`
`ARTICLES
`
`Thorough QT Study with Recommended and
`Supratherapeutic Doses of Tolterodine
`
`BK Malhotral, P Glue‘, K Sweeneyz, R Anzianoz, I Mancusoz and P Wickerz
`
`The objective of our study was to determine the QTc effects of tolterodine. A crossover-design thorough QT study of
`recommended (2 mg twice daily) and supratherapeutic (4 mg twice daily) doses of tolterodine, moxifloxacin (400 mg
`once daily), and placebo was performed. Electrocardiograms (ECGs) and pharmacokinetic samples were obtained on
`days 1-4; time—matched baseline ECGs were taken on day 0. Mean p|acebo—subtracted change from baseline Fridericia—
`corrected QT (QTcF) during peak drug exposure on day 4 was the primary end point. Mean QTcF prolongation of
`moxifloxacin was 8.9 ms (machine-read) and 19.3 ms (manual-read). At recommended and supratherapeutic tolterodine
`doses, mean QTcF prolongation was 1.2 and 5.6 ms (machine-read), respectively, and 5.0 and 11.8 ms (manual-read),
`respectively. The QTc effect of tolterodine was lower than moxifloxacin. No subject receiving tolterodine exceeded the
`clinically relevant thresholds of 500 ms absolute QTc or 60 ms change from baseline. In conclusion, tolterodine does not
`have a clinically significant effect on QT interval.
`
`Tolterodine is an antimuscarinic agent approved for the
`treatment of overactive bladder.‘‘3 The effects of tolterodine
`on cardiac ion channels have been evaluated in virro. A
`
`thorough QT (TQT) study, reported here, was performed to
`determine any effect of tolterodine on cardiac repolarization.
`Tolterodine is eliminated primarily by metabolism invol-
`ving cytochrome P450 (CYP) 2D6 and 3A4.‘‘‘5 CYPZD6
`exhibits genetic polymorphism, and thus the metabolic
`disposition of tolterodine is different between subjects with
`normal CYPZD6 activity (extensive metabolizers (EMs)) and
`those with CYPZD6 deficiency (poor metabolizers (PMs)).6’7
`Metabolism of tolterodine by CYP2D6 in EMs results in the
`formation of an equipotent pharmacologically active meta-
`bolite, 5-hydroxy-methyltolterodine
`(DD01).8‘9
`In PMs,
`however, CYP3A4 is involved in the formation of an inactive
`metabolite (Figure l).8 The unbound fraction of tolterodine
`and DD01 in serum is 3.7% and 36%, respectively.7‘m'”P
`tent CYP3A4 inhibitors increase peak tolterodine exposure
`about twofold, requiring tolterodine dose reduction to 2 mg)‘
`day when used concomitantly}; Tolterodine is available as an
`immediate-release (IR) tablet (1 or 2 mg twice daily) and as
`an extended—release (ER) capsule (2 or 4mg once daily).m‘”
`The peak concentrations (Cmx) of tolterodine and DD0l
`with ER capsules are approximately 61% and 67% lower,
`respectively,
`than with IR tablets.” For both IR and ER
`formulations, EMs have comparable C,,,,,,. of tolterodine and
`
`DD01.“"” In PMs, DD01 is not formed and tolterodine Cm“
`is six— to eightfold higher than in EMs.'"‘” However, because
`of the 10—fold
`lower unbound fraction of tolterodine,
`exposure to the active moiety (i.e.,
`sum of unbound
`tolterodine and DDOI)
`is comparable between EMs and
`PMs, and no dose adjustment
`is necessary by CYPZD6
`metabolizer status.]°’“
`
`A recent in vitro study of the effects of tolterodine on
`cardiac ion channels showed that it was a potent inhibitor of
`both human ether a—go—go—re1ated gene (HERG) cardiac
`potassium and L—type calcium channels.” The HERG
`inhibitory concentration 50% (IC50) value was reported to
`be 17nM, with some inhibition evident at concentrations as
`
`low as 3 nM. In comparison, the steady-state Cm,‘ of unbound
`tolterodine at the therapeutic dose is much lower, averaging
`0.31 nM in EMs and 1.33 nM in PMS after IR administration.”
`
`In the guinea pig ventricular myocytes model, even at
`concentrations far exceeding those associated with therapeu-
`tic exposures of tolterodine,
`there was a much smaller
`prolongation of action potential duration than that observed
`with the pure HERG/repolarizing cardiac potassium current
`(In) antagonist dofetilide.”
