Rodriguez I, Kilborn MJ, Liu X, Pezzullo JC, Woosley RL. Drug-Induced QT Prolongation in Women During the Menstrual Cycle. JAMA. 2001;285(10):1322-1326. doi:10.1001/jama.285.10.1322
Author Affiliations: Division of Clinical Pharmacology, Departments of Medicine and Pharmacology, Georgetown University Medical Center, Washington DC.
Context Women have a higher incidence of torsades de pointes than men, but it
is not known if the risk of drug-induced torsades de pointes varies during
the menstrual cycle.
Objectives To determine if the degree of QT prolongation in response to ibutilide
varies with the menstrual cycle phase and to compare QT prolongation between
women and men.
Design and Setting Cohort study of men and women who received the same intervention conducted
between November 1998 and November 2000 at a general clinical research center
of a university hospital.
Participants A volunteer sample of 58 healthy adults (38 men and 20 women) aged 21
to 40 years.
Intervention A low dose of ibutilide (0.003 mg/kg), infused intravenously for 10
minutes. Subjects were monitored for 120 minutes. Women received the intervention
on 3 separate occasions to correspond with menstrual cycle phases, which were
verified by using hormonal assays.
Main Outcome Measure QT interval, recorded from electrocardiogram at timed intervals during
and after ibutilide infusion and standardized for variations in heart rate
Results Maximum (mean [SD]) millisecond increase in QTc after ibutilide infusion
was greater for women during menses (63 ) and the ovulatory phase (59
) compared with women during the luteal phase (53 ) and compared with
men (46 ; P = .002 vs menses and P = .007 vs ovulation). Progesterone (r = −0.40)
and progesterone-to-estradiol ratio (r = −0.41),
but not estradiol (r = 0.14) or testosterone (r = 0.09), were inversely correlated with ibutilide-induced
Conclusions Menstrual cycle and sex differences exist in QTc responses to ibutilide,
with the greatest increase in QTc corresponding to the first half of the menstrual
Women have slower cardiac repolarization than men, which manifests as
longer heart rate corrected QT intervals (QTc) on the electrocardiogram (ECG).1 This sex difference is apparent only after puberty.2 Furthermore, women are more prone than men to develop
torsades de pointes ventricular arrhythmias after administration of drugs
that prolong cardiac repolarization (eg, antiarrhythmic drugs, terfenadine,
erythromycin, etc).3- 5
These findings suggest a role for sex hormones in the response to drugs that
alter cardiac repolarization, and animal studies have demonstrated that sex
hormones can alter potassium channel expression, ion currents, cardiac repolarization,
and QT response to drugs.6- 9
During the menstrual cycle there is a dynamic change in circulating
levels of estrogen and progesterone. In the absence of a drug that alters
cardiac repolarization, QTc does not change during the menstrual cycle,10 but the possibility that the variation in the hormonal
milieu may cyclically modulate the action and/or disposition of drugs has
not been studied.
Ibutilide is an antiarrhythmic agent that is used for termination of
atrial fibrillation and flutter. It prolongs QTc in a dose-dependent manner
with a rapid onset and return to near baseline within 2 to 6 hours. The plasma
concentrations of the drug fall rapidly following intravenous infusion, and
it has no known active metabolites,11 making
ibutilide an excellent probe to study differences in drug-induced QT prolongation.
The purpose of this study was to compare QT prolongation after the administration
of a dose of ibutilide in women during 3 phases of the menstrual cycle and
to compare the degree of QT prolongation in response to ibutilide between
women and men.
We studied 58 healthy volunteers who were not taking any medications,
38 men and 20 women, between ages 21 and 40 years and within 10% of their
ideal body weight (Metropolitan Life Table). All volunteers had normal clinical and laboratory evaluations
and normal ECG with QTc no more than 440 milliseconds. Women had regular menstrual
cycles and were neither pregnant nor taking hormonal contraceptives. Exclusion
criteria included family history of long QT syndrome, arrhythmias or sudden
death; concomitant use of any licit or illicit drug, including tobacco; and
current lactation. The study was approved by the institutional review board
of Georgetown University Medical Center, Washington, DC, and all participants
provided written informed consent.
Participants were studied in the General Clinical Research Center of
the Georgetown University Hospital. The women were studied 3 times coinciding
with the menses, ovulation, and luteal phases of the menstrual cycle. Men
were each studied once. Menses phase evaluation for women was performed within
24 to 60 hours after the onset of menses. Ovulation phase evaluation was done
24 to 48 hours after a urinary ovulation predictor test turned positive (OvuQuick,
Quidel Corporation, San Diego, Calif). Luteal phase evaluation was performed
7 to 9 days after ovulation. Women entered the study at different phases of
their menstrual cycle, but the majority (14/20) started with the menses phase
After a supine rest period of 20 minutes, a baseline ECG was recorded
using a MacVU ECG recorder (Marquette Electronics, Milwaukee, Wis). Then each
participant received ibutilide (Corvert, The Upjohn Co, Kalamazoo, Mich) 0.003
mg/kg, diluted in 20 mL of normal saline and infused over a period of 10 minutes.
