Alfuzosin Delays Cardiac Repolarization by a Novel Mechanism
Antonio E. Lacerda, Yuri A. Kuryshev, Yuan Chen, Muthukrishnan Renganathan,1
Heather Eng,2Sanjay J. Danthi,3James W. Kramer, Tianen Yang, and Arthur M. Brown
ChanTest Corporation, Cleveland, Ohio (A.E.L., Y.A.K., Y.C., M.R., H.E., S.J.D., J.W.K., T.Y., A.M.B.); and Department of
Physiology and Biophysics, School of Medicine, Case Western Reserve University, Cleveland, Ohio (A.M.B.)
Received July 10, 2007; accepted November 5, 2007
The United States Food and Drug Administration (FDA) uses
alfuzosin as an example of a drug having QT risk in humans that
was not detected in nonclinical studies. FDA approval required
a thorough clinical QT study (TCQS) that was weakly positive at
high doses. The FDA has used the clinical/nonclinical discor-
dance as a basis for mandatory TCQS, and this requirement
has serious consequences for drug development. For this rea-
son, we re-examined whether nonclinical signals of QT risk for
alfuzosin were truly absent. Alfuzosin significantly prolonged
action potential duration (APD)60in rabbit Purkinje fibers (p ?
0.05) and QT in isolated rabbit hearts (p ? 0.05) at the clinically
relevant concentration of 300 nM. In man, the QT interval
corrected with Fridericia’s formula increased 7.7 ms, which
exceeds the 5.0-ms threshold for a positive TCQS. Effects on
hKv11.1, hKv4.3, and hKv7.1/hKCNE1 potassium currents and
calcium current were not involved. At 300 nM, ?30? Cmax,
alfuzosin significantly increased whole-cell peak sodium
(hNav1.5) current (p ? 0.05), increased the probability of late
hNav1.5 single-channel openings, and significantly shortened
the slow time constant for recovery from inactivation. Alfuzosin
also increased hNav1.5 burst duration and number of openings
per burst between 2- and 3-fold. Alfuzosin is a rare example of
a non-antiarrhythmic drug that delays cardiac repolarization not
by blocking hKv11.1 potassium current, but by increasing so-
dium current. Nonclinical studies clearly show that alfuzosin
increases plateau potential and prolongs APD and QT, consis-
tent with QT prolongation in man. The results challenge the FDA
grounds for the absolute primacy of TCQS based on the claim
of a false-negative, nonclinical study on alfuzosin.
Alfuzosin hydrochloride is marketed in the United States
(Uroxatral, alfuzosin HCl-extended release) as a treatment
for benign prostatic hyperplasia. Alfuzosin is a selective an-
tagonist of postsynaptic ?1-adrenergic receptors located in
the prostate, bladder base, bladder neck, prostate capsule,
and prostate urethra. Outside the United States, the imme-
diate-release formulation (2.5 mg three times a day) was
marketed since 1987, the sustained-release formulation (5
mg two times a day) was marketed since 1993, and the
extended-release formulation (10 mg once a day) was mar-
keted since 1999 without reports of adverse cardiac events
clearly attributable to alfuzosin (http://www.fda.gov/cder/foi/
nda/2003/021287_uroxatral_toc.htm). A United States Food
and Drug Administration (FDA) search of the World Health
Organization adverse event database included data from
22,912 patients in 194 clinical trials and found that the
most frequent adverse events were hypotension (57), syn-
cope (53), postural hypotension (42), palpitations (28), an-
gina (14), myocardial infarction (14), tachycardia (13), and
atrial fibrillation (13) (http://www.fda.gov/cder/foi/nda/2003/
21-287_Uroxatral_Admindocs_P2.pdf). The World Health
Organization data did not include the specific terms of QT
prolongation and Torsade de Pointes; however, seven cases of
arrhythmia and three cases of sudden death were listed.
In a recent evaluation of cardiovascular tolerability of the
extended-release formulation (10 mg once a day), age, car-
diovascular comorbidity, and antihypertensive co-medication
had no impact on the safety profile of alfuzosin (Hartung et
During a review of the New Drug Application (NDA), the
FDA commented that alfuzosin might increase the rate-cor-
rected QT interval (QTc) and that the clinical pharmacology
This work was supported in part by Grant HL082060 from the NHLBI,
National Institutes of Health and ChanTest Corporation.
1Current affiliation: Boehringer Ingelheim Pharmaceuticals, Ridgefield,
2Current affiliation: Pfizer Global Research and Development, Groton,
3Current affiliation: Genzyme Corporation, Waltham, Massachusetts.
Article, publication date, and citation information can be found at
ABBREVIATIONS: FDA, United States Food and Drug Administration; NDA, New Drug Application; QTc, rate-corrected QT interval; TCQS,
thorough clinical QT study; AP, action potential; ICa,L, cardiac myocyte L-type calcium channel current; DMSO, dimethyl sulfoxide; HB-PS,
HEPES-buffered physiological solution; PFT, Purkinje Fiber Tyrode; BCL, basic cycle length; APD, action potential duration; KH, Krebs-Henseleit;
LQT3, long QT syndrome 3; PF, Purkinje fiber; HEK, human embryonic kidney.
THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS
Copyright © 2008 by The American Society for Pharmacology and Experimental Therapeutics
JPET 324:427–433, 2008
Vol. 324, No. 2
Printed in U.S.A.
by guest on February 19, 2013
data submitted in the NDA were insufficient to evaluate
the QTc risk (http://www.fda.gov/cder/foi/nda/2003/21-287_
Uroxatral_Admindocs_P3.pdf). To better evaluate the car-
diac risk, the FDA required a thorough clinical QT study
(TCQS). In discussion summaries, the FDA stated that in
vitro data would not be helpful in determining the QT risk
and that postmarketing surveillance (phase IV studies)
would be inadequate (http://www.fda.gov/cder/foi/nda/2003/
21-287_Uroxatral_Admindocs_P3.pdf). Sanofi-Synthelabo (now
known as sanofi-aventis, Bridgewater, NJ) performed the
TCQS that resulted in the inclusion of summary QTc data in
the prescribing information for alfuzosin (http://products.
sanofi-aventis.us/uroxatral/uroxatral.html). The outcome of
the TCQS was that, at four times the therapeutic dose (40
mg; chosen to mimic increased alfuzosin plasma levels in the
presence of a potent inhibitor of liver metabolism of alfuzosin,
like ketoconazole) (http://products. sanofi-aventis.us/uroxatral/
uroxatral.html), alfuzosin prolonged QTc, although the ex-
tent of the prolongation depended on the method of correc-
tion. For TCQS, the threshold of positivity for mean QTc was
5 ms, and the upper bound for the 95% confidence interval
was less than 10 ms. For alfuzosin, the mean corrected QT
interval by Fridericia’s formula was 7.7 ms, and the upper
bound of the 95% confidence interval was 13.5 ms (http://
products.sanofi-aventis.us/uroxatral/uroxatral.html). The posi-
was contrasted with the negative nonclinical data included in
the NDA submission. These contained an IC50value of 83 ?M
for Kv11.1 block, a piglet cardiac Purkinje fiber (PF) action
potential (AP) assay demonstrating no effect of 1 ?M alfuzo-
sin and a large safety margin (Table 1). In this study, we
show that alfuzosin does, in fact, produce positive nonclinical
signals for repolarization risk and provide a mechanism for
the positive TCQS. We claim that there is no absolute pri-
macy for TCQS and that both nonclinical and clinical assays
are necessary for QT risk assessment.
Materials and Methods
Two adult Hartley guinea pigs and 11 New Zealand white rabbits
of either gender were maintained in the Association for Assessment
and Accreditation of Laboratory Animal Care-accredited facilities at
Case Western Reserve University (Cleveland, OH) and Northeastern
Ohio Universities College of Medicine (Rootstown, OH), respectively.
All procedures for harvesting tissues were approved by the Institu-
tional Animal Care and Use Committee of each facility.
Methods for voltage clamp of cells, solution preparation, drug
application, data acquisition, and analysis hardware and software
were essentially performed as described previously (Kirsch et al.,
2004). HEK293 cells were stably transfected with hKv11.1 (hERG,
KCNH2), hKv4.3, hKv7.1/hKCNE1 (hKvLQT1/hminK), or hNav1.5
cDNA (Kuryshev et al., 2001; Lacerda et al., 2001). Guinea pig
cardiac myocyte L-type calcium channel currents (ICa,L) were re-
corded from acutely isolated, enzymatically dispersed guinea pig
ventricular myocytes (Kuryshev et al., 2005), which were prepared
from adult animals of either gender. Unless otherwise indicated,
chemicals were obtained from Sigma-Aldrich (St. Louis, MO). Alfuzo-
sin was obtained from SynFine Research (Richmond Hill, ON, Can-
ada). Single-channel data were analyzed with Clampfit 9.2 (Molec-
ular Devices, Sunnyvale, CA). Experiments were performed at room
Patch-Clamp Solutions. With the exception of hNav1.5 and ICa,L
recordings, alfuzosin concentrations were prepared by diluting di-
methyl sulfoxide (DMSO) stock solutions into a HEPES-buffered
physiological solution (HB-PS) (Kirsch et al., 2004). The pipette
solution for whole-cell recordings of hKv11.1 and hKv4.3 potassium
channels was prepared as described previously (Kirsch et al., 2004).
