Transcript Tolerability and safety: Predictability of QT prolongation

Tolerability and safety:
Predictability of QT prolongation
in healthy subjects from
preclinical experiments
Henri Caplain
2nd Joint Annual Meeting 2007
Bad Homburg v.d.H., Germany
1
Sanofi-Aventis (SA) Evaluation
of Cardiac Repolarization Data
for 15 Drugs
2
SCOPE OF EVALUATION
• 16 completed TES:
– 2003-2007;
– 15 compounds.
• Degree of concurrence between
preclinical and clinical data:
– Comparable methodology for preclinical
and clinical assessments.
3
Regulatory Environment
• Guidance ICH S7B: Nonclinical Evaluation of
the Potential for Delayed Ventricular
Repolarization (QT Interval Prolongation) by
Human Pharmaceuticals;
• Guidance ICH E14: Clinical Evaluation of
QT/QTc Interval Prolongation and
Proarrhythmic Potential for NonAntiarrhythmic Drugs.
4
ICH S7B
• One preclinical assay not sufficient;
• In vitro and in vivo complementary;
• Potential to delay ventricular repolarization,
related to the extent to concentration (parent +
metabolites);
• Estimate human risk;
• Integrated risk assessment.
5
ICHS7B: NonClinical Testing Strategy
Chemical/
Pharmacological
Class
Ion
In channel
vitro IKr
assay
assay
Relevant Nonclinical and Clinical
Information
In vivo QT
assay
*
Follow-up
-up studies if
Studies
Integrated
Risk
Integrated
risk assessment
necessary
Assessment
Signal of risk
Evidence
of Risk
None
Weak
Quantitative SCORING?
Strong
Impact on Clinical
Investigations
6
ICH E14: Scope
• Recommendations for clinical studies to
assess the potential of a drug to delay
cardiac repolarization;
• Determine whether a drug has a threshold
pharmacological effect on cardiac
repolarization, as detected by QT/QTc
prolongation;
7
ICH E14: Positive Study
• Prolongation QT/QTc:
– Central tendency analysis:
• 5 msec maximal mean ↑ QTc (no TdP to date);
• Difference of 10 msec on UB 95% confidence
interval around the mean effect on QTc.
– Categorial analysis:
• Absolute QTc prolongation: > 500 msec;
• Change from baseline in QTc: > 60 msec.
• Documented arrhythmias;
8
ICH E14: Potential Impact
• Determine whether or not the effect of a
drug on the QT/QTc in target patient
populations should be studied intensively
during later stages of development;
• Basis for nonapproval of a drug or
discontinuation of its clinical development.
9
SA Preclinical Experiments for
these 15 Compounds
10
SA Core Battery Studies
• In vitro hERG/Ikr assay:
– Human ether-a-go-go-related gene
(hERG) channel in CHO cells;
– APD: Papillary muscle (discovery)
/Rabbit Purkinje Fiber (development).
• In vivo QT assay:
– Conscious telemetered Dog;
– Toxicology Study Non-Rodent (at least 4
weeks).
11
Criteria of evaluation
• hERG: IC50 µM (concentration ng/mL);
• Effect on APD:
– Global (concentration ng/mL);
– Increase of APD: concentration ng/mL.
• Telemetered dog: ↑QTc (concentration
ng/mL);
• Toxicology Non-Rodent: ↑QTc
(concentration ng/mL).
Parent Compound
+ Relevant Metabolites
12
SA Thorough ECG Studies
(TES)
13
TES: Design
• Population:
– Healthy subjects (18/65 years): 16/16
[=Pharmacological model];
– Males AND Females: 16/16 (50/50)
[Gender effect = +4 msec];
– Mostly Caucasian.
• Repeated administration: 15/16 (SS of
unchanged compound + metabolites).
14
TES: Study Drug Doses
2 doses: 16/16 (1 with 3)
• Expected maximal therapeutic dose;
• + One supratherapeutic dose: DDI +
concentration/effect:
• 2 to 8x the expected maximal therapeutic
dose;
• 2 to 12x the expected maximal concentration
at the maximal therapeutic dose.
15
TES: Design
• Cross-over/Parallel:
– Cross-over: 6/16 – 1 to 7 days;
– Parallel: 10/16 studies – 3 to 28 days.
• Sample size:
– Cross-over: 16 to 45;
– Parallel: 64 to 114.
• Randomized, Double Blind: 16/16;
• Placebo controlled: 16/16.