`Electrocardiogram (ECG) data from both healthy subjects
`and patients in clinical trials have not shown any clinically
`relevant QTc interval prolongation. Since the first approval in
`1997, nearly 10 million patients have been exposed to
`
`‘Pfizer Inc., New York, New York, USA; 2Pfizer |nc., Groton, Connecticut, USA. Correspondence: BK Malhotra (bimal.k.ma||1otra@pfizer.com)
`Received 8 August 2006; accepted 25 November 2006. doi:10.1038/'sj.c|pt.61000B9
`
`CLINICAL PHARMACOLOGY 8: THERAPEUTICS | VOLUME 81 NUMBER 3 | MARCH 2007
`
`377
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2071 - 0001
`
`
`
`ARTICLES
`
`Tolterocline
`
`DD01
`
`Tolierodine acid
`
`
`
`Dealkylaied tolterodine
`
`Dealkylated hydroxylated tolterodine
`
`Dealkylated tolterodine acid
`
`Figure 1 Metabolic pathways of tolterodine. In CYPZD6 EMs, tolterodine is preferentially metabolized to an active metabolite DD01, whereas in CYPZD6 PMS,
`it is metabolized to pharmacologically inactive dealkylated tolterodine via CYP3A4.
`
`tolterodine (data on file). A review of all postmarketing
`cardiovascular events associated with tolterodine use has
`
`revealed no cases where tolterodine caused QT prolongation
`or torsade de pointes (TdP; data on file). A Prescription Event
`Monitoring study in the United Kingdom found no reports
`of TdP or sudden death in more than 14,000 patients
`receiving a prescription for tolterodine IR; only 17 cases
`(0.1%) of tachycardia or palpitations were possibly or
`probably related to tolterodine.” Four of 29 other arrhyth-
`mias were judged to be possibly or probably related to
`tolterodine use. None represented a serious arrhythmia, such
`as ventricular tachycardia or fibrillation.
`Despite the lack of cardiac arrhythmic events potentially
`related to QT interval prolongation in the extensive clinical
`experience accumulated so far with tolterodine, considering
`positive results in the preclinical HERG assay, a TQT study
`was deemed to be important for a definitive assessment of the
`effects of this agent on the QT interval. With a
`few
`exceptions, regulatory agencies now require a TQT study,
`not only for drug candidates in clinical development, but also
`for marketed drugs.l5‘”' The analysis and interpretation of
`the results of this TQT study took into account
`the
`International Conference on Harmonization (ICH) harmo-
`nized tripartite E14 guidelines.” In addition, we compared
`results using both machine- and manual-read interval data.
`Previous experience has shown that manual over-reading
`may introduce increased variability and bias into the ECG
`data sets.”
`
`RESULTS
`
`Subject demography
`Forty-eight subjects (25 men and 23 women) entered the
`study and received 21 close of study medication. The study
`population was enriched in PMs (45% ) compared with the
`natural frequency of about 7%.” MeaniSD of the subjects’
`age, height, and body mass index were 36.7J_r10.9 years,
`1.71 i0.09 m, and 25.4i3.5 kglmz,
`respectively. Approxi-
`
`378
`
`mately one-third of the subjects were 45-55 years of age.
`Three subjects discontinued from the study (one each
`because of adverse event, protocol deviation, and personal
`reasons).
`
`Adverse events
`
`Forty-eight adverse events were reported by 23 subjects given
`tolterodine 2 mg twice daily, 48 adverse events were reported
`by 25 subjects given tolterodine 4mg twice daily, 31 adverse
`events were reported by 18 subjects given placebo, and 70
`adverse events were reported by 25 subjects given moxiflox-
`acin 400 mg once daily. The majority of adverse events
`reported were treatment related. The most frequent adverse
`events were headache, dry mouth, and nausea. One subject
`reported moderate chest pain and was noted to have ST-
`segment depression after administration of placebo and
`discontinued from the study.
`
`Pharmacokinetics of tolterodine and moxifloxacin
`
`Tolterodine was quickly absorbed with median time of Cm,
`(tmax) of 1h, regardless of dose and genotype. Tolterodine
`C,,,,,,, and area under the concentration4ime curve (AUC)
`increased proportionally with dose and were approximately
`3-5 and 10 times, respectively, higher in PMs than in EMS
`(Table 1). Exposure ofDD01 in PMs was 6-7 times less than
`EMS. Median r,,,,,,, of DD01 was approximately 1 h regardless
`of dose. Moxifloxacin Cm“ was consistent with its published
`data in the package insert. The median rm, of moxifloxacin
`was 2h. These pharmacokinetic data support the statistical
`analysis of QTc values at 1 and 2 h postdose for assessment of
`QTC prolongation at the time of maximum exposures of
`tolterodine and moxifloxacin, respectively.
`The alpha,-acid glycoproteinl-acid glycoprotein (AAG)
`concentrations were generally similar across all four periods.