For hNav1.5-channel recordings, alfuzosin concentrations were pre-
pared by diluting DMSO stock solutions into the following low-
sodium HB-PS solution (sodium channel-isolating physiological sa-
line): 40 mM NaCl, 97 mM L-aspartic acid, 4.0 mM KCl, 1.8 mM
CaCl2, 1 mM MgCl2, 5 mM HEPES, and 10 mM glucose, and pH was
adjusted to 7.4 with N-methyl-D-glucamine. The intracellular pipette
solution for hNav1.5 measurements was as follows: 130 mM cesium
aspartate, 5 mM MgCl2, 5 mM EGTA, 2 mM Na2ATP, 0.1 mM GTP,
and 10 mM HEPES, and pH was adjusted to 7.2 with CsOH. Guinea
pig myocytes were superfused with a calcium channel-isolating ex-
ternal solution of the following composition (calcium channel-isolat-
ing physiological saline): 137 mM NaCl, 5.4 mM CsCl, 1.8 mM CaCl2,
1 mM MgCl2, 10 mM HEPES, and 10 mM glucose, and pH was
adjusted to 7.4 with NaOH. The internal solution for ICa,Lmeasure-
ments was designed to prevent outward currents through sodium
and potassium channels, and it had the following composition: 130
mM cesium methanesulfonate, 20 mM tetraethylammonium chlo-
ride, 1 mM MgCl2, 10 mM EGTA, and 10 mM HEPES, and pH was
adjusted to 7.2 with methanesulfonic acid. Internal solutions for ICa,L
and hKv7.1/hKCNE1 measurements were supplemented with ATP,
GTP, phosphocreatine, and creatine phosphokinase to prevent run-
down of calcium channels (Bean, 1992). The pipette solutions were
prepared in batches, aliquoted, and stored frozen at ?20°C (?80°C
for solutions containing enzyme), and they were freshly thawed each
day of use. Measurements were made at room temperature.
Kv11.1 Assay. HEK293 cells stably expressing hKv11.1 were held
at ?80 mV. Onset and steady-state block of hKv11.1 current due to
alfuzosin was measured using a pulse pattern with fixed amplitudes
(depolarization, ?20 mV for 2 s; repolarization, ?50 mV for 2 s)
repeated at 10-s intervals. Peak tail current was measured during
the 2-s step to ?50 mV.
Nav1.5 Assay. For concentration-response measurements, cells
stably expressing hNav1.5 were held at ?80 mV. Onset and steady-
state activation of hNav1.5 current due to alfuzosin was measured
using a pulse pattern with fixed amplitudes (hyperpolarization,
?120 mV for 200 ms; depolarization, ?15 mV for 10 ms) repeated at
10-s intervals. Peak inward current was measured during the step to
?15 mV. The protocol for measurements of recovery from inactiva-
tion is described in the legend to Fig. 6.
Alfuzosin concentrations measured in human plasma and hKv11.1 safety ratio
Safety Ratio (Kv11.1 IC50)/(CmaxFree)
aAlfuzosin is 82 to 90% protein bound in plasma.
bA single 10-mg dose of Cmaxis 13.6 ng/ml, and the repeat oral dosing is 1.2 to 1.6 higher. Taking the upper limit (1.6) makes the Cmaxfor repeat oral dosing 21.8 ng/ml.
Lacerda et al.
by guest on February 19, 2013
Kv4.3 Assay. HEK293 cells stably expressing hKv4.3 were held at
?80 mV. Onset and steady-state block of hKv4.3 current due to
alfuzosin was measured using a pulse pattern with fixed amplitudes
(depolarization, ?20 mV for 300 ms) repeated at 15-s intervals. Peak
current and current after 70 ms were measured during the step to
Kv7.1/KCNE1 Assay. Cells stably expressing hKv7.1/hKCNE1
were held at ?80 mV. Onset and steady-state block of hKv7.1/
hKCNE1 current was measured using a pulse pattern with fixed
amplitudes (depolarization, ?20 mV for 2 s; repolarization, ?40 mV
for 0.5 s) repeated at 15-s intervals. Current was measured at the
peak of current at ?40 mV.
Calcium Current Assay. Cells were held at ?40 mV to inacti-
vate sodium channels. Onset and steady-state block of ICa,Lcurrent
due to alfuzosin was measured using a pulse pattern with fixed
amplitudes (depolarization, 0 mV for 300 ms) repeated at 20-s inter-
vals. Peak current was measured during the step to 0 mV.
Purkinje Fiber AP Assay
Purkinje Fiber Preparation. Purkinje fibers were excised from
adult rabbit ventricles by a procedure adapted from the standard
methods for canine Purkinje fiber dissection (Gintant et al., 2001).
On the day of testing, a rabbit was heparinized (sodium heparin,
1000 U/kg i.v.) and anesthetized (i.v.) with ketamine (37 mg/kg),
xylazine (2.5 mg/kg), and acepromazine (1 mg/kg). Purkinje fibers
were mounted in a Plexiglas chamber (approximate volume, 2 ml),
affixed to a heated platform, and superfused at approximately 4
ml/min with standard Purkinje Fiber Tyrode (PFT) solution warmed
to 37 ? 1°C. Alfuzosin concentrations were prepared by diluting
DMSO stock solutions into the following standard PFT solution: 131
mM NaCl, 4.0 mM KCl, 2.0 mM CaCl2, 0.5 mM MgCl2, 18.0 mM
NaHCO3, 1.8 mM NaH2PO4, and 5.5 mM glucose. Before use, the
PFT solution was aerated with a mixture of 95% O2and 5% CO2(pH
7.2 at room temperature) and warmed to 37°C.
