← Product Code NPT · P110021

# EDWARDS SAPIEN TRANSCATHETER HEART VALVE AND ACCESSORIES (P110021)

_Edwards Lifesciences, LLC · NPT · Oct 19, 2012 · Cardiovascular · APWD_

**Canonical URL:** https://fda.innolitics.com/device/P110021

## Device Facts

- **Applicant:** Edwards Lifesciences, LLC
- **Product Code:** NPT
- **Decision Date:** Oct 19, 2012
- **Decision:** APWD
- **Device Class:** Class 3
- **Review Panel:** Cardiovascular
- **Attributes:** Therapeutic, Expedited Review

## Intended Use

The Edwards SAPIEN Transcatheter Heart Valve, model 9000TFX, sizes 23 mm and 26 mm, is indicated for transfemoral delivery in patients with severe symptomatic calcified native aortic valve stenosis without severe aortic insufficiency and with ejection fraction >20% who have been examined by a heart team including an experienced cardiac surgeon and a cardiologist and found to either be: 1) inoperable and in whom existing co-morbidities would not preclude the expected benefit from correction of the aortic stenosis; or 2) be operative candidates for aortic valve replacement but who have a Society of Thoracic Surgeons predicted operative risk score ≥8% or are judged by the heart team to be at a ≥15% risk of mortality for surgical aortic valve replacement. The RetroFlex 3 Delivery System is indicated for the transfemoral delivery of the Edwards SAPIEN transcatheter heart valve. The Edwards SAPIEN transcatheter heart valve, model 9000TFX, sizes 23 mm and 26 mm, is indicated for transapical delivery in patients with severe symptomatic calcified native aortic valve stenosis without severe aortic insufficiency and with ejection fraction >20% who have been examined by a heart team including an experienced cardiac surgeon and a cardiologist and found to be operative candidates for aortic valve replacement but who have a Society of Thoracic Surgeons operative risk score ≥8% or are judged by the heart team to be at a ≥15% risk of mortality for surgical aortic valve replacement. The Ascendra Balloon Catheter is indicated for the transapical delivery of the Edwards SAPIEN transcatheter heart valve.

## Device Story

The Edwards SAPIEN Transcatheter Heart Valve (THV) is a balloon-expandable, radiopaque, stainless steel (316L) frame with three bovine pericardial tissue leaflets and a PET fabric skirt. It is used for transcatheter aortic valve implantation (TAVI) in patients with severe symptomatic native aortic valve stenosis. The device is delivered via transfemoral or transapical access using specialized delivery systems (RetroFlex 3 or Ascendra). The procedure involves pre-dilation of the stenotic valve using a balloon catheter, followed by crimping and deployment of the SAPIEN valve. The device is used in a catheterization laboratory or OR by a heart team. The valve functions as a replacement for the native aortic valve, restoring hemodynamic function and improving patient symptoms. Potential benefits include avoiding open-heart surgery in high-risk patients. Clinical outcomes are monitored via echocardiography and clinical assessment.

## Clinical Evidence

The PARTNER trial (Cohort A) was a prospective, randomized, controlled, multi-center pivotal trial (n=699) comparing TAVI (SAPIEN) to surgical AVR in high-risk patients. Primary endpoint: freedom from all-cause mortality at 1 year. Results: TAVI was non-inferior to AVR (1-year mortality: 24.27% TAVI vs 26.80% AVR; p=0.0014). Secondary endpoints included MACCE, NYHA class, and 6-minute walk test. TAVI showed higher rates of stroke (5.5% vs 2.9%) and vascular complications (11.0% vs 3.8%) but lower bleeding rates compared to AVR. Clinical data supported the benefit-risk profile for high-risk patients.

## Technological Characteristics

Materials: 316L stainless steel frame, bovine pericardial tissue (Carpentier-Edwards ThermaFix process), PET fabric. Sensing/Actuation: Balloon-expandable frame. Energy: Mechanical deployment via balloon inflation. Connectivity: None. Sterilization: Terminal liquid sterilization (glutaraldehyde) for valve; Ethylene Oxide (EO) for delivery systems. Dimensions: 23mm and 26mm sizes. MR Conditional (1.5T/3.0T).

## Regulatory Identification

To replace a patient's aortic heart valve.  They are different from the classified device (heart valves) in that they are placed percutaneously and do not require open chest surgery or a cardiotomy for placement.

## Reference Devices

- Cribier-Edwards™ Aortic Bioprosthesis, Model 9000
- Carpentier-Edwards® PERIMOUNT® Pericardial Bioprosthesis
- Cordis Palmaz Genesis stents

## Submission Summary (Full Text)

> This content was OCRed from public FDA records by [Innolitics](https://innolitics.com). If you use, quote, summarize, crawl, or train on this content, cite Innolitics at https://innolitics.com.
>
> Innolitics is a medical-device software consultancy. We help companies design, build, and clear FDA-regulated software and AI/ML devices, including [a PMA](https://innolitics.com/services/regulatory/), [a 510(k)](https://innolitics.com/services/510ks/), [a SaMD](https://innolitics.com/services/end-to-end-samd/), [an AI/ML medical device](https://innolitics.com/services/medical-imaging-ai-development/), or [an FDA regulatory strategy](https://innolitics.com/services/regulatory/).

