What should be the optimal Design for TAVI Procedure. CRT 2010 Washington DC, January 21, 2010 What should be the optimal Design for TAVI Procedure. Eberhard Grube, MD, FACC, FSCAI St.Elisabeth Hospital, Heart Center Rhein-Ruhr, Essen, Germany Instituto Cardiologico Dante Pazzanese, São Paulo, Brazil

Eberhard Grube, MD DISCLOSURES Consulting Fees Honoraria Abbott Vascular, Boston Scientific Corporation, Cordis, a Johnson & Johnson Company, Medtronic CardioVascular, Inc. Honoraria Biosensors International , Boston Scientific Corporation, Medtronic CardioVascular, Inc Ownership Interest (Stocks, Stock Options or Other Ownership Interest) Biosensors International , Medtronic CardioVascular, Inc. I intend to reference unlabeled/ unapproved uses of drugs or devices in my presentation. I intend to reference off-label use of stents and valve prosthesis.

Evidence is key in Establishing TAVI as a new standard Evidence is critical is gaining Regulatory approvals New Indications and Applications Coverage and reimbursement Adoption by implanters Adoption by referrers

To make TAVI the Gold Standard, what data to collect? – Everything ! The key aspects need to be collected: Safety Effectiveness Clinical Outcomes Cost Effectiveness Durability Need to cover all timeframes: Procedural / Accute Short Term Long Term We need to progressively cover all population groups: Inoperable High Risk Medium Risk Lower Risk patients All Applications: Transfemoral, Transapical, Direct Aortic Re-Dos, Valve-in-Valve

Issues with current data accumulated Comparisons of outcomes between TAVI and Surgical are neither fair nor favorable Key issues are disparities: Different population groups being compared Different maturities of the technologies Other issues: Little long term data accumulated so far Initial studies were started with little understanding of the landscape, patients groups, and potential outcomes

Differences of perspective by specialty/therapy are key Higher Risk Patients Lower Risk Patients TAVI Candidates Surgical Candidates Surgeons’ view Surgical Candidates TAVI Candidates Interventionalists’ view Clearly TAVI area Clearly Surgical TAVI/Surgical zone of contention

And they feed the debate on comparative outcomes Higher Risk Patients Lower Risk Patients Surgeons’ view TAVI Candidates Surgical Candidates Interventionalists’ view TAVI Candidates Surgical Candidates

Randomized Trials have strong appeal but they have their drawbacks RCTs are trusted by all agencies They offer really good controls over the population groups being studied They are very effective in pre-market conditions They are extremely expensive, require huge infrastructure They only offer a narrow, specific slice of the population Risk of selective enrolling once commercial options are available, on or off-label

Observational studies have another set of drawbacks Observational Studies are cheaper and faster to enroll They offer a broad slice of real-world population They are effective in post-market conditions The results are sometimes second-guessed The results tend to be blurred by the mix of different patients subgroups They are hard to use to get new indications or approvals

There are specific issues with these trials in the context of TAVI The comparison with the established Surgical outcomes looks unfavorable: Because of adverse patient selection Because of early user experience in TAVI vs experienced users in Surgical There is no clarity on what the patients’ risks profiles are: Some surgeons will operate on anybody, others will be more conservative Differences between what people would say and what they would do with these patients Risk profiles STS or EuroScores are outdated with the current technologies Hence the difficulty for TAVI to establish itself: No clarity for which exact patient group TAVI is indicated No clarity of what the true comparative advantages and disadvantages of TAVI are over surgery in groups where both are indicated

A potential third way? TAVI Observational Arm If Surgeons agree they are not surgical candidates Review by Surgeons and Interventionalists Surgical Observational Arm If Interventionalists agree they are surgical candidates Patients Randomized group between TAVI and Surgical If there is a doubt

The three-arms design has advantages in terms of control of the population TAVI Observational Arm If Surgeons agree they are not surgical candidates Confirms that there are truly inoperable patients Lack of viable alternative legitimates observational approach Outcomes in this cohort would be isolated and not compared to straight surgical results Surgical Observational Arm If Interventionalists agree they are surgical candidates Would keep the truly surgical candidates in surgery Would keep the outcomes as specific to that population Randomized group between TAVI and Surgical If there is a doubt or conflict Would allow randomization without adverse selection bias Would allow somewhat more homogeneous population selection (reverse consensus)

TAVI Observational Arm advantages Would prevent adverse selection bias in RCT group by isolating the highest risk TAVI patients Could provide meaningful data on outcomes and long term benefits for high risk / inoperable Would quantify concretely the magnitude of the inoperable / high risk population Could be used for coverage and/or indications expansion in that category

Surgical Observational Arm Advantages Would prevent adverse selection bias in RCT group by isolating the lowest risk Surgical patients Would establish a clear benchmark for TAVI to meet in order to expand into lower risk patients

Randomized TAVI/Surgical Arms Randomized group between TAVI and Surgical Obvious TAVI or surgical candidates would be sent to their specific arms RCT Arm would be made up of more homogeneous risk profile Outcomes’ comparisons would be more valid between TAVI and Surgical randomized groups

In summary A hybrid approach could be advantageous, leveraging the observational arms to maximize the effectiveness of randomization over a naturally well-selected group