Purkinje Fiber Electrophysiology Protocols. Membrane po-
tentials were recorded using conventional intracellular microelec-
trodes filled with 3 M KCl solution connected to an electrometer
amplifier (IE 210; Warner Instruments, Hamden, CT). Action poten-
tials were evoked by repetitive electrical stimuli (0.1–3-ms duration,
approximately 1.5? threshold amplitude) generated by a photo-iso-
lated stimulator (S-900; Dagan Corporation, Minneapolis, MN). An-
alog signals were low-pass filtered at 20 kHz before digitization at 50
kHz (DT3010; Data Translation, Inc., Marlboro, MA) and stored on a
hard disk using a PC-compatible computer controlled by Notocord-
Hem 3.5 software (Notocord Systems SA, Croissy-sur-Seine, France).
Purkinje fibers were continuously paced at a basic cycle length
(BCL) of 1 s during a stabilization period of at least 25 min before
obtaining control AP responses. Acceptable fibers were continuously
stimulated at a BCL of 1 s for 20 min. At the end of this period,
baseline action potential duration (APD) rate dependence under
control conditions was measured using stimulus pulse trains consist-
ing of approximately 50 pulses at a BCL of 1 and 0.5 s. After
returning to a BCL of 1 s, a test article at the lowest concentration
was applied for 20 min to allow equilibration, and the stimulus
trains were repeated. The entire sequence (20 min of equilibration
followed by two cycles of stimulus trains at decreasing BCL) was
repeated at increased alfuzosin concentrations. The average re-
sponses from the last five recorded action potentials from each stim-
ulus train were analyzed for each test condition.
QT in Isolated, Perfused Rabbit Heart
Alfuzosin concentrations were prepared daily by diluting stock
solutions in Krebs-Henseleit (KH) solution: 129 mM NaCl, 3.7 mM
KCl, 1.3 mM CaCl2, 0.64 mM MgSO4, 2.0 mM sodium pyruvate, 17.8
mM NaHCO3, and 5 mM glucose. The solution was aerated with a
mixture of 95% O2and 5% CO2, pH 7.3 to 7.45. On the day of testing,
a rabbit was heparinized (sodium heparin, 500 U/kg i.v.) and anes-
thetized with sodium pentothal (50 mg/kg i.v.). The heart was rap-
idly removed via a midsternal thoracotomy and placed in chilled
oxygenated (95% O2? 5% CO2) KH solution. The heart was mounted
in a Langendorff heart perfusion apparatus and perfused with KH
solution (37°C) in a retrograde fashion through the aorta at a con-
stant pressure. A polyethylene drainage tube was inserted into the
left ventricle via the left atrial appendage to diminish the metabolic
effects of myocardial work on the coronary vasomotor tone and to
continuously assess the competency of the aortic valves. The A-V
node was ablated to slow the intrinsic heart rate to a ventricular
escape less than 60 beats/min. The heart was immersed in a ther-
mostatically controlled bath into which the heart’s effluent drained.
The heart was rejected if visual inspection revealed signs of clotting
or if the coronary flow was less than 20 ml/min.
Stimulus Generation and Data Acquisition. The heart was
paced by repetitive electrical stimuli (0.1–5-ms duration, approx-
imately 1.5? threshold amplitude). A bipolar, insulated (except at
the tip) electrode was used to deliver pulses generated by a stim-
ulator (Master 8; A.M.P.I., Jerusalem, Israel). ECG signals were
conditioned by an AC-coupled preamplifier (Grass Model P511;
Astro-Med, Inc., West Warwick, RI) with low-pass filtering to
achieve a bandwidth of 10 to 300 Hz. Analog signals were digitized
at 1 kHz and stored on a hard disk using a PC-compatible com-
puter controlled by LabView software (version 7.1; National In-
struments, Austin, TX).
Test Procedures. The concentration-response relationship was
determined by continuous pacing at a basic cycle length of either 0.5
or 1 s throughout the experimental period. A stabilization period of
at least 30-min duration was obtained before the start of measure-
ments. Baseline control data were acquired at the end of the vehicle
control period at BCLs of 1 and 0.5 s. Alfuzosin concentrations were
sequentially applied in ascending order for exposure periods of at
least 15 min/concentration to allow equilibration with the tissue. The
average responses (at BCLs of 1 and 0.5 s) from four to five hearts
were analyzed for each alfuzosin concentration.