`The mean:SD values of unbound fractions of tolterodine
`and DD01, calculated from the AAG concentrations in
`
`serum, were 0.026i0.005 and 0.302i0.044, respectively.
`
`VOLUME 81 NUMBER 3 | MARCH 2007 l www.nature.comlcpt
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2071 - 0002
`
`
`
`ARTICLES
`
`Table 1 Mean (SD) pharmacokinetic values of tolterodine, DDD1, and moxifloxacin following single dose (clay 1) and at steady
`state (day 4)
`
`Anamel study day
`
`Tolterodine 2mg b.r'.d.
`
`Tolterodine, day 1
`
`Tolterodine, day 4
`
`DD01, day 1
`DD01, day 4
`
`Tolterodine 4mg b.i.d.
`
`Tolterodine, day 1
`
`Tolterodine, day 4
`
`DD01, day 1
`DD01, day 4
`
`Moxifloxacin 400mg q.d.
`
`Moxifloxacin, day 1
`
`Moxifloxacin, day 4
`
`EMs
`
`3.1 (2.4)
`
`3.4 (2.9)
`
`2.5 (1.0)
`2.8 (1.0)
`
`6.1 (5.5)
`
`6.6 (5.7)
`
`4.9 (2.1)
`5.5 (1.7)
`
`Cm. (ng/ml)
`PMs
`
`All
`
`EMS
`
`AUC"’ (ng h/ml)
`PMs
`
`All
`
`9.1 (3.7)
`
`15.8 (8.3)
`
`0.2 (0.6)
`0.3 (0.7)
`
`20 (11)
`
`34 (19)
`
`0.5 (1.1)
`0.7 (1.4)
`
`5.9 (4.3)
`
`9.1 (8.7)
`
`1.5 (1.4)
`1.7 (1.5)
`
`12 (11)
`
`19 (19)
`
`2.9 (2.8)
`3.3 (2.8)
`
`12 (14)
`
`13 (5.1)
`
`26 (31)
`
`NC
`28 (8.4)
`
`NC
`
`128 (78)
`
`NC
`2.0 (4.7)
`
`NC
`
`266 (166)
`
`65 (79)
`
`8.4 (7.7)
`
`136 (166)
`
`4.7 (8.6)
`
`17 (15)
`
`NA
`
`NA
`
`3000 (718)
`
`3610 (782)
`
`NA
`
`NA
`
`NC
`
`NC
`
`AUC, area under the concentration—time curve; b.i.d., twice daily: EM, extensive metabolizer, NA, not applicable; moxifloxacin pharmacokinetics unaffected by CYP2D6
`genotype: NC, not calculated; samples were collected and/or analyzed up to 4h postdose; PM, poor metabolizer; q.d., once daily. ‘AUC calculated up to the time of last
`quantifiable concentration during the 12-h dosing interval.
`
`Central tendency (mean) analysis of QTc effects of toIter-
`odine and moxifloxacin
`
`The time course of AA Fridericia-corrected QT (QTcF)
`values on day 4, by treatment,
`is shown in Figures 2
`and 3, using the machine-
`and manual-read ECGS,
`respectively. The mean increase of heart rate associated with
`tolterodine 2mg twice daily was 2.0, and 6.3beatsfmin
`with tolterodine 4mg twice daily. The change in heart rate
`with moxifloxacin was 0.5 beats/min. The results for AAQTC
`at
`11,,“
`(1 h for
`tolterodine, 2h for moxifloxacin)
`are
`summarized in Tables
`2
`and 3. Mean AAQTCF for
`tolterodine 2mg twice daily was 13% (machine) and 26%
`(manual) of that after moxifloxacin. For 4mg twice-daily
`tolterodine, AAQTcF was 63% (machine) and 61% (manual)
`of that after moxifloxacin.
`
`It is apparent from Figures 2 and 3 that the maximum
`AAQTCF, an ICH E14-recommended end point, occured for
`moxifloxacin at 4h as opposed to the pharmacokinetic rm,
`of 2h postdose;
`the mean (90% confidence interval
`(C1))
`were 13.5 (9.9,
`l7.1)ms for machine-read ECGs and 22.4
`(19.3, 27.l)ms for manual-read ECGS. For tolterodine, the
`time of maximum AAQTCF effect coincided with its
`pharmacokinetic rm“ of lh postdose, with the exception of
`manual-read ECGS at the 2mg twice-daily dose, where the
`mean (95% CI) of the maximum AAQTCF was 1.38 (-3.61,
`
`6.37) ms at 3 h postdose.