16
TES: Design
• Positive control (establish assay
sensitivity): Moxifloxacin 16/16;
• Active control (same chemical
/pharmacological class): 2/16;
• ECG:
– Manual ECG reading + central lab: 16/16;
– Tangent method: 16/16;
– Multiple ECGs at baseline + during study:
intrinsic variability;
17
TES Design
– 0-12 h + window around the potential
Cmax;
– Timepoints in analysis window: 5 to 8.
• Central tendency analysis:
– Primary endpoint: average effect around
the median Cmax;
– Cross-over: raw QTcF;
– Parallel: QTcF change from baseline, with
baseline as covariate.
18
Drug
hERG
PF
In-vitro
In-vivo
PC Signal
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
hERG: G = > 10 µM; Y = 1 to 10 µM ; O < 1 µM
PF: G = 0; Y = ↑APD; O = ↑APD+EAD+triangulation
In vivo: G = 0; Y = ↑QTc (trend); O = ↑QTc (DP)
19
Drug
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
QTcF/TD
QTcF/STD
Dose/Conc 4/4 4/12 2.5/2.5 5/7 8/8 3/3 2/2 4/4 2/2 2/2 6/5 4/3 4/3 3/2 2.5/2.5
HR effect -/+ -/- -/- -/- -/- -/- -/+ -/- -/- +/+ -/- -/- -/- +/+ -/QTcN
C Signal
QTcF/QTcN: G < 5 msec MMI; O > 5 msec MMI
QTcN = Study specific correction
TD = Expected Max Ther Dose; STD = Supra Ther Dose
Dose = STD/TD; Conc = STDconc/TDconc
HR effect: - = No effect; + = Significant effect (TD/STD)
20
Drug
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
PC Signal v
C Signal
SUMMARY TN
h h/v
h h
h
v
FN TN FP TN TN TN TP TN TN FN TP FN FP FN
TN = true negative – 7/15
FP = False positive – 2/15
FN = False negative – 4/15
TP = True positive – 2/15
21
Case Study #1: False Positive (Drug 4)
• hERG: IC50 =
•
•
0.45 nM
↑ APD: 10 µM (DP)
In-vivo concious
non-rodent: ↑ QTc
•
•
•
•
•
POSITIVE
SIGNAL
QTcF TD = 0.04 msec
QTcF STD = 2.18 msec
Dose STD/TD = 5
Conc STD/TD = 7
(8.32 ng/mL)
No HR effect
NEGATIVE
TES
FALSE POSITIVE
22
Case Study #2: True Positive (Drug 12)
• hERG: IC50 =
•
•
0.22 µM
↑ APD + EAD
In-vivo concious
non-rodent: ↑ QTc
(trend)
• QTcF TD = 0.7 msec
• QTcF STD = 6.98**
•
•
•
POSITIVE
SIGNAL
msec
Dose STD/TD = 4
Conc STD/TD = 3
(35.6 ng/mL)
No HR effect
POSITIVE
TES
TRUE POSITIVE
23
Case Study #3: True Negative (Drug 5)
• hERG: IC50 = 5.15 • QTcF TD = -1.42 msec
µM/M = 25.7 µM
• QTcF STD = -2.95
• ↑ APD: No
• In-vivo concious
non-rodent: No
•
•
•
NEGATIVE
SIGNAL
msec
Dose STD/TD = 5
Conc STD/TD = 8
(2250 ng/mL)
No HR effect
NEGATIVE
TES
TRUE NEGATIVE
24
Case Study #4: False Negative (Drug 11)
• hERG: IC50 > 11.1 µM
•
•
(6133 ng/mL/M = 2.8
µM
↑ APD: No
In-vivo concious nonrodent: No ↑ QTc
NEGATIVE
SIGNAL
• QTcF TD = 0.0 msec
• QTcF STD = 10.0***
•
•
•
msec
Dose STD/TD = 6
Conc STD/TD = 5 (881
ng/mL/M = 115 ng/mL)
No HR effect
POSITIVE
TES
FALSE NEGATIVE
25
Summary of Evaluation of SA
Cardiac Repolarization Data for
15 Drugs
26
CONCLUSIONS ON PREDICTIVITY OF
PRECLINICAL MODELS
• Consistent with previous data:
– ILSI/HESI program;
– ABPI retrospective analysis;
– FDA experience on 19 drugs (2005).
• No universal predictivity:
– False positive and false negative for every test;
– Only one result should not to be considered.