Data Analysis. The QT interval was determined using the tan-
gent method (Surawicz and Knoebel, 1984; Sides, 2002; Al-Khatib et
al., 2003). QRS duration and P-P intervals were measured from the
ECG tracings. The mean ? S.E.M. values were calculated from the
last four beats in each equilibration period.
Data were reported as the mean ? S.E.M. and were calculated
with Microsoft Excel 2003 (Microsoft, Redmond, WA). Statistical
analyses were performed with SAS JMP 5.0.1 software (SAS Insti-
tute, Cary, NC). Changes in paired data were evaluated for statisti-
cal significance using a two-tailed Student’s t test for paired samples
with significance at p ? 0.05.
Kv11.1 Concentration Response to Alfuzosin. hKv11.1
current (Fig. 1) was used as a surrogate for IKr, the rapidly
repolarizing cardiac potassium current. The concentration-
response relationship for block of hKv11.1 peak tail currents
was fit with an IC50value of 13.9 ?M. The safety margin for
hKv11.1 repolarization risk was greater than 1000 (Table 1).
Prolongation of Rabbit Purkinje Fiber Action Poten-
tial. Alfuzosin produced a concentration-dependent increase
in rabbit Purkinje fiber APD parameters (Fig. 2) measured at
a basic cycle length of 2 s. The effect was greater on APD60
than APD90, and late AP plateau potential was more positive
(Table 2; Fig. 2). The increase of APD was larger at slower
rates and showed reverse-use dependence. Prolongation of
APD was statistically significant (p ? 0.05 using the Dun-
nett’s test). Mean time differences from control at a BCL of
Alfuzosin Nonclinical Assays
by guest on February 19, 2013
1 s corresponding to statistically significant increases of
APD60in Table 2 for alfuzosin at 0.3, 1, and 10 ?M were
10.9 ? 3.3, 17.0 ? 4.8, and 41.7 ? 10.2 ms, respectively. For
APD90, corresponding time differences for alfuzosin at 1 and
10 ?M were 15.1 ? 4.4 and 47.2 ? 10.2 ms, respectively. At
a BCL of 0.5 s, corresponding APD60and APD90mean time
differences for alfuzosin at 10 ?M were 22.7 ? 7.1 and 26.4 ?
5.5 ms, respectively.
Prolongation of QT in Isolated Rabbit Heart. The
effects of alfuzosin on the QT interval measured in a paced
Langendorff-perfused rabbit heart model (CT-QT) showed a
concentration-dependent increase of QT, similar to the effect
of alfuzosin on rabbit Purkinje fiber APD (Fig. 3). The effect
on QT was detected at 300 nM concentration and was statis-
tically significant (p ? 0.05).
Effects on hKv4.3, hKv7.1/hKCNE1, and ICa,LCur-
rents. hKv4.3 current was used as a surrogate for Ito, the
transient outward potassium current. Alfuzosin had no sig-
nificant effect on hKv4.3 currents at 10 ?M concentration
(increased current by 6.1 ? 2.5%, n ? 2; data not shown).
hKv7.1/hKCNE1 current was used as a surrogate for IKs, the
slowly activating component of the cardiac-delayed rectifier
potassium current, and alfuzosin had no significant effect on
this current (data not shown). Alfuzosin had no significant
effect on ICa,Lcurrents at 10 ?M concentration (data not
Effects on Cardiac Sodium Currents. Because alfuzo-
sin had no effect on hKv11.1, hKv4.3, or hKv7.1/hKCNE1
potassium currents and no effect on L-type calcium currents,
we hypothesized that it might increase sodium current. In
Fig. 2. Alfuzosin delays cardiac repolarization. Alfuzosin
was sequentially applied at increasing concentrations to a
rabbit Purkinje fiber. Superimposed records before (con-
trol) and after equilibration with 0.3, 1, and 10 ?M alfuzo-
sin. Temperature ? 37 ? 1°C, BCL ? 2 s. Alfuzosin pro-
longed APD and increased plateau potential at all of the
tested concentrations at this cycle length.
Fig. 1. Block of hKv11.1 current by alfuzosin. Sample of hKv11.1 current records [current (pA); time (s)] during application of control and alfuzosin
solutions at the indicated concentrations. A steady state was maintained for at least 30 s before and after application of each alfuzosin concentration.
Peak tail current was measured until a new steady state was achieved. Voltage wave form [voltage (mV); time (s)] stimulus used to activate hKv11.1
currents is shown in the bottom panel. Cells stably expressing hKv11.1 were held at ?80 mV. Onset and steady-state block of hKv11.1 current due
to alfuzosin was measured using a pulse pattern with fixed amplitudes (depolarization, ?20 mV for 2 s; repolarization, ?50 mV for 2 s) repeated at
10-s intervals. Peak tail current was measured during the 2-s step to ?50 mV. Records were not corrected for capacity transient or leak current.
