficult to interpret. The study permitted baseline QTc values of G450 ms and observed actual values ranging up to 449 ms at baseline. Hence, several subjects at baseline were already close to or above the 440-ms threshold specified to be of concern. This raises concern that at least some of the eight reported QTc values of Q440 ms may have been caused by a high baseline QTc that persisted during the study. Regarding subjects with large QTc change, ICH E14 guidance notes that interpretation of QT interval change is complicated by the potential for variability not related to drug therapy, specifically, the phenomenon of regression to the mean. This is illustrated by the unusually large QTc change in one subject of 114 ms. Given the reported range of QTc values observed during the study (ranging up to 470 ms), this subject_s baseline QTc was at most 356 ms (a relatively low value), suggesting that the observed change at least in part resulted from regression to the mean (it would be helpful if the authors could provide the actual observed baseline values for this subject, as well as the eight subjects with QTc values Q450 ms). Fifth, in contrast to the results reported by Blair et al. (2005), a recent, fixed-dose study (interim analysis presented by DelBello et al., 2005) on ziprasidone treatment in 63 children and adolescents (ages 10Y17 years) with bipolar mania or schizophrenia suggests that ziprasidone is safe and well tolerated in this population. This study not only obtained ECGs under standardized conditions for fasting, postural changes, activity, and dose relationship, it also assessed the central tendency of changes occurring around Cmax at fixed doses. In this study, no subject experienced a baseline QTcF increase

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 45:6, JUNE 2006

exceeding 60 ms, and, in per-protocol completers, the mean QTcF change from baseline around Cmax at week 3 was 1.3 ms for the lower dose group (80 mg/day) and 11.2 ms for the higher dose group (160 mg/day). These results are consistent with QTc changes demonstrated in adult populations treated with ziprasidone (using similar, rigorously controlled methods) and differ substantially from the Blair et al. study. They provide added support for our view that the Blair et al. (2005) study, which used low doses of ziprasidone (target maximum 40 mg/day), is likely to have overstated the effect of ziprasidone on the QT interval. Given the above-noted methodological and data analysis limitations, we believe this study does not provide an accurate assessment of the effect of ziprasidone on the QT interval and is potentially misleading. Antony Loebel, M.D. Jeffrey Miceli, Ph.D. Phillip Chappell, M.D. Pfizer, Inc. New York Cynthia Siu, Ph.D. Data Power (DP), Inc. Ringoes, NJ Disclosure: Drs. Loebel, Miceli, and Chappell are employees of Pfizer, Inc. Dr. Siu is with Data Power (DP), Inc. and is a paid consultant for Pfizer. This research was supported by Pfizer. Blair J, Scahill L, State M, Martin A (2005), Electrocardiographic changes in children and adolescents treated with ziprasidone: a prospective study. J Am Acad Child Adolesc Psychiatry 44:73 Y 79 DelBello M, Ice K, Fisher DO, Versavel M, Miceli JJ (2005), Ziprasidone in the treatment of children and adolescents with bipolar mania or schizophrenia: an open-label, dose-ranging safety and tolerability study. Poster presented at New Clinical Drug Evaluation Unit, 45th Annual Meeting, Boca Raton, FL, June 6 Y 9 Fenichel RR, Malik M, Antzelevitch C et al. (2004), Drug-induced torsades de pointes and implications for drug development. J Cardiovasc Electrophysiol 15:475 Y 495 International Conference on Harmonization (ICH) E14 (2005). The clinical evaluation of QT/QTc interval prolongation and proarrhythmic potential for non-antiarrhythmic drugs. http://www.fda.gov/cder/guidance/guidance.htm. Accessed August 10, 2004 Labellarte MJ, Crosson JE, Riddle MA (2003), The relevance of prolonged QTc measurement to pediatric psychopharmacology. J Am Acad Child Adolesc Psychiatry 42:642 Y 650 Malik M (2001), Problems of heart rate correction in assessment of druginduced QT interval prolongation. J Cardiovasc Electrophysiol 12:411 Y 420 Malik M (2003), Drug-induced QT interval prolongationVwhy is it of concern? Presented at the FDA Anesthetic and Life Support Drugs Advisory Committee Meeting; Gaithersburg, MD, November 18;http:// www.fda.gov/ohrms/dockets/ac/03/slides/4000S1_05_Malik_files/ frame.htm. Accessed October 5, 2005

637

Copyright @ 2006 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.