27
CONCLUSIONS ON PREDICTIVITY OF
PRECLINICAL MODELS
• Careful search for confounding factors (PK, metab).
• Knowledge in that field is rapidly evolving (CPMP,
Points to consider in 12/1997 recommended PF test;
ICH S7B consider this test as only an additional test
and ask for hERG. In the (near?) future,
recommendations will evolve…
– Need to find a more accurate scoring for preclinical
assessment;
– Need of innovative methods of preclinical investigation.
28
Preclinical data captured the majority but
not all the drugs that prolong QT (2/15),
even with a conscientious methodology
consistent with ICH S7B and E14
29
Preclinical Evaluation
Strong signal, shallow water…don’t go there!
Same strong signal, but deep water…safe!
Regulatory guidance must not expect no
signal…..very safe but potentially no
drugs!
Never ?
30
SUMMARY ON PREDICTIVITY OF CLINICAL
THOROUGH ECG STUDY
NEXT STEP ….. BE PATIENT
AND ENJOY THIS NICE
MEETING!
31
Back-up
32
The First in vitro Model: hERG Channel
Whole cell Patch clamp method: hERG channel
stably expressed in Chinese Hamster Ovarian
cell (IKr)
Used Protocol
Recording
Ikr Inhibition / Cumulative
Concentrations
33
hERG Channel: Sources of Variability
Experimental design
Statistics
Criteria of acceptation
Analysis of conc response curve
Expression system
Number of conc to built
a conc - response curve
[K + ] in superfusion
medium
Main hERG
patch -clamp
issues
Number of cells/
concentration
Superfusion
temperature
Validation with a
reference
Stimulation
protocol
Current reading
protocols
Reversibillity
Nature of blockade
Rundown quantification
Incubation time
Limitations of in vitro models
34
The Second in vitro Model
PAPILLARY MUSCLE: GUNEA-PIG (Discovery)
PURKINJE FIBER: RABBIT (Development)
•Measure the action potential of a conductive
tissue (or cell);
•Action potential = result of the activity of the
various channels involved in the depolarization
and in the repolarization phases;
35
The Second in vitro Model
PAPILLARY MUSCLE: GUNEA-PIG (Discovery)
PURKINJE FIBER: RABBIT (Development)
•Microelectrode technique;
•Tissue paced at fixed frequency; bradycardia
can be simulated by pacing at low rate.
36
The Second in vitro Model:
Purkinje/Papillary
EXAMPLE:
1 Hz
37
Action Potential: Source of Variability
a: nœud sinusal
b: tissu atrial (humain)
c: fibre de Purkinje (chien)
d: tissu épicardique (chien)
e: tissu endocardique (chien)
f: tissu myocardique (rat)
38
Action Potential: Effect of Various Channels
Control
Control
Control
IKr
INa
ICa
But also:
IKto1
IKto2
IKs
INa-Ca
IK1
39
ACTION POTENTIAL EXPERIMENTS
•Not be used as a stand-alone experiment to appreciate the
risk of TdP linked to hERG inhibition;
•Global view on the AP, taking
into account all effects on the various channels;
•Considered as follow-up studies (in addition to hERG and
in vivo tests) and can be used to perform the Integrated
Risk Assessment before First-entry-Into-Man and then at
every time new information is coming either from clinical
trials or fromlong-term toxicological studies.
40
Central Tendency:
Method 1 = Maximum Mean DDQTc
DDQTc
ANALYSIS
Method 2 = Mean DDQTc at Cmax
DDQTc
Time
Time of Cmax
41
CALCULATING MAX MEAN CHANGE
Drug
Maximum
Mean DDQTc
Time
Baseline
Subtracted:
DQTc
Baseline
Placebo
QTc
Time
Baseline
Time
DQTc =
Response
adjusted (D) for
baseline effect
DDQTc
QTc
Mean DDQTc
Placebo
Subtracted:
Time
DDQTc =
Response
adjusted (DD)
for baseline &
placebo effect42
CALCULATING MEAN CHANGE AT CMAX
Mean DDQTc
at Cmax
Drug
Time
Baseline
Subtracted:
DQTc
Baseline
Placebo
QTc
Time
Baseline
Time
DQTc =
Response
adjusted (D) for
baseline effect
DDQTc
QTc
Mean DDQTc
Placebo
Subtracted:
Time
DDQTc =
Response
adjusted (DD)
for baseline 43
&