Lacerda et al.
by guest on February 19, 2013
whole-cell currents, alfuzosin at 300 nM and higher produced
a statistically significant (p ? 0.05) increase (330 ? 13 pA/
9 ? 5% at 0.3 ?M, n ? 2 and 269 ? 68 pA/11 ? 2% at 10 ?M,
n ? 3) of peak current amplitude relative to control. Alfuzo-
sin decreased the time to peak. We hypothesized that alfuzo-
sin could modulate sodium channel gating in a manner sim-
ilar to the hereditary long QT syndrome 3 (LQT3). The LQT3
arises from a low frequency of atypical sodium channel open-
ings characterized by the failure of inactivation to close the
channel after it is opened, thereby increasing inward current
during the AP plateau and delaying repolarization (Bennett
et al., 1995; Dumaine et al., 1996). The increase of late
openings seen in LQT3 gating was distinct in unitary sodium
current recordings (Dumaine et al., 1996), and we performed
similar studies. Figure 4 shows superimposed records with
pronounced bursting activity after application of alfuzosin.
Figure 5 shows averaged currents with an increase relative
to control of early peak current and late current (Table 3).
Alfuzosin at 10 ?M increased the number of openings in a
burst from 3 ? 1 to 7 ? 6 ms (n ? 2) and burst duration from
6.3 ? 4.2 to 18.3 ? 16.8 ms (n ? 2). Alfuzosin at 300 nM
increased late-opening NPoby 52 ? 10% (n ? 3), and the
increase was statistically significant (p ? 0.05) (Table 3).
Macropatch average peak currents in alfuzosin decayed
faster, crossed over control current, and had a shorter time to
peak (Fig. 5). The average increase of macropatch peak so-
dium current after exposure to 300 nM alfuzosin was 55 ?
5% (n ? 6) and was statistically significant (p ? 0.05).
Because alfuzosin altered inactivation gating of single-
sodium channels, we looked at recovery from inactivation in
whole-cell currents. Consistent with an increase in late chan-
nel openings, alfuzosin increased the rate of recovery from
inactivation measured with a double-pulse procedure (Wan
et al., 2001). Alfuzosin at 300 nM concentration significantly
shortened the slow time constant for recovery from inactiva-
tion, increasing the number of channels available for activa-
tion at late times during the voltage step (Fig. 6). This result
was consistent with the increase of channel openings seen in
the overlaid single-channel records and averaged late cur-
rents and with the increased late AP-plateau potential and
APD prolongation (Fig. 2). Alfuzosin only had small effects on
activation and inactivation gating of sodium channels (data
Unlike the conclusion of the FDA and the drug’s sponsor,
our results showed that alfuzosin tests were positive for
delayed repolarization in nonclinical studies. We found sig-
nificant prolongation of APD in rabbit Purkinje fibers and
significant QT prolongation in Langendorff-perfused rabbit
hearts at basic cycle lengths of 1 and 0.5 s. The piglet Pur-
kinje fiber measurements included in the NDA submission
did show prolongation of APD at a basic cycle length of 4 s,
which was reversed to APD shortening at a basic cycle length
of 1 s. This result was interpreted by Sanofi-Synthelabo as a
reverse use-dependent effect of alfuzosin. Our data from rab-
bit Purkinje fibers showed reverse-use dependence with basic
cycle lengths of 1 and 0.5 s. The rabbit PF preparation more
accurately predicts the QT risk of non-antiarrhythmic drugs
than the canine PF preparation (Lu et al., 2001), and Pur-
kinje fibers from these two species have been tested much
more extensively than piglet PFs. The QT prolongation seen
in the paced, isolated Langendorff-perfused rabbit heart was
consistent with the Purkinje fiber results and showed a de-
tectable effect at 300 nM.
The prolongation of APD and QT was not due to block of
hKv11.1 current, which usually accounts for 95% plus of
hKv7.1/hKCNE1, and L-type calcium currents were not af-
fected. Previous experiments showed that alfuzosin does not
interrupt trafficking of hKv11.1 protein to the cell surface
(Wible et al., 2005).
Our results showed that alfuzosin delays cardiac repolar-
Fig. 3. Alfuzosin prolongs QT. QT prolongation by alfuzosin was concen-
tration-dependent in Langendorff-perfused rabbit hearts paced at a basic
cycle length of 1 s (n ? 3–5). The AV node was crushed to suppress sinus
rate. The percentage change of QT was measured relative to control in
each heart. QT increased significantly (?, p ? 0.05) at all concentrations
(Student’s paired t test).