`The study sensitivity was confirmed because a mean QT
`prolongation effect
`in excess of 5ms was
`seen with
`moxjfloxacin. For each of the analyses, whether machine-
`
`CLINICAL PHARMACOLOGY 8: THERAPEUTICS | VOLUME 81 NUMBER 3 | MARCH 2007
`
`Machine-read iolierodine TOT
`
`25
`20
`
`—o—Tui1erodina 2 my BID
`—O—Toiterodine 4 mg BID
`—I—Muxillm(acln 400 rngfday
`n Moxilioxacin in otherTOT studies
`
`AAQTCF
`
`(rns) 6
`
`B
`
`'5
`
`Time postdose (h)
`
`Figure 2 PIacebo—subtracted time-—r'natched change from baseline Fridericia
`QTc values (machine-read) on day 4. Dashed line shows the 5 ms threshold
`criterion for a negative TQT study. Moxifloxacin QTc effect from other trials
`is based on manua|—read ECGs.
`
`the QTc interval effects for both
`read or manual-read,
`tolterodine doses were lower than with moxifloxacin.
`
`Categorical (outlier) analysis of QTc effects of tolterodine and
`moxifloxacin
`
`None of the QTc intervals measured in this study exceeded
`500 ms. One subject after moxifloxacin and no subject after
`tolterodine treatment had a QTc change from baseline
`>60 ms. Overall, the observed frequency of subjects with
`QTCF or population-corrected QT (QTCP) changes between
`30 and 60 ms appeared to be lower for placebo (e<_2.1"/’o for
`manual-read, 0% for machine-read analyses) and tolterodine
`
`379
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2071 - 0003
`
`
`
`ARTICLES
`
`l\JU1
`
`IO0
`
`.5 U1
`
`
`
`AAQTCF(ms) Eu‘
`
`0
`
`Machine-read tolterodine TQT
`
`
`
`-0- Tolterodine 2 mg am
`-0- Tolterodine 4 mg BID
`—I— Moxifloxacin 400 mglday
`D Moxifloxacin in oIherTOT studies
`
`
`
`12
`
`Time postdose (h)
`
`14
`
`Figure 3 P|acebo—subtracted time—matched change from baseline Fridericia
`QTc values (manual-read) on day 4. Dashed line shows the 5 ms threshold
`criterion for a negative TQT study. Moxifloxacin QTc effect from other trials
`is based on manual-read ECGs.
`
`Table 2 Steady-state, placebo-adjusted change from baseline
`in machine-read QTIQTc values at r,,,,,, following tolterodine
`and moxifloxacin administration
`
`Tolterodine
`
`Tolterodine
`
`Moxifloxacin
`
`2 mg b.i.d.a
`
`4 mg b.i.d."
`
`400 mg q.d.'°
`
`Table 3 Steady-state, placebo-adjusted change from baseline
`in manual-read QTlQTc values at t,,,,,, following tolterodine
`and moxifloxacin administration
`
`QT (uncorrected)
`Point estimate
`
`CI
`SE for difference
`
`OTcF (Fridericia)
`Point estimate
`
`CI
`SE for difference
`
`QTCB (Bazett)
`Point estimate
`
`CI
`SE for difference
`
`Tolterodine
`2 mg b.i.d.a
`
`Tolterodine
`4 mg b.i.d.a
`
`Moxifloxacin
`400 mg q.d.b
`
`1.1
`
`(-5.9, 8.0)
`3.0
`
`-0.3
`
`(-7.2, 6.6)
`3.0
`
`17.7
`
`(11.9, 23.4)
`3.5
`
`5.0
`
`(0.3, 9.7)
`2.4
`
`11.8
`
`(7.1, 16.6)
`2.4
`
`19.3
`
`(15.5, 23.0)
`2.3
`
`6.9
`
`(1.9, 12.0)
`2.6
`
`18.3
`
`19.9
`
`(13.2, 23.3)
`2.6
`
`(15.5, 24.3)
`2.7
`
`19.1
`
`OT (uncorrected)
`Point estimate
`
`-2.9
`
`-6.4
`
`7.1
`
`CI
`SE for difference
`
`(-10.5, 4.6)
`3.8
`
`(-14.0, 1.1)
`3.8
`
`(-0.0, 14.1)
`4.3
`
`QTCF (Friderfcia)
`Point estimate
`
`CI
`SE for difference
`
`OTCB (Bazett)
`Point estimate
`
`CI
`SE for difference
`
`1.2
`
`(-3.0, 5.3)
`2.1
`
`5.6
`
`(1.5, 9.8)
`2.1
`
`8.9
`
`(4.8, 13.0)
`2.5
`
`3.2
`
`(-0.6, 7.0)
`1.9
`
`11.9
`
`(8.1, 15.3)
`1.9
`
`9.6
`
`(5.5, 13.7)
`2.5
`
`OTcP (population)
`Point estimate
`
`CI
`SE for difference
`
`2.0
`
`(-1.8, 5.8)
`1.9
`
`QTcP (population)
`8.3
`
`(4.5, 12.2)
`1.9
`
`9.3
`
`(5.3, 13.2)
`2.4
`
`b.i.d., twice daily; CI, confidence interval; SE, standard error; q.d., once daily. “At t,,,,,,,
`of 1 f1; 95% Cl. “At cm of 2 h; 90% CI.
