Dr. Jay Mason, Chief Medical Officer, Spaulding Clinical Research presents on Cardiac Surveillance Strategies to Improve Safety and Avoiding the Thorough QT Study

1. The Thorough QT Study
Considerations for its Improvement
Jay W. Mason, MD
February 7, 2013
Webinar Sponsored by Spaulding Clinical research

2. Drugs Withdrawn for ↑QT
Year Drug Indication
1988 Prenylamine (Elecor) Angina
1989 Lidoflazine (Angex) Angina
1991 Terodiline (Mictrol) Bladder spasm
1998 Terfenadine (Seldane) Allergy
1998 Sertindole (Serdolect) Psychosis (revived)
1999 Grepfloxacin (Raxar) Infection
1999 Astemizole (Hismanal) Allergy
2000 Cisapride (Propulsid) GERD, gastroperesis
2001 Droperidol (Inapsine) Nausea
2001 Levacetylmethadol (Orlaam) Opiate addiction
2004 Dofetilide (Tikosyn) Atrial fib/flut (not US)
2005 Thioridazine (Mellaril) Psychosis (not US)
2007 Clobutinol (Silomat) Cough
2010 Dolasetron i.v. (Anzemet) Nausea
2011 Propoxyphene (Darvon) Pain

3. Outline
• The basis for and history of the ICH E-14 Guidance
• Current status of the Guidance
– Q&A issued by ICH, FDA, Health Canada
– Additional clarifications in FDA response letters
– Changes and updates in a nutshell
• What could be improved?
– Timing
– Moxifloxacin exposure
– More informative endpoints
– Cost
• Alternative strategies
– Front load and taper
– Eliminate the active control arm
– Add better endpoints: T-wave quantification
– Control cost
• Automation
• Purpose-built device and analysis

5. E-14 Requirements: Thorough QT Study
• Applies to almost all new drugs (>95%)
• Detect 5 msec (UCB 10 msec) change in QTc
• Between 4,000 and 30,000 ECGs per study
• Centralized ECG interpretation
• Placebo control
• Active control (moxifloxacin)
• Supratherapeutic dose (4X – 10X)
• Performed prior to approval, usually after Phase 2
• Annotated XML file

6. Additions and Clarifications
• Thorough analysis of heart rate, PR and QRS should be
included
• Clinical dose arm not required
• Blinding of moxifloxacin not required
• Automation is acceptable when a positive control is
included
• Zhang/Machado sample size recommendations
• Bazett QT correction no longer required
• Multiple endpoint adjustment to control type I error
required for the assay sensitivity test
• Account for possible delayed effect
• Perform PK-PD modeling

7. What could be improved?: Timing
Phase 1 Phase 2 Phase 3 Phase 4
ECG Quality
ECG Quantity
ECG Information
Two Problems with this timing
One
• ECG quality, quantity and resulting information
are lowest in Phase 1
• Ironically, the highest doses and most careful
exploration of pharmacokinetics occur in Phase 1
• A huge opportunity lost
Two
• Subjects are exposed to the drug in Phases 1 and
2 at supratherapeutic doses
• Often several hundred subjects and patients are
exposed in Phases 1 and 2
• Aren’t they entitled to the same protection the
TQTS is designed to guarantee in Phase 3?

8. Alternative Timing Strategy
Strategy
• Front load ECG acquisition
• Centralize and maintain high quality
throughout the program
• Reduce frequency of time points and
replicates as newly gained information
warrants
Results
• High quality ECG information gathered
at maximum dosage
• Early warning of QT and other ECG
issues
• Subject/patient safety improved
• Early out
• TQTS avoided
• Cost reduced?
Phase 1 Phase 2 Phase 3 Phase 4

9. What could be improved?: Moxifloxacin Exposure
Moxifloxacin is not totally benign
– QT prolongation and torsades de pointes
– Hypersensitivity reaction (SJS, TEN)
– C. difficile (CDAD)
– Tendon rupture, etc.
Alternative Moxifloxacin strategies
– High quantity and quality of ECGs reduce need for active control
– Methods and investigative sites that accurately measure QTc change are now
well known
– Maintenance of high quality in Phase 3 (and 4) reduces chance of missed
repolarization liability
– Other sound methods for demonstrating assay sensitivity have been proposed
• Positional changes in QTc
• Statistical methods to discern QTc precision

10. What could be improved?: Endpoints
• QTc prolongation has extremely low specificity for torsades de pointes
• T-wave change has higher specificity and sensitivity in animal models
• T-waves are currently assessed subjectively without quantification
• T-wave segmentation (e.g., Tpeak – Tend)
• Serial quantification of eigenvectors (e.g., BioQT)
Alternative strategy: Quantify T-wave

11. Holter T-Wave Monitoring

12. What could be improved?: Cost
,000,000
0,000

13. Alternative Cost Strategy: Automation
• Faster
• Less expensive
• Human error eliminated
• Variability decreased
• Sample size decreased
• The impossible made possible
– Well-founded QTcI
– T-wave morphology tracking
– Other analyses requiring large quantities or
continuous data.

14. Types of Automation
• On-board computer algorithms
– GE SL12
– Mortara Veritas
– Philips ECL (now DXL)
• New stand-alone software
– BioQT (OBS Medical)
– QTinno (NewCardio)
– Compass (iCardiac)

15. • Reduce the cost of manufacturing and leasing the
electrocardiograph
• Reduce the cost of shipping the device
• Simplify ECG acquisition to reduce training costs
• Eliminate the cost of human error
• Enhance data quantity, quality and reliability
Alternative Cost Strategy: Purpose-
built devices

16. Ahhh! No
more TQTs!!

17. TQT Alternative Strategy
• Front-load and taper ECG acquisition to get
earlier repolarization answer
• Eliminate the active control arm for patient
safety and cost reduction
• Monitor T-wave morphology for better
assessment of risk
• Automate ECG analysis to reduce cost
• Use more cost effective devices and analyses

18. • Implement a sufficiently simple, inexpensive, reliable ECG
process to allow high-quality ECG collection from FIM to
pivotal phase 3 trials and beyond.
• Continuously update ECG safety prediction.
• Adjust safety monitoring requirements in accordance
with prediction.
• Carry requirements into phase 4 if necessary.
• No TQT!
• “Thorough ECG Strategy”
Thorough ECG Strategy