Alfuzosin effects on rabbit cardiac AP parameters at 0.3, 1, and 10 ?M
6.1 ? 2.8
7.8 ? 2.2*
6.0 ? 4.2
11.5 ? 3.2*
15.6 ? 4.7*
28.3 ? 6.2*
3.2 ? 1.9
3.9 ? 1.4
4.3 ? 2.8
7.8 ? 2.3*
14.3 ? 2.8*
24.7 ? 5.1*
?1.4 ? 0.8
?1.2 ? 0.7
?0.2 ? 0.8
?0.1 ? 0.7
1.1 ? 2.3
0.9 ? 1.9
1.4 ? 1.6
2.4 ? 0.8
?2.4 ? 3.5
?1.6 ? 3.3
?3.9 ? 3.0
?2.3 ? 2.7
5.2 ? 7.7
5.5 ? 2.3
?0.4 ? 10.7
4.2 ? 12.0
?11.4 ? 109
?1.9 ? 9.5
* Statistically significant (p ? 0.05) Student’s t test for paired data (n ? 3–5). Alfuzosin-dependent changes in AP parameters were calculated as the difference from control
in each fiber and expressed in the measurement units or as a percentage of change.
aRMP, resting membrane potential; APA, action potential amplitude; Vmax, action potential upstroke velocity.
Alfuzosin Nonclinical Assays
by guest on February 19, 2013
ization due to an agonist effect on sodium current. Alfuzosin
is a rare example of a drug that increases QT interval in this
manner. Alfuzosin significantly increased peak sodium cur-
rent. Alfuzosin also reduced the time to peak current (p ?
0.05). We interpreted this to indicate that alfuzosin reduced
the latency to first opening of sodium channels. A reduced
latency to first opening could result in more overlapping
channel openings earlier, increasing peak current but also
increasing the peak current decay that we observed due to
fewer first openings after the peak. These effects explain the
rabbit Purkinje fiber data that showed an increase in APD60
at 300 nM alfuzosin. Our single-channel data and recovery
from inactivation data showed that alfuzosin increases the
Fig. 5. Alfuzosin increases early and late sodium current. The main
figure shows the average of control (47 current records; black lines) and
10 ?M alfuzosin (106 current records; red lines) responses of hNav1.5
channels from the cell-attached patch shown in Fig. 4 and obtained with
the voltage protocol shown and described in Fig. 4. Alfuzosin increased
the early mean peak sodium current (main figure) and the late mean
sodium current (inset) relative to control. The inset shows currents from
the main figure at a higher vertical resolution. Mean current records
were corrected for leak and capacity current transients. Currents shown
were evoked by the stimulus wave-form components within the dashed-
line bounding rectangle of the inset shown in Fig. 4.
Fig. 4. Alfuzosin increases opening probability of late car-
diac sodium channels. Unitary currents are recorded from
an HEK293 cell stably expressing hNav1.5. Top, 47 super-
imposed control records; bottom, 50 superimposed records
after equilibration with 10 ?M alfuzosin. Stimulation fre-
quency is 0.4 Hz (2.5-s interval). Solutions used for record-
ing were high potassium-depolarizing solution in bath and
in the inset is as follows: holding potential, ?80 mV; condi-
tioning voltage steps, ?120 mV for 200 ms; second voltage
step, ?80 mV for 2.5 ms; test voltage step to ?40 mV for
125 ms; and return to ?80 mV (holding potential). Single-
channel currents are displayed during the portion of the volt-
age protocol within the dashed-line bounding rectangle.
Alfuzosin stimulation of late sodium current
Unitary currents were recorded from HEK293 cells stably expressing hNav1.5 with
the cell-attached configuration of the patch-clamp method. Patches contained mul-
tiple sodium channels, and openings were evoked with the voltage protocol described
in the legend to Fig. 4. NPofor late openings was measured over the last 65 ms of a
125-ms step to ?40 mV. Alfuzosin was applied to the bath at 300 nM and 10 ?M
concentrations and increased late openings. The mean increase at 300 nM was 52 ?
10% (n ? 3) and was statistically significant (p ? 0.05).
CellAlfuzosin Control NPo
Lacerda et al.
by guest on February 19, 2013
availability of sodium channels at long times similar to the Download full-text
duration of the cardiac action potential and the QT interval.
The resulting increase of late inward current was consistent
with the increase of late AP plateau potential, APD, and QT.
Our experiments with alfuzosin showed that thorough non-
clinical testing of cardiac depolarization and repolarization
abrogated the false-negative result of concern to the FDA. By
extension, we argue that nonclinical QT risk will be observed
for other unnamed drugs that the FDA refers to in support of
mandatory TCQS if appropriate, accurate, nonclinical tests
were done. In conclusion, what is required is standardization
of the nonclinical assays recommended in the International
Conference on Harmonization of Technical Requirements for
Registration at Pharmaceuticals for Human Use (ICH) S7B
(www.ICH.org) guidance similar to the standardization for
clinical QT measurements in the ICH E14 guidance.
Clinical QT is a weak surrogate for torsade de pointes (Sides,
2002). It is likely that the FDA threshold values for positive
thorough clinical QT studies will result in far too many false-
positives and seriously impair drug development and approval.
QT prolongation risk, the FDA is premature in assigning pri-
macy to the TCQS. A more balanced approach that takes into
account both nonclinical and clinical studies (namely, the Eu-
ropean and Japanese regulatory bodies) seems to be more ap-
propriate at this time.