Although this dose of ibutilide is only one third of the recommended antiarrhythmic
dose,11 it is known to produce significant
prolongation of the QT interval.
Timed ECGs were obtained from time = 0 (before ibutilide infusion) to
at least time = 120 minutes (at 0, 5, 10, 15, 20, 30, 40, 50, 60, 90, and
120 minutes) and, for initial safety assessment in the first 39 subjects studied,
to time = 300 minutes. All ECGs were 12 lead and were recorded on computer
disk and on paper at 50 mm per second speed with the subject in a stationary
resting supine position.
The ECGs were coded and randomized to allow blinded measurement of QT
intervals using a validated computer-operator interactive method developed
in our laboratory.12 Measured QT intervals
were corrected for heart rate using the formulae of Bazett13
(QTc = QT/RR1/2) and Fridericia14
(QTc = QT/RR1/3).14
Ibutilide concentrations in plasma were measured using high-performance
liquid chromatography and mass spectrometry detection (Agilent Technologies
series 1100 LC/MSD system, Palo Alto, Calif) after liquid-liquid extraction.
The extracted standard curve in plasma using this method was linear for ibutilide
concentrations from 25 to 500 pg/mL. The percentage coefficient of variation
(CV) at the lower limit of quantification of 50 pg/mL was 6.0; percentage
CV for the higher limit of quantification of 500 pg/mL was 1.3 on 7 runs.
Sample size was calculated to allow detection of a 30% difference in
QTc prolongation between menstrual cycle phases and each sex, with α
of .05 and power of .80. We used a 2-tailed, unpaired t test to compare single time points among men and individual phase
studies of women. Assessment of significance of difference in mean change
in QTc interval between groups was performed using analyses of variance (ANOVA).
To assess the cumulative "burden" of QTc increase over time after ibutilide
infusion, we also compared the areas under the curve (AUC) of the change in
QTc vs time over 60 minutes from onset of infusion. A similar AUC analysis
with ANOVA was performed for ibutilide concentrations measured over time =
10 to 40 minutes. A P value <.05 was considered
significant. The correlation of QTc prolongation with serum hormone levels
was calculated using the Pearson correlation coefficient.
The mean ages of the women and men were similar (Table 1). Mean weight and height of the women were both less than
those of the men, but body mass index was similar. Baseline heart rate was
the same in both sexes, but men had shorter QTc intervals than women.
Sex hormone levels fluctuated as expected with low levels of estradiol
and progesterone during menses, a peak of estradiol during ovulation, and
the highest values of progesterone during the luteal phase (Table 2). Men had the lowest estradiol serum levels, and their testosterone
was 15 times higher than that of the women. There were no significant differences
in the baseline QTc intervals during the 3 phases of the women's menstrual
cycle (Figure 1).
After ibutilide infusion there were no significant changes in the heart
rate or blood pressure and the only adverse effect observed was a self-terminating
short run (10 seconds) of asymptomatic bigeminal premature ventricular contractions
at the end of the infusion in 1 woman. All QTc results shown use the Bazett13 correction for heart rate. Analyses using Fridericia14 correction were qualitatively the same and quantitatively
very similar (data not shown).
Ibutilide infusion induced an increase in QTc in all subjects, typically
peaking within 5 minutes (t = 15 minutes) of completion
of the 10-minute infusion. In women, there was a trend, which did not reach
statistical significance, for the mean (SD) maximal change in QTc interval
to be greatest during menses (63 ), intermediate in ovulation (59 ),
and least in the luteal phase (53 ) (maximal ΔQTc, milliseconds).
Overall, men had smaller maximal increases (46  milliseconds; P = .002 vs menses, P = .007 vs ovulation, P = .12 vs luteal, and P = .004
vs the combined mean responses). Figure 2 depicts time courses of mean (SEM) increases in QTc for the 3 menstrual
cycle phases compared with men. The mean increment in QTc at every time point
measured in the first 2 hours was less in men than women during their menses
phase (Figure 2). The mean QTc change
over 0 to 60 minutes was significantly smaller in men than in women studied
at both the menses (P = .03) and ovulatory (P = .04) phases (Figure
3). Similarly, women in the luteal phase of their menstrual cycle
had the least QTc prolongation secondary to ibutilide infusion compared with
the other 2 phases (P = .06 vs menses, P = .02 vs ovulatory). For comparisons of menses phase with ovulatory
phase (P = .91) and of luteal phase with men (P = .75), there was no difference in mean QTc change over
0 to 60 minutes.
Concentrations of ibutilide were the same in men and women (mean [SD]
563  pg/mL in men vs 507  pg/mL in women; P
= .76 at t = 15 minutes; 187  pg/mL in men vs
186  pg/mL in women; P = .72 at t = 40 minutes). Comparison of AUCs of concentrations did not show
the same pattern as the QTc AUCs. In increasing order, the values (pg ×
min/mL) were 1444 for menses, 1531 for luteal, and 1658 for ovulatory, with
no significant differences between phases (repeated measures ANOVA, P = .23).
In women, both progesterone levels (r = −
0.40) and the progesterone-to-estradiol ratio (r
= − 0.41) were inversely correlated with the ibutilide-induced QT interval
prolongation (P = .001 for both). Neither testosterone
(r = 0.09, P = .46) nor
estradiol (r = 0.14, P =
.28) serum levels showed any significant correlation with the mean change
This is the first study to compare QT-prolonging effects of a drug during
phases of the menstrual cycle. It yielded the novel finding that the QTc prolongation
seen after a single infusion of a relatively low dose of ibutilide varies
with a greater response found during the first half of the cycle. Furthermore,
we found an inverse correlation between progesterone level and mean QTc change
after ibutilide but no such correlation for estradiol concentration nor for
testosterone. Our results also demonstrate a sex difference in ibutilide response
with a greater QT prolongation in women than in men, as described for other
QT-prolonging drugs.3,15 The lack
of difference between plasma ibutilide concentrations does not support a pharmacokinetic
explanation for greater QT response to ibutilide in women and suggests sex
differences in cardiac sensitivity to the drug. It is important to note that
in our protocol, we infused only one third of the clinically used initial
dose of ibutilide and due to the dose-dependent characteristics of this drug,11 it is possible that the full clinical dose would
exhibit even larger differences. Our results confirmed the well-known sex
difference in baseline QTc and are in agreement with those found by Burke
et al,10 in which no differences in the baseline
QTc intervals were seen during the 3 phases of the menstrual cycle.
The mechanism of action of ibutilide has been attributed to both activation
of a slow inward sodium current and inhibition of the potassium delayed rectifier
(IKr).16,17 It prolongs
the action potential and hence the QT interval in a dose-dependent manner.11,18 After intravenous infusion, plasma
concentrations decline multiexponentially.18
Proarrhythmia is the major adverse event with a reported incidence that ranges
up to 8% or even 36% in some reports.19,20
Most subjects enrolled in ibutilide studies have been men, but the proportion
of patients who have developed ibutilide-induced polymorphic tachyarrhythmias
is higher in women (eg, 17.5% vs 5.7% in men).19
In our study, we used a relatively small dose of ibutilide, which was sufficient
to induce moderate short-lived repolarization changes. The sex difference
in QT interval prolongation found in our study is consistent with the reported
incidence of ibutilide-induced proarrhythmia and was seen despite adjusting
dose by body weight to avoid higher plasma concentrations in smaller participants.
The absolute dose of ibutilide administered to women in our study was 25%
lower than that given to men (0.19 mg in women vs 0.25 mg in men; P = .001), and the plasma concentrations of ibutilide were similar
in both sexes. These findings indicate that the greater response in women
in this study is not attributable to differences in dosage or pharmacokinetics.
The first 60 minutes after initiation of ibutilide infusion are of prime
interest for many reasons: (1) ibutilide is in widespread use clinically,
(2) its maximum effect on QT is usually seen immediately after the end of
infusion with rapid declines in blood levels and in QT, (3) there are no known
active metabolites that could produce a delayed response, and (4) most ibutilide-induced
arrhythmias have been found within the first 30 minutes after infusion.19,20
Previous observations such as a greater drug-induced QTc interval and
a higher incidence of torsades de pointes in women and a shorter QTc interval
in men after puberty have led to the hypothesis that sex hormones influence
cardiac repolarization. To date much of the research in this field has focused
on the effect of estradiol on ion currents and cardiac repolarization,7 although its role has not been completely defined.
It is more clear that testosterone may exert a protective role, enhancing
potassium currents, shortening the action potential duration, and diminishing
the QT response to potassium channel blockers (eg, quinidine).9
Little is known about the direct or indirect effects of progesterone on cardiac
repolarizing currents and/or the effects of progesterone on the QT interval
prolongation secondary to drugs. This requires further study.
In agreement with our results, Burke et al21
reported that, after autonomic blockade, both women during the luteal phase
and men have a shorter QTc interval compared with women in the menstrual and
ovulation phases of their menstrual cycles (univariate analysis P<.005). Of note, after controlling for covariates, Burke et al
found that only the difference between sexes continued to be statistically
significant. In another study, Rashba et al22
demonstrated that the postpartum period is associated with a significant increase
in risk of cardiac events among women with congenital long QT syndrome. Since
there is an abrupt decline in estradiol and progesterone after delivery, this
finding may also reflect a role of these sex hormones.
Our study supports the observation that women are more likely than men
to develop ibutilide-induced torsades de pointes. Women during the menstrual
and ovulation phases of the menstrual cycle had the greatest QTc response,
and the findings support a complex role of sex hormones in this regard, possibly
including a protective effect of progesterone. These are novel findings that
require further investigation and confirmation. Physicians caring for patients
who are receiving drugs with potential actions on cardiac repolarization currents
should closely monitor the QT interval and beware of other risk factors for
the development of torsades de pointes. These risk factors may include the
phase of the menstrual cycle and recent pregnancy, in addition to female sex,
serum electrolyte levels (especially low potassium and/or magnesium), ischemia,
and concurrent use of other drugs with the ability to potentiate QT-prolonging