`
`treatments (€_ 16.7% for manual-read, S6.3% for machine-
`
`(€_45.7% for
`than moxifloxacin treatment
`read analyses)
`c<‘23.9"/o for machine-read analyses). In the
`manual-read,
`manual-read data set,
`there were no occurrences of QTcF
`>470 ms in women; QTc >450 ms occurred in one male
`subject after each of the three active treatments. In the
`machine-read data set, no women had QTcF >470ms;
`
`380
`
`QTCP (population)
`Point estimate
`
`CI
`SE for difference
`
`4.5
`
`(-0.4, 9.3)
`2.4
`
`10.3
`
`(5.5, 15.1)
`2.4
`
`(15.3, 22.9)
`2.3
`
`b.i.d., twice daily; CI, confidence interval; SE, standard error; q.d., once daily. ‘At t,..a,,
`of 1 h; 95% CI. “A1 rm, of 2 h; 90% CI.
`
`among men, the occurrence of QTCF >450 ms was seen in
`one (4%),
`three (12%), and one (4%) subjects receiving
`tolterodine 2 mg twice daily, tolterodine 4 mg twice daily, and
`moxifloxacin 400 mg once daily, respectively.
`The model-based QTcP increase for moxifloxacin was
`
`consistent with previously reported
`10.8 ms, which is
`prolongation values.19‘23. For tolterodine, this estimate was
`7.9ms for
`the supratherapeutic dose and 2.0 ms for the
`recommended dose. For both tolterodine and moxifloxacin,
`
`the model-predicted results agreed well with the point
`estimate from the statistical analysis of the machine—read
`data (Table 2).
`
`DISCUSSION
`
`According to the ICH E14 guidance, unless precluded by
`considerations of safety or tolerability owing to adverse
`effects, the QT interval prolongation of the study drug should
`be assessed at
`substantial multiples of the anticipated
`maximum therapeutic exposure.” Alternatively,
`if
`the
`concentrations of a drug can be increased by drug—drug or
`drug—food interactions, the QT assessment may be performed
`under conditions of maximum inhibition.” The 4 mg twice-
`daily dose of tolterodine was chosen to represent a worst-case
`scenario of supratherapeutic exposures to tolterodine, based
`on the twofold increase in tolterodine exposures in PMs
`
`VOLUME 81 NUMBER 3 | MARCH 2007 l www.nature.com)’cpt
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2071 - 0004
`
`
`
`ARTICLES
`
`when used concomitantly with potent CYP3A4 inhibitors. It
`should be noted that such a scenario would be non-
`
`compliant with the prescribing information for tolterodine,
`which recommends
`a dose reduction in such circum-
`
`stances.”"” Furthermore, the IR formulation of tolterodine
`was used in this TQT study, which gives peak concentrations
`approximately 1.5-fold higher than the more widely pre-
`scribed ER formulation.” Exposures much above those in
`PMs receiving tolterodine 4 mg twice daily would be difficult
`to tolerate owing to the dose—limiting anticholinergic side
`effects. Urinary retention was reported to be a major
`tolerability problem in healthy volunteers receiving 12.8 mg
`single and higher than 4 mg twice-daily doses.“ Therefore,
`4mg twice daily was considered as an adequate and
`appropriate supratherapeutic dose of tolterodine for
`this
`study.
`The E14 guidance recommends that all TQT trials should
`include a positive control that has mean QT/QTc interval
`prolongation of about 5 ms (i.e., an effect that is close to the
`QT/QTC effect that represents the threshold of regulatory
`concern, around 5rns).15 Based on both the machine— and
`manual-read ECG data (Tables 2 and 3), the study sensitivity
`was confirmed because a >5 ms QT prolongation effect was
`seen with moxifloxacin. It was remarkable, however, that the
`
`QT effect of moxifloxacin was approximately twofold higher
`when using the manual-read QTc intervals than the machine-
`read intervals. The manual-read QTc changes of I1'l0Xifl0X-
`acin, a widely used and well-characterized positive control,
`were considerably higher in this study than other TQT trials
`(shown in Table 4 and plotted on both Figures 2 and 3).19'23
`Across several trials with both single and multiple doses of
`moxifloxacin, the mean QTC changes ranged from 6 to 13 ms,
`compared with the >19 ms manual-read mean change
`estimated in this study (Figure 3). The corresponding 90%
`CI was (155-230) and did not include the range of mean
`values reported in other studies. The moxifloxacin QTC effect
`measured by the machine-read ECG data was 8.9 ms (Table 2)
`and more consistent with previous studies.