We thank Sharon Beringuel for preparation of cell cultures, Bryce
Waldal for sodium channel measurements, and Lisa Dewey for per-
forming Purkinje fiber measurements.
Al-Khatib SM, LaPointe NM, Kramer JM, and Califf RM (2003) What clinicians
should know about the QT interval. JAMA 289:2120–2127.
Bean BP (1992) Whole-cell recording of calcium channel currents. Methods Enzymol
Bennett PB, Yazawa K, Makita N, and George AL Jr (1995) Molecular mechanism
for an inherited cardiac arrhythmia. Nature 376:683–685.
Dumaine R, Wang Q, Keating MT, Hartmann HA, Schwartz PJ, Brown AM, and
Kirsch GE (1996) Multiple mechanisms of Na?channel-linked long-QT syndrome.
Circ Res 78:916–924.
Gintant GA, Limberis JT, McDermott JS, Wegner CD, and Cox BF (2001) The canine
Purkinje fiber: an in vitro model system for acquired long QT syndrome and
drug-induced arrhythmogenesis. J Cardiovasc Pharmacol 37:607–618.
Hartung R, Matzkin H, Alcaraz A, Emberton M, Harving N, van Moorselaar J,
Elhilali M, and Vallancien G (2006) Age, comorbidity and hypertensive co-
medication do not affect cardiovascular tolerability of 10 mg alfuzosin once daily.
J Urol 175:624–628.
Kirsch GE, Trepakova ES, Brimecombe JC, Sidach SS, Erickson HD, Kochan MC,
Shyjka LM, Lacerda AE, and Brown AM (2004) Variability in the measurement of
HERG potassium channel inhibition: effects of temperature and stimulus pattern.
J Pharmacol Toxicol Methods 50:93–101.
Kuryshev YA, Ficker E, Wang L, Hawryluk P, Dennis AT, Wible BA, Brown AM,
Kang J, Chen XL, Sawamura K, et al. (2005) Pentamidine-induced long
QT syndrome and block of HERG trafficking. J Pharmacol Exp Ther 312:316–
Kuryshev YA, Wible BA, Gudz TI, Ramirez AN, and Brown AM (2001) KChAP/
Kvbeta1.2 interactions and their effects on cardiac Kv channel expression. Am J
Physiol Cell Physiol 281:C290–C299.
Lacerda AE, Kramer J, Shen K-Z, Thomas D, and Brown AM (2001) Comparison of
block among cloned cardiac potassium channels by non-antiarrhythmic drugs. Eur
Heart J Suppl 3:K23–K30.
Lu HR, Marien R, Saels A, and De Clerck F (2001) Species plays an important role
in drug-induced prolongation of action potential duration and early afterdepolar-
izations in isolated Purkinje fibers. J Cardiovasc Electrophysiol 12:93–102.
Sides GD (2002) QT interval prolongation as a biomarker for torsades de pointes and
sudden death in drug development. Dis Markers 18:57–62.
Surawicz B and Knoebel SB (1984) Long QT: good, bad or indifferent? J Am Coll
Wan X, Chen S, Sadeghpour A, Wang Q, and Kirsch GE (2001) Accelerated inacti-
vation in a mutant Na?channel associated with idiopathic ventricular fibrillation.
Am J Physiol Heart Circ Physiol 280:H354–H360.
Wible BA, Hawryluk P, Ficker E, Kuryshev YA, Kirsch G, and Brown AM (2005)
HERG-Lite: a novel comprehensive high-throughput screen for drug-induced
HERG risk. J Pharmacol Toxicol Methods 52:136–145.
Address correspondence to: Dr. Antonio E. Lacerda, ChanTest Corpora-
tion,14656 NeoParkway, Cleveland,
Fig. 6. Alfuzosin shortens recovery from inactivation. The
main figure shows normalized recovery from inactivation
data (n ? 6) in control (filled square symbols) and after
exposure to 300 nM alfuzosin (filled diamond symbols). The
control and 300 nM alfuzosin data were fit with a sum of
two exponential functions (solid curved lines through the
data). Fit parameter values in control were as follows: fast
amplitude, 0.62; fast tau, 3.3 ms; slow amplitude, 0.38; and
slow tau, 218 ms. Fit parameter values in 300 nM alfuzosin
were as follows: fast amplitude, 0.70; fast tau, 4.4 ms; slow
amplitude, 0.30; slow tau, 166 ms. The alfuzosin-mediated
faster recovery from inactivation reflected in the smaller
slow-time constant value was statistically significant at
p ? 0.05. The inset shows the voltage protocol used to
obtain recovery data. The holding potential was ?80 mV,
the conditioning pulse potential was ?15 mV for 500 ms,
and a variable duration step to ?120 mV was followed by a
10-ms step to ?15 mV to measure recovery of initial cur-
rent during the 500-ms-conditioning step. Pulse sequences
were repeated at 8-s intervals.
Alfuzosin Nonclinical Assays
by guest on February 19, 2013