# SUMMARY OF SAFETY AND EFFECTIVENESS DATA

## 1. GENERAL INFORMATION

Device Generic Name: Replacement Heart Valve

Device Trade Name: Edwards SAPIEN™ Transcatheter Heart Valve Model 9000TFX, 23 and 26mm, and accessories (RetroFlex 3™ Delivery System, Models 9120FS23 and 9120FS26, RetroFlex™ Balloon Catheter, Models 9120BC20 and 9120BC23, Ascendra™ Balloon Catheter, Models 9100BCL23 and 9100BCL26, Ascendra™ Balloon Aortic Valvuloplasty Catheter, Model 9100BAVC, Ascendra™ Introducer Sheath Set Model 9100IS, Crimper, Models 9100CR23 and 9100CR26)

Applicant Name and Address: Edwards Lifesciences LLC, One Edwards Way, Irvine, CA 9261

Date of Panel Recommendation: June 13, 2012

PMA Application Number: P110021

Date of FDA Notice of Approval: October 19, 2012

Expedited: Granted expedited review status on May 2, 2011 because the SAPIEN device represents a breakthrough technology that provides a clinically meaningful option in a high risk patient population.

The SAPIEN Transcatheter Heart Valve was previously approved for use in inoperable patients under P100041. The Summary of Safety and Effectiveness supporting the previous indication is available on the CDRH web site http://www.accessdata.fda.gov/cdrh_docs/pdf10/P100041b.pdf. Previously, the Indication for Use (for the inoperable patient) read as follows:

The Edwards SAPIEN Transcatheter Heart Valve (THV), model 9000TFX, sizes 23mm and 26mm, is indicated for transfemoral delivery in patients with severe symptomatic native aortic valve stenosis who have been determined by a cardiac surgeon to be inoperable for open aortic valve replacement and in whom existing co-morbidities would not preclude the expected benefit from correction of the aortic stenosis.

8
# 2. INDICATIONS FOR USE

## TRANSFEMORAL PROCEDURE

The Edwards SAPIEN Transcatheter Heart Valve, model 9000TFX, sizes 23 mm and 26 mm, is indicated for transfemoral delivery in patients with severe symptomatic calcified native aortic valve stenosis without severe aortic insufficiency and with ejection fraction >20% who have been examined by a heart team including an experienced cardiac surgeon and a cardiologist and found to either be: 1) inoperable and in whom existing co-morbidities would not preclude the expected benefit from correction of the aortic stenosis; or 2) be operative candidates for aortic valve replacement but who have a Society of Thoracic Surgeons predicted operative risk score ≥8% or are judged by the heart team to be at a ≥15% risk of mortality for surgical aortic valve replacement.

The RetroFlex 3 Delivery System is indicated for the transfemoral delivery of the Edwards SAPIEN transcatheter heart valve.

## TRANSAPICAL PROCEDURE

The Edwards SAPIEN transcatheter heart valve, model 9000TFX, sizes 23 mm and 26 mm, is indicated for transapical delivery in patients with severe symptomatic calcified native aortic valve stenosis without severe aortic insufficiency and with ejection fraction >20% who have been examined by a heart team including an experienced cardiac surgeon and a cardiologist and found to be operative candidates for aortic valve replacement but who have a Society of Thoracic Surgeons operative risk score ≥8% or are judged by the heart team to be at a ≥15% risk of mortality for surgical aortic valve replacement.

The Ascendra Balloon Catheter is indicated for the transapical delivery of the Edwards SAPIEN transcatheter heart valve.

# 3. CONTRAINDICATIONS

The bioprosthesis and delivery system are contraindicated in patients who cannot tolerate an anticoagulation/antiplatelet regimen or who have active bacterial endocarditis or other active infections.

# 4. WARNINGS AND PRECAUTIONS

The warnings and precautions can be found in the labeling for the SAPIEN Transcatheter Heart Valve with the Ascendra Delivery System and in the SAPIEN Transcatheter Heart Valve with the Retroflex 3 Delivery System.

# 5. DEVICE DESCRIPTION

## 5.1 SAPIEN Transcatheter Heart Valve

The Edwards SAPIEN Transcatheter Heart Valve (bioprosthesis), intended for transcatheter valve implantation (TAVI), shown in Figure 1, is comprised of a balloon-expandable, radiopaque, stainless steel (316L) frame, three bovine pericardial tissue P110021: FDA Summary of Safety and Effectiveness Data Page 2 of 55

leaflets, and a polyethylene terephthalate (PET) fabric. The bioprosthesis is treated according to the Carpentier-Edwards ThermaFix process, packaged, and terminally sterilized in glutaraldehyde.

![img-0.jpeg](img-0.jpeg)
Figure 1 - Edwards SAPIEN Transcatheter Heart Valve

## 5.2 The RetroFlex 3 Delivery System

The RetroFlex 3 Delