HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Request Reprints Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (7) Citing Articles via Google Scholar Google Scholar Articles by Noel, G. J. Articles by Natarajan, J. Search for Related Content PubMed PubMed Citation Articles by Noel, G. J. Articles by Natarajan, J. Social Bookmarking                   What's this?

Measuring the Effects of Supratherapeutic Doses of Levofloxacin on Healthy Volunteers Using Four Methods of QT Correction and Periodic and Continuous ECG Recordings

From Johnson & Johnson Pharmaceutical Research and Development, L.L.C., Raritan, New Jersey (Dr. Noel, S. Chien, B. Solanki, Dr. Natarajan); Department of Pediatrics, University of Medicine and Dentistry of New Jersey, Newark, New Jersey (Dr. Noel); Covance Central Diagnostics, Reno, Nevada (Dr. Goodman); and Smith-Hanley Consulting Group, Lake Mary, Florida (M. Padmanabhan).

Address for reprints: Gary J. Noel, MD, Clinical Professor of Pediatrics, UMDNJ-Newark, Johnson & Johnson Pharmaceutical Research and Development, L.L.C., 920 Route 202, Box 300, Raritan, NJ 08869.

	ABSTRACT

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
A clinical trial was conducted in healthy volunteers using both periodic and continuous ECG recordings to assess the effect of increasing doses of levofloxacin on the QT and QTc interval. Periodic and continuous ECGs were recorded before and after subjects were dosed with placebo and increasing doses of levofloxacin (500 mg, 1000 mg, 1500 mg) that included doses twice the maximum recommended dose of 750 mg in a double-blind, randomized, four-period, four-sequence crossover trial. Mean heart rate (HR) and the QT and QTc interval after dosing with levofloxacin and placebo were compared, and HR-QT interval relationships defined by linear regression analysis were calculated. After single doses of 1000 and 1500 mg of levofloxacin, HR increased significantly, as measured by periodic and continuous ECG recordings. This transient increase occurred at times of peak plasma concentration and was without symptoms. Mean QT intervals after placebo and mean intervals after levofloxacin were indistinguishable. Using periodic ECG recordings, single doses of 1500 mg were associated with small increases in QTc that were statistically significant. In contrast, an effect on QTc was shown only using the Bazett formula with data obtained from continuous ECG recordings. Together with the finding that levofloxacin does not influence HR-QT relationships, these findings suggest that levofloxacin has little effect on prolonging ventricular repolarization and that small increases in HR associated with high doses of levofloxacin contribute to the drug's apparent effect on QTc. Single doses of 1000 or 1500 mg of levofloxacin transiently increase HR without affecting the uncorrected QT interval. Differences in mean QTc after levofloxacin compared to placebo vary depending on the correction formula used and whether the data analyzed are from periodic or continuous ECG recordings. This work suggests that using continuous ECG recordings in assessing QT/QTc effects of drugs may be of value, particularly with drugs that might influence HR.

Key Words: Levofloxacin • QT-heart rate relationship • ventricular repolarization • periodic and continuous ECG recordings • QT and QTc intervals

A single-dose, dose-ranging study was conducted in volunteers using a four-way crossover, placebo-controlled study design to assess increases in QT and QTc intervals associated with plasma levels expected to be achieved after dosing with up to twice the highest recommended dose of levofloxacin. The methods used in this analysis included those used in the recently published placebo-controlled comparative trial that involved moxifloxacin, ciprofloxacin, and levofloxacin.⁵ Because the maximum effect of drug dose on QTc after treatment with levofloxacin occurred during the 4 hours after dosing in this comparative trial,⁵ this dose-escalating trial focused on measuring QT intervals over this period. In previous work, small effects on heart rate in the absence of effects on uncorrected QT raised concern that small differences in QTc calculated using widely accepted formulas (Bazett and Fridericia) may not be accurately reflecting the drug's effect on ventricular repolarization.⁶ To explore this issue further, comparisons of the measurement of QT and of QT intervals corrected for heart rate using four different formulas were made from ECGs collected using 12-lead Holter monitoring as well as standard periodic ECG recordings.

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Overview
A double-blind, randomized, placebo-controlled, four-treatment crossover, single-dose trial was conducted at a single center (PPD Development, Inc., Austin, TX). The study protocol (LOFBO-PHI-111) was reviewed and approved by Research Consultants' Review Committee (Austin, TX). Healthy subjects older than 18 years who had a normal 12-lead ECG, a heart rate between 50 and 100 beats/min, and no past medical history of cardiac disease; were not taking concomitant medications; and had a calculated creatinine clearance of > 50 mL/min were eligible for the trial. Subjects were stratified at entry by gender and age group (< 65 years, 65 years) so that there were approximately equal numbers of males and females and that approximately 16 of the 48 subjects were 65 years of age. Subjects were assigned randomly to one of four treatment sequence groups that involved four treatments with p.o. capsules (placebo and 500-mg, 1000-mg, and 1500-mg doses of levofloxacin). Each treatment was followed by a 96-hour washout period, and subjects were housed in the study unit for the duration of the trial (approximately 15 days).

Pharmacokinetic Parameters
A mean plasma levofloxacin concentration-time profile for each dose of levofloxacin was generated for the following parameters: peak concentration after dose (C_max), time to reach peak concentration after dose (t_max), and area under the concentration curve from 0 to 24 hours after dosing (AUC_0-24).

Procedures
Subjects were admitted to the study unit at least 24 hours before the first dose. Eight ECGs were recorded using a Marquette model MAC 1200 device at approximately 24, 23.5, 23, 22.5, 22, 20, 16, and 12 hours prior to dosing. Placement of leads was marked to ensure that placement was the same for each recording. The same ECG machine was used for each subject's recordings. Subjects were fasted for at least 8 hours prior to dosing and were not fed until 2 hours after dosing. Dosing was given with 240 mL of water between 7 a.m. and 9 a.m. Treatment day ECGs were recorded immediately before dosing (0 h) and then 0.5, 1, 1.5, 2, 4, 8, 12, and 24 hours after dosing. Immediately after each treatment day ECG, venous blood was sampled for measurement of levofloxacin plasma concentration. Drug concentration was measured using high-performance liquid chromatography. The assay used has been validated over concentration ranges of 0.125 to 13.75 µg/mL in plasma. ECG recordings were transmitted digitally to a central reader (Covance Diagnostics, Inc., Reno, NV). QT and RR intervals were measured manually by an experienced cardiologist who was blinded to subject treatment and treatment period. QT intervals were measured from the beginning of the QRS complex to the end of the T wave from the lead with the longest QT interval. In addition to standard ECGs, each subject had 12-lead Holter monitoring obtained for approximately 1 hour prior to dosing to 4 hours after dosing. This monitoring constituted 10-second recordings of the standard 12-lead ECG from 10 electrodes using GE/Marquette SEER-MC continuous 12-lead Holter recorders. Therefore, each subject had approximately 6 ECG recordings per minute or 180 per 30-minute epoch. ECGs from these recordings were analyzed by the automated GE/Marquette software 12-SL with no manual interpretive review. Each of these procedures was repeated for each of the four treatment periods.

Data Analyses
Primary analyses involved assessment of the relation between levofloxacin dose and QTc intervals derived from manually read 12-lead ECGs. For periodic ECG recordings, QT intervals were measured from three successive heartbeats from the lead with the longest QT interval. The RR preceding each measured QT interval was used in calculations. The average of the three QTc values calculated with each of the correction formulas was used for analysis. For continuous ECG recordings, QT values were extracted from the automated algorithmic results. The algorithm determines the QT from a set of 12 median beats electronically generated from all the normally conducted complexes for each lead. The interval is defined as the earliest onset of the Q wave in any lead to the latest offset of the T wave in any lead. Heart rate and RR were calculated in this algorithm over the entire trace.

For periodic recordings, heart rate-corrected QT intervals were calculated using the average of the three QTc values calculated using the Bazett formula (), the Fridericia formula (), and the Framingham formula (QTc = QT + 0.154 (1 - RR)), where RR is expressed in seconds. Individual correction formulas were calculated based on linear correction methods described by Malik⁶ using a total of 44 ECGs on each subject recorded predose in each of the four treatment periods. Data from 44 ECG recordings are similar to the number reported by Malik (i.e., 41) to calculate this correction⁶ but are less than the minimum of 50 suggested by Malik and Camm⁷ to be sufficient for optimal calculation of this correction. For each subject, a linear regression model was fit to the predose QT data with RR as a predictor, and the intercept and slope of the regression line were estimated. The postdose QT values were corrected for heart rate using the formula QTc = QT + b(1 - RR), where b is the estimated slope for that subject.

For continuous recordings, heart rate-corrected QT intervals were calculated using the Bazett, Fridericia, and Framingham formulas defined above. Individual corrected QTc was also calculated as described above using the estimated slope from the best fit (least squares) for the QT-RR relationship for all values of QTRR from the predose ECGs for each subject. The equation used for linear fit was E(QT| RR) = (intercept + slope • RR), where E(QT| RR) denotes the expected value of QT given RR. The postdose QT values were corrected for heart rate using the formula QTc = (QT + estimated slope • (1 - RR)). QT and RR intervals were averaged over 30-minute periods (or epochs) that began 1 to 0.5 hours before dosing and continued for 4 to 4.5 hours after dosing. Epochs were labeled based on the time (minute or hour) that began the epoch (i.e., 1.5-2 h = epoch 1.5 h or epoch 90 min).

The heart rate, QT, and QTc values obtained at defined times or epochs after levofloxacin or placebo administration were evaluated using mixed-effects models with treatment sequence group, period, treatment, scheduled postdose time of measurement, and the interaction between treatment and postdose time point of measurement as fixed effects and subject as a random effect. At each postdose time of measurement, the effect of each dose of levofloxacin was compared to that of placebo using appropriate linear contrasts. An overall 5% level of significance was used for these comparisons.

Change from baseline of QTc intervals postdosing was measured using two commonly used heart rate correction methods (Bazett and Fridericia) and four methods of determining baseline from data collected from periodic ECG measurements. Baselines were calculated using the following four methods: (1) the mean of the 0-, 4-, and 8-hour values on the predose day and the 0-hour value on the dosing day for a subject at the given period; (2) the mean of the values from the 4 predose days at 0, 0.5, 1, 1.5, 2, 4, 8, and 12 hours for a subject at the given time point of measurement for all four periods; (3) the mean of the 0-, 0.5-, 1-, 1.5-, 2-, 4-, 8-, 12-, and 24-hour values on the placebo dosing day for a subject for all four periods; and (4) the mean of the 0-, 0.5-, 1-, 1.5-, 2-, 4-, 8-, and 12-hour values on the predose day and the 0-hour value on the dosing day for a subject at the given period.

The effect of the treatment on the relationship between heart rate and QT was assessed using the QT and RR values obtained from continuous ECG during the period of maximal drug exposure (1.5-3.0 h postdose = epochs 1.5-2.5 h) after levofloxacin and placebo dosing. For each subject and treatment (placebo, 500 mg, 1000 mg, and 1500 mg), a linear regression model was fit to the QT values with RR as a predictor, using all QT-RR data obtained 1.5 to 3.0 hours postdose, and the intercept and the slope of the regression line were estimated. The effect of treatment on the QT-RR relationship was evaluated by comparing the estimated intercept and slope across the four treatments. Mixed-effects models were fitted to the data with the estimated intercept or slope as the dependent variable, treatment as a fixed effect, and subject as a random effect. The treatment effect was tested at a 5% level of significance.

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Subject Demographics
Forty-eight of the 50 healthy volunteers (mean age = 47.1 years; age range = 21-77; 17 subjects 65 years) who were randomized and started the trial completed the study. One subject, a 43-year-old woman, withdrew for personal reasons after receiving placebo and 1500 mg of levofloxacin. A second subject, a healthy 73-year-old male, was withdrawn from the study after being noted to have an episode of ventricular bigeminy after receiving placebo.

Pharmacokinetics
In subjects completing the trial receiving the 500-mg, 1000-mg, and 1500-mg levofloxacin doses, the mean C_max values (± SD) were 5.43 ± 1.76 µg/mL, 10.2 ± 2.91 µg/mL, and 14.6 ± 3.49 µg/mL, respectively. In subjects completing the trial receiving the 500-mg, 1000-mg, and 1500-mg levofloxacin doses, the mean t_max values (± SD) were 1.7 ± 0.5 hours, 2.0 ± 0.9 hours, and 2.2 ± 1.0 hours, respectively. In subjects completing the trial receiving the 500-mg, 1000-mg, and 1500-mg levofloxacin doses, the mean AUC_0-24 values (± SD) were 54.1 ± 14.0 µg•h/mL, 122 ± 33.2 µg•h/mL, and 203 ± 50.2 µg•h/mL, respectively. The mean plasma concentrations of levofloxacin over the 4 hours after dosing are shown in Figure 1 (lower panel).

View larger version (28K): [in this window] [in a new window]   Figure 1. Mean heart rate, mean QT interval, and mean plasma concentration of levofloxacin in 48 healthy subjects after receiving a single dose of either placebo () or 500 mg (), 1000 mg (), or 1500 mg () of levofloxacin. Mean values of heart rate and of the QT interval are shown for measurements obtained from continuous ECG recordings (upper panel) and from periodic ECG recordings (middle panel). Asterisks indicate values that are significantly different from the mean value after placebo at the corresponding time or epoch after dosing.

Effects on Heart Rate and Uncorrected QT Intervals
The mean heart rate and QT interval, measured using continuous monitoring (Figure 1, upper) and periodic measurements (Figure 1, middle panel) for the 4 hours after treatment dose, were similar but not identical for the two methods used to collect these data in these healthy adult volunteers. As expected, because continuous monitoring permitted subjects to be active and periodic measurements were done in subjects at rest, the mean heart rate measured with continuous monitoring was approximately 10 beats per minute greater than the mean heart rate measured with periodic ECG recordings. The mean heart rate of subjects 1.5 and 2 hours after dosing with 1000 and 1500 mg of levofloxacin using periodic ECG recordings and of subjects at epochs 1, 1.5, and 2 hours after dosing were significantly greater (p < 0.05) than the mean heart rate of these subjects after treatment with placebo. These data indicate that single 1000-mg or 1500-mg doses of levofloxacin increased heart rate in healthy volunteers. For the highest dose given (1500 mg), the increase in mean heart rate compared to placebo was 7.75 beats per minute at epoch 2 hours (measurements 2.0-2.5 h after dosing), as measured by continuous ECG, and was 4.7 beats/min at 2 hours, as measured by periodic ECG recordings.

Although mean QT intervals trended to be slightly shorter, especially with values measured in continuous monitoring, after 1000-mg and 1500-mg doses of levofloxacin, differences in mean QT values were not statistically significant compared to mean values after placebo (Figure 1).

Effects on heart rate-corrected QT intervals. To estimate the effect of drug on changes in ventricular repolarization, it is necessary to consider the normal physiologic relationship between heart rate and QT interval. This relationship was considered in calculating heart rate-corrected QT intervals (QTc) using two methods commonly used by clinicians (Bazett and Fridericia) and two less commonly used but widely accepted methods (individual and Framingham) that are considered more precise than either Bazett or Fridericia in correcting for heart rate when rates differ substantially from 60 beats/min.⁶

Mean QTc (Bazett and Fridericia) values calculated for subjects at selected times 24 hours before and after dosing are shown in Table I. Times selected include those periods of maximal effect on QTc (1.5-2 h after dose) and corresponding periods during the day before dosing (-22.5 and -22 h). Mean values were statistically significant from placebo for the 1500-mg dose at both 1.5- and 2-hour periods. Mean QTc values at 8 and 12 hours after dosing with levofloxacin were indistinguishable from mean values calculated after treatment with placebo and from those values calculated for the corresponding periods (-16 and -12 h) prior to dosing (data not shown). Consistent with results previously published,⁵ the period of greatest effect of levofloxacin on QTc was evident during the first 4 hours after exposure (Table I and Figure 2).

View larger version (39K): [in this window] [in a new window]   Figure 2. Mean heart rate-corrected QT (QTc) intervals of 48 subjects after receiving a single dose of either placebo () or 500 mg (), 1000 mg (), or 1500 mg () of levofloxacin. QTc values were calculated using four different methods: individual correction formula (upper left panel), Framingham formula (upper right), Fridericia formula (lower left), and Bazett formula (lower right). Values are shown for measurements obtained on these 48 subjects for the 4-hour period after dosing with levofloxacin or placebo from periodic (upper) and continuous (lower) ECG recordings in each panel. Asterisks indicate values that are significantly different from the mean value after placebo at the corresponding time or epoch after dosing.

The mean QTc intervals of subjects for the 4 hours after treatment are shown in Figure 2 for measurements taken using continuous and periodic ECG recordings. The difference between mean QTc values after treatment with levofloxacin and mean QTc values after placebo varied depending on whether measurements were made using continuous or periodic ECG recordings and on the correction methods used to calculate QTc. Using the individual correction method, the only mean QTc value after treatment with levofloxacin that was significantly greater than the corresponding mean QTc value after placebo was the value measured at 1.5 hours after dosing using data from periodic ECG recordings (402.4 msec for placebo vs. 409.5 msec for 1500 mg levofloxacin). Using values obtained from periodic ECG recordings, analyses using the Framingham and Fridericia formulas yielded similar results. Using these formulas to calculate values obtained from periodic ECG recordings, mean values were significantly greater at 1.5 and 2 hours after dosing with 1500 mg of levofloxacin than after placebo. For the Framingham correction method, the difference between mean QTc values at 1.5 and 2 hours was 10.9 msec (398.4 vs. 409.3 msec) and 11.2 msec (398.1 vs. 409.3 msec), respectively. For the Fridericia correction method, the difference between mean QTc values at 1.5 and 2 hours was 9.2 msec (400.5 vs. 409.7 msec) and 9.5 msec (400.1 vs. 409.6 msec), respectively. For the continuous ECG recordings, mean QTc values after levofloxacin treatment calculated using individual correction, Framingham, and Fridericia formulas were not significantly different from mean QTc values after placebo.

Compared to the other correction formulas, analysis using the Bazett formula demonstrated the largest differences between mean QTc values after levofloxacin compared to placebo. In contrast to the other correction formulas, these differences were evident in values obtained from both periodic and continuous ECG recordings. The difference in mean values of QTc (Bazett) after the 1500-mg dose was significantly greater compared to placebo at 1.5 and 2 hours after dosing using values calculated from periodic ECG recordings (1.5 h: 399.5 vs. 415.3 msec; 2.0 h: 398.9 vs. 414.1 msec). Mean QTc (Bazett) values were significantly greater at epochs 1 through 3 hours after dosing compared to mean values after placebo, using values calculated from continuous ECG recordings. These differences ranged from 8.5 msec (epoch 2 h: 429.7 vs. 438.3 msec) to 5.3 msec (epoch 3 h: 426.9 vs. 432.2 msec).

Comparison of mean change in QTc over the 24-hour period after exposure to baseline QTc intervals also demonstrated an effect of levofloxacin on these intervals (Table II). In these analyses using Bazett and Fridericia correction methods and baselines calculated using four different methods, the degree of change varied with dose and methods used. This change from baseline was not statistically significant using any of these methods in subjects exposed to 500 mg of levofloxacin. Statistically significant changes from baseline were evident using some but not all methods in subjects after exposure to 1000 mg. The degree of mean change (1.55-3.93 msec for QTc Bazett and 1.38-2.79 msec for QTc Fridericia) was consistent with those mean changes previously reported⁵ after single 1000-mg doses of levofloxacin. In all of the methods used, changes from baseline were statistically significant after the 1500-mg dose. The observation that the effect of drug on QTc may vary with the method used is consistent with that previously reported⁵ and underscores the importance of considering methods and factors that may potentially influence methods in assessing the relationship between a drug's effect on QTc.

Heart Rate-QT Relationships
The small effect of high doses of levofloxacin on heart rate and the discrepancies among the size and identification of differences in mean QTc values, depending on the methods used to correct QT for heart rate, suggested that heart rate effects may be contributing to the apparent differences between mean QTc calculated in subjects given high-dose levofloxacin and the mean QTc values in subjects given placebo. To explore this issue further, an analysis was performed using data obtained from continuous ECG recordings. It is expected that drugs slowing ventricular repolarization will change the relationship between heart rate and the QT interval. Recent studies exploring the value of these types of analyses underscore the importance of establishing this relationship by performing multiple, perhaps hundreds, of ECGs on individual subjects to establish relationships between heart rate and the QT interval in the absence and presence of a drug.^6,7 Continuous ECG recordings provided this opportunity. Linear regression analysis was performed using heart rates and QT values to establish the relationship between these two values for individual subjects after placebo and levofloxacin treatments. Epochs 1.5 to 2.5 hours were selected for this evaluation because this time interval corresponded to the time of maximal drug exposure. This time interval also corresponded to the period associated with the largest differences between mean QTc (Bazett) after 1500 mg and mean QTc (Bazett) after placebo using methods involving both periodic and continuous ECG recordings. Table III shows the slope and y-intercept of lines defined by these analyses for each of the four treatments. The mean slope and mean intercept of lines defining the heart rate-QT interval relationship for subjects after levofloxacin treatment were indistinguishable from those measured after placebo treatment. This finding supports that in this group of healthy volunteers, levofloxacin does not have an effect on the heart rate-QT relationship, consistent with the drug-prolonging ventricular repolarization.

Adverse Events
Nine subjects had adverse events of special interest (events consistent with delayed ventricular repolarization and/or dysrhythmia) that included palpitation, syncope, and dizziness. In these 9 subjects, eight episodes of dizziness occurred after levofloxacin (one after 500 mg, two after 1000 mg, and five after 1500 mg), three episodes of palpitations occurred after levofloxacin (two after 500 mg and one after 1500 mg), and one episode of syncope occurred 2 days after a subject was dosed with 1000 mg of levofloxacin.

Six episodes of dizziness and one episode of palpitation occurred during the period of continuous ECG recording. In five of these episodes (four dizziness episodes after 1500 mg and one dizziness episode after 1000 mg), no abnormality in heart rate, heart rhythm, or prolongation of QT interval was evident on review of the continuous ECG recordings. One 68-year-old woman had symptoms of dizziness and palpitations over an hour before five isolated unifocal premature ventricular contractions (PVCs) were noted over a 50-second period about 2.5 hours after she had been dosed with 1500 mg of levofloxacin. No other abnormalities in rate, rhythm, or QT interval were detected in this subject's ECG, and the relationship between her symptoms and her cardiac function was not evident. In summary, none of symptoms experienced by subjects dosed with levofloxacin appeared to be related to changes in heart rate or abnormalities in ECG recordings.

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
The results presented here demonstrate that levofloxacin can increase the heart rate of healthy volunteers after single doses of 1000 and 1500 mg. These small effects on heart rate did not appear to be associated with symptoms. These effects appeared to be maximal at times, coincident with the highest plasma concentration of levofloxacin. Interestingly, effects on heart rate were not evident by 4 hours after dosing. At 4 hours, plasma concentrations of the drug were comparable to those measured at 1.5 hours after dosing when an effect on heart rate was evident. The basis for this inconsistent association of plasma concentration and increased heart rate is not understood but may reflect a chronotropic effect of the drug that is specific to the rise of drug concentration to C_max or the sensitivity of detecting small effects on heart rate during periods of higher heart rates that occur as part of normal diurnal variation. Regardless of the underlying cause of this effect, small influences on heart rate can affect the calculation of the heart rate-corrected QT interval,⁶ and therefore this observation was considered to be important in developing our understanding of the potential for levofloxacin to prolong ventricular repolarization, as demonstrated by increases in QTc.

Uncorrected QT intervals were not increased compared to placebo by levofloxacin, even at doses as high as 1500 mg. This was demonstrated using both periodic ECG and continuous ECG recordings. Interestingly, the values derived from continuous ECG recordings showed the mean QT intervals to be consistently shorter by as much as 10 msec at 1.5 and 2 hours after dosing with 1000 or 1500 mg of levofloxacin. Although these differences were not statistically significant, they are consistent with a trend that would be expected of an effect on QT interval associated with small increases in heart rate. This trend of a decreased uncorrected QT interval associated with increasing plasma concentrations of drug contrasts that which has been described for moxifloxacin, a fluoroquinolone that has been shown to increase the uncorrected QT interval,⁵ have a substantially greater effect on IKr channel function than levofloxacin,⁴ and have an influence on the heart rate-corrected QT interval.⁸ The results reported here further underscore the variability of the degree of effect on QTc depending on methods and analyses used to assess this effect. Using periodic ECG recordings and the heart rate correction method, which is generally accepted as the most accurate in correcting for effects of heart rate on the calculation of QTc,^6,8 small increases in QTc intervals after a single 1500-mg dose of levofloxacin were demonstrated. This effect on QTc was substantially less than that demonstrated in the analysis using the Bazett correction method. At the 1.5-hour time period, the difference between mean QTc values calculated using an individual correction method after dosing 1500 mg and after dosing placebo was less than half that calculated using the Bazett method (see Figure 2; 7.12 vs. 15.85 msec). In analysis of data collected from continuous ECG recordings, no dose-response effect of levofloxacin on QTc could be demonstrated using the individual correction method, whereas analysis using the Bazett method demonstrated significant increases in QTc after the 1500-mg dose of levofloxacin compared to placebo.

As is the case with most trials focused on measuring effects of a drug on QTc intervals, the results of this single, escalating dose trial do not establish the clinical importance of levofloxacin's effect on QTc. A variety of factors were not examined in this trial involving healthy volunteers, including synergistic effects with other drugs that cause slowing of ventricular repolarization and underlying medical conditions, and these factors should be considered when assessing the clinical importance of a drug's effect on the QTc interval. Furthermore, this trial was conducted with single doses of levofloxacin. Because levofloxacin pharmacokinetics are linear and highly predictable and it has been shown that differences in C_max and AUC_0-24 after single doses differ minimally from C_max and AUC_0-24 after multiple daily doses, it seems unlikely that the effect of multiple doses of levofloxacin on QTc would be markedly different from the results observed after single doses. The results reported here indicate that single, albeit high, doses of levofloxacin can influence QTc in healthy volunteers, and this should be considered to establish that exposure to levofloxacin has the potential for influencing this calculated interval. Although this does not establish that levofloxacin can cause clinically meaningful changes in cardiac function, the observation that 1500-mg doses of levofloxacin can affect the calculated QTc interval should be considered when treating patients with multiple doses of this drug.

The value of continuous ECG recordings in measuring the effect of drugs on ventricular repolarization has been considered but is not widely accepted. It has been suggested that continuous ECG recordings may have advantages over periodic measurements to assess effects of drugs on the QT interval. The basis for this suggestion is that continuous recordings provide hundreds of measurements, and averaging the large number of these measurements over defined periods may best consider the normal physiologic variation in heart function. Concerns over using continuous ECG recording include the reliance on machine-based measurement of intervals and the potential to miss critical abnormalities in ECG recordings such as U waves, which may be better defined by experienced cardiologists manually reading recordings than by machine-based readings. The experience reported here demonstrates that continuous ECG recordings are capable of detecting similar trends in heart rate and QT interval that were measured with standard periodic ECG recordings. Data from continuous and periodic ECG recordings both demonstrated the largest effects of the 1500-mg dose of levofloxacin on QTc (Bazett). Although these differences became less apparent when non-Bazett correction formulas were applied to measurements from periodic ECG recordings, these differences were not apparent when non-Bazett formulas were applied to measurements from continuous recordings. Last, the collection of hundreds of ECGs over the period of highest drug exposure provided the opportunity to assess the effect of drug on the QT-heart rate relationship. This analysis suggested that levofloxacin does not influence this relationship in a manner as would be the case for a drug-prolonging ventricular repolarization. Furthermore, this is consistent with levofloxacin's effect on heart rate being the primary basis for apparent effects on QTc and the variation in QTc differences associated with levofloxacin exposure that occurs when different formulas are used to correct for heart rate.

Taken together, the analyses performed using data collected from periodic and continuous recordings of ECGs in healthy subjects challenged with increasing doses of levofloxacin demonstrate that high doses of levofloxacin can increase heart rate at times immediately before and at points of maximal levofloxacin plasma concentration. This effect on heart rate confounds analyses that have been used to assess the influence of drugs on ventricular repolarization, especially the use of the Bazett formula to correct QT for heart rate. Although periodic ECG measurement remains the "gold standard" for assessing effects on QTc, the experience reported here with continuous ECG recordings suggests that these methods may be valuable in assessing drug effects on ventricular repolarization. These methods could have advantages over using periodic ECG recordings because they could be used readily to assess QT-heart rate relationships associated with ventricular repolarization. In addition, these methods might be used to complement traditional QT/QTc trial designs to address specific issues that may arise with drugs that influence heart rate independent of repolarization events.

	ACKNOWLEDGEMENTS

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
We thank Dr. Robert Abels and Dr. Thomas Hunt for their contribution to this work and Dr. Christian Funck-Brentano for his thoughtful comments and review of these data.

	FOOTNOTES

 
DOI: 10.1177/0091270004264643

Submitted for publication December 18, 2003; Revised version accepted February 15, 2004.

	REFERENCES

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 

1. Owens RC: Risk assessment for antimicrobial agent-induced QTc interval prolongation and torsades de pointes. Pharmacotherapy 2001;21: 310-319.

2. Rubenstein E, Camm J: Cardiotoxicity of fluoroquinolones. J Antimicrob Chemother 2002;49: 593-596.[Free Full Text]

3. Stahlman R, Lode H: Safety overview, in: Andriole V (ed.), The Quinolones. 3rd ed. New York: Academic Press, 2000; 397-453

4. Kang J, Wang L, Chen XL, Triggle DJ, Rampe D: Interactions of a series of fluoroquinolone antibacterial drugs with the human cardiac K⁺channel HERG. Mol Pharmacol 2001;59: 122-126.[Abstract/Free Full Text]

5. Noel GJ, Natarajan J, Chien S, Hunt TL, Goodman DB, Abels R: Effects of three fluoroquinolones on QT interval in healthy adults after single doses. Clin Pharmacol Ther 2003;73: 292-303.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]

6. Malik M: Problems of heart rate correction in assessment of drug-induced QT interval prolongation. J Cardiovasc Electrophysiol 2001;12: 411-420.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]

7. Malik M, Camm AJ: Evaluation of drug-induced QT interval prolongation. Drug Saf 2001;24: 323-351.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]

8. Demolis JL, Kubitza D, Tenneze L, Funck-Brentano C: Effect of a single oral dose of moxifloxacin (400mg and 800mg) on ventricular repolarization in healthy subjects. Clin Pharmacol Ther 2000;68: 658-666.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]

9. Chien S-C, Rogge MC, Gisclon LG, Curtin C, Wong F, Natarajan J, et al: Pharmacokinetic profile of levofloxacin following once-daily 500mg oral or intravenous doses. Antimicrob Agents Chemother 1997;41: 2256-2260.[Abstract]
CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    What's this?

This article has been cited by other articles:

				M. Iwamoto, J. T. Kost, G. C. Mistry, L. A. Wenning, S. A. Breidinger, T. C. Marbury, J. A. Stone, K. M. Gottesdiener, D. M. Bloomfield, and J. A. Wagner Raltegravir Thorough QT/QTc Study: A Single Supratherapeutic Dose of Raltegravir Does Not Prolong the QTcF Interval J. Clin. Pharmacol., June 1, 2008; 48(6): 726 - 733. [Abstract] [Full Text] [PDF]

				D. Hutschala, K. Skhirtladze, A. Zuckermann, W. Wisser, P. Jaksch, B. X. Mayer-Helm, H. Burgmann, E. Wolner, M. Muller, and E. M. Tschernko In Vivo Measurement of Levofloxacin Penetration into Lung Tissue after Cardiac Surgery Antimicrob. Agents Chemother., December 1, 2005; 49(12): 5107 - 5111. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Request Reprints Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (7) Citing Articles via Google Scholar Google Scholar Articles by Noel, G. J. Articles by Natarajan, J. Search for Related Content PubMed PubMed Citation Articles by Noel, G. J. Articles by Natarajan, J. Social Bookmarking                   What's this?