`The machine-read QT interval data from this study were
`deemed to be most appropriate for
`the basis of study
`conclusions. Although the exact reason(s) for the discrepancy
`in manual-read results remains unclear, several methodolo-
`
`gical
`
`issues were identified. Although this was a crossover
`
`study, the involvement of multiple readers could introduce
`subjectivity and non-standardized interval adjustment,
`in-
`creasing the overall variability of manual-read QTC values.
`The ECG tracings for the entire study were read over a
`relatively long period of 3 months, which could further add a
`within-reader component of variability. Additionally,
`the
`tangent method for determining the end of T-wave may be
`inappropriate. Combined, these logistical constraints, meth-
`odological
`issues, and a
`few operational
`inconsistencies
`(described in the Methods section) have the potential
`to
`affect the precision and accuracy of the manual-read QTC
`determinations. The potential
`for manual
`techniques to
`introduce bias is discussed further in a recent publication.“
`A broad review of ECG tracings confirmed the absence of
`T-wave morphology changes. This supported the validity of
`using machine-read QT intervals in this healthy volunteer
`study. More than 14,000 ECGs were collected during the four
`treatment periods of the study. Among the machine-read
`ECGs, only six subjects showed nonspecific ST- and/or T-
`wave abnormalities at their screening visit but not at any of
`the measurement times during the four periods. Only one
`subject exhibited T-wave abnormality in a
`single ECG
`collected on the baseline day during moxifloxacin treatment
`period. The lack of any notable T-wave abnormalities was
`also confirmed by the central ECG readers’ comments, where
`such observations were made in only a minimal number of
`ECGs (56) across more than 10,300 ECGs that were manually
`over-read.
`
`The present regulatory determination of a negative TQT
`study is one where “the largest time-matched mean difference
`between the drug and placebo (baseline subtracted) for the
`QTc interval is about S5ms, with a one-sided 95% CI that
`excludes an effect > 10 ms.’’‘5 It is recognized that drugs that
`prolong the mean QT/QTc interval by about $5 ms do not
`appear to cause TdP. Data on drugs that prolong the mean
`QT/QTC interval by >5 and <20 ms are inconclusive. In
`such cases, the clinical relevance of these modest changes may
`be best determined by careful examination of the clinical trial
`and postmarketing adverse events possibly related to QT/QTc
`interval prolongation, such as TdP, cardiac arrest, sudden
`cardiac death, and ventricular arrhythmias (e.g., ventricular
`tachycardia and ventricular fibrillation). Drugs that have an
`average QTIQTC interval prolongation of >20 ms have an
`
`Table 4 Mean (95% CI) changes in QTcF produced by moxifloxacin 400 mg in other TQT studies
`
`TQT study drug
`
`Moxifloxacin dose, regimen
`
`Population sex (age, mean, or range)
`
`Mean QTcF change (msla
`
`Tolterodine
`
`400 mg. q.d. x 4 days
`
`Men/women (37 years)
`
`Vardenafilzé
`
`A|fuzosin25
`
`So|ifenacin23
`
`400 mg, SD
`
`400 mg, so
`
`400 mg, SD
`
`Men (53 years)
`
`Men (27 years)
`
`Women (51 years)
`
`400 mg, q.d. x 5 days
`Trospiumz“
`400 mg, q.d. x 6 days
`Darifenacin”
`CI, confidence interval; q.d., once daily. ‘Results are based on manual-read ECGs.
`
`Men/women (range, 18-45 years)
`Men/women (43 years)
`
`CLINICAL PHARMACOLOGY 8: THERAPEUTICS | VOLUME 81 NUMBER 3 | MARCH 2007
`
`19.3
`
`7.7
`
`10.3
`
`11.0
`
`6.4
`11.6
`
`381
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2071 - 0005
`
`
`
`ARTICLES
`
`increased likelihood of being proarrhythmic. A basis of these
`interpretation guidelines is the survey of mean peak QTc
`prolongation by several drugs.”
`Using the machine-read QTCF data, a clinically relevant 5-
`ms prolongation of QTcF could be ruled out;
`that
`is, a
`negative TQT study conclusion could be made for both the
`recommended and the supratherapeutic doses of tolterodine.
`The manual-read data indicated that a 5-ms QTcF prolonga-
`tion could be ruled out for tolterodine at the recommended
`
`is
`It
`the supratherapeutic dose.
`daily close, but not at
`noteworthy that none of the subjects,
`irrespective of their
`metabolic profile, exceeded the clinically relevant thresholds
`of 500 ms for absolute QTc or 60 ms for change from
`baseline,
`following both doses of tolterodine. The QTc
`interval changes
`following both recommended and su-
`pratherapeutic tolterodine doses were consistently smaller
`relative to the effect seen with the therapeutic dose of
`moxifloxacin (Figures 2 and 3).
`Although the study-specified primary end point was the
`change from baseline QTcF at the pharmacokinetic rm,‘ of
`tolterodine,
`the results showed that it was also coincident
`
`with the maximum change from baseline QTcF, the ICH E14-
`specified end point.” Therefore, the same study conclusions
`can be made whether using the pre-E14 study end point or
`that specified in the contemporary E14. The actual clinical
`impact of these QTc changes is unknown, but needs to be
`considered when prescribing tolterodine to patients taking
`certain drugs known to prolong QT interval, including Class
`IA or Class III antiarrhythmic medications, or to patients
`with congenital or documented acquired QT prolongation.
`The relationship between QTc changes and the serum
`concentrations of tolterodine and its active metabolite was
`
`best described using a linear model. Based on this model, we
`can estimate the probable changes in QTc using the more
`widely prescribed ER formulation. The peak exposures of
`tolterodine and DD0l after dosing with the ER capsules are
`about 61% and 67%, respectively, compared with IR tablets,
`and thus any QTc effects of the ER capsules would also be
`expected to be proportionally smaller. The concentra-
`tion—QTc modeling results also showed an exposure-related
`higher QTc effect in PMs than in EMs. Furthermore, QTc
`interval increases in PMs treated with tolterodine 2 mg twice
`daily were comparable to those observed in EMs receiving
`4 mg twice daily.
`The current understanding of cardiac safety in relation to
`the utility of QT prolongation as a biomarker for TdP and
`sudden cardiac death is poorly defined and is an area of
`continued investigation. Affinity for HERG channels in vitro
`is commonly used as a preclinical screen for
`a drug’s
`likelihood to prolong QTc. The modest QTc changes
`reported here for tolterodine suggest that I1“.-binding affinity
`in the HERG assay may be an imperfect screening tool for
`determining the likelihood for clinical QT prolongation. If
`HERG channel results had been used as the sole criterion for
`
`continuing the development of tolterodine, it is possible that
`it would not have been tested in humans. However, other
`
`382
`
`testing with tolterodine demonstrated a lack of
`preclinical
`effect on action potential duration. This TQT study,
`confirming the lack of clinically significant prolongation, is
`in agreement with the postmarketing cardiac safety assess-
`ment of tolterodine and supports the use of the latter type of
`analysis for the detection of safety signals, given the extensive
`clinical practice experience with tolterodine since 1997 (data
`on file). Broad searches within Pfizer’s postmarketing
`database, as well as those maintained by the US Food and
`Drug Administration and World Health Organization, have
`not shown any signal of an association between tolterodine
`and QT prolongationJTdP in multiple sources searched. This
`includes the lack of any published case reports.
`These non—clinical and clinical data for
`
`tolterodine
`
`demonstrate that drugs with high HERG affinity in vitro,
`do not necessarily produce clinically relevant QT prolonga-
`tion or clinical cardiac events. Another example of a drug
`with high HERG affinity, but no cardiac safety liability,
`is
`verapamil.26 It is possible that tolterodine and verapamil have
`activities at other ion channels, most notably calcium channel
`blockade,
`that mitigate any negative effects of HERG
`blockade; however,
`the precise mechanism(s) for such an
`effect have not been identified.13'19
`
`In conclusion, this TQT study demonstrated that tolter-
`odine does not have any clinically significant effect on the
`QTc interval. Compared with the therapeutic dose of
`moxifloxacin, the QTCF changes were up to one-fourth at
`the recommended dose and up to two-thirds at
`the
`supratherapeutic dose of tolterodine. Manual-read QTc data
`were more variable and the mean QTc change was
`consistently larger than machine data; however, this did not
`affect the aforementioned conclusions. The disparity between
`tolterodine’s high affinity for HERG channels in vitro and its
`lack of clinical cardiac safety signals suggest caution in over-
`reliance on HERG assays as a sole preclinical screening tool
`for selecting drug candidates for further development.
`
`METHODS
`
`Study design. This was a positive- and placebo-controlled, multi-
`ple—dose, four—way crossover study conducted at two centers. The
`study was double-blinded with respect to tolterodine and placebo,
`with open-label moxifloxacin. This study evaluated the single-dose
`and steady—state QTc effects of the recommended (2 mg twice daily)
`and supratherapeutic (4 mg twice daily) doses of tolterodine IR and
`the positive control, moxifloxacin (400 mg once daily), each
`compared with placebo. Moxifloxacin was included as a positive
`control to confirm the sensitivity of the study to detect small QTc
`changes. Moxifloxacin is frequently used in TQT studies because it
`has a well-defined QTIQTC effect, usually about 6-12 ms.
`At the screening visit, subjects underwent physical examination,
`assessment of previous medical history,
`laboratory screen, and
`CYPZD6 genotyping. Volunteers were admitted to the clinic in the
`evening 2 days before closing on day 1 and remained in the clinic
`until
`the evening of day 4 in each period. Treatments were
`administered for 4 clays (morning close only on day 4), with a
`washout period of 25 days between periods. A follow—up visit was
`performed 5-7 days after period 4 ended. The mean elimination
`half-lives were 2.2h (EM) and 9.6h (PM) for tolterodine, 2.4h (EM)
`for DDO1, and 12h for moxifloxacin. Therefore, the dosing and
`
`VOLUME 81 NUMBER 3 | MARCH 2007 l www.nature.comlcpt
`
`Patent Owner, UCB Pharma GmbH — Exhibit 2071 - 0006
`
`
`
`ARTICLES
`
`washout periods were adequate for reaching steady state and for
`drug elimination between periods, respectively.
`
`if the U wave merges with the T wave.
`in the QT measurement
`Otherwise, the U wave is excluded from the QT measurement.
`
`Subject population. Healthy adult men and women (aged 18-55
`years) with a body mass index of 18-30 kg/m2 were eligible.
`Screening blood samples were analyzed for CYP2D6 *3, *4, *6, *7,
`*8, and ‘I0 alleles to determine their CYPZD6 genotype. Subjects
`with two or more non-functional alleles were identified as PMs. It
`
`was targeted to enroll 2(L24 PMs. All subjects provided written
`informed consent. The study was conducted in accordance with
`Title 21 of the US Code of Federal Regulations, Good Clinical
`Practice guidelines, and the Declaration of Helsinki Principles.”
`
`Sample size. To meet the primary objective, 43 completed subjects
`were more than adequate to provide 95% probability that the half-
`width [i.e., margin of error) of the 95% CI for the true mean
`difference in change from baseline in QTcF between active treat-
`ment and placebo would not exceed Sms. To assess the sensitivity
`of the study, 43 subjects were required to complete treatment
`to provide 85% power
`to detect a 5-ms difference between
`moxifloxacin and placebo in change from baseline in QTCF, based
`on a one—sided paired t—test at the 5% significance level. Thus, the
`sample size of 48 subjects was adequate, assuming a dropout rate
`of 10%.
`
`ECG acquisition. Single I2-lead ECGs were obtained at screening
`and follow-up visits. During each period, baseline and on-treatment
`ECGS were obtained in triplicate at prespecified time points, with
`the consecutive replicates about
`2 min apart. All ECGS were
`performed after the subject had rested quietly for at least 10 min,
`with no food permitted 2h before measurement. When the timing
`of an ECG coincided with blood collection, the ECG was obtained
`before the nominal time of blood collection.
`
`ECGS were obtained predose on the mornings of days I-4 in each
`period. At steady state,
`the ECGs were taken at 0 (within lh
`predose), 0.5, 1, 2, 3, 4, 6, 9, and 12h postdose on day 4. After the
`first close on Clay 1, the ECGs were taken just before and at 0.5, 1, 2,
`and 4 h after the morning dosing. Baseline ECGs were performed on
`day 0 at the same times of day as those on day 4.
`All 12-lead ECGs were recorded on GE Marquette’s MAC 1200 R
`ECG recorders (GE Medical Systems, Milwaukee, WI) and the same
`ECG machine was used for all readings on a single subject. Ten
`seconds of all 12 leads of ECG data were simultaneously collected
`from each subject based on the specifics outlined in the protocol and
`stored in the MAC 1200“ . All ECGS were digitally transferred to a
`dedicated MUSE system at the core ECG laboratory (Biomedical
`Systems, St Louis, MO).
`All machine-read measurements for each ECG were calculated
`using the GE/Marquette l2SL program and were based on a “Global
`Median" beat, a superimposition of 12 median beats (one from each
`lead). The median beat within each lead was calculated with the
`following characteristics:
`
`1. Beats of the same shape (within each lead) were combined into
`an accurate, representative cycle,
`thereby reducing the noise
`dramatically.
`2. A “primary beat” was recognized as the group of beats with the
`most ECG information (1'.e.,
`the beat most
`informative of
`normal conduction).
`3. Middle (median) voltage of primary beats was taken for each
`sample (most informative of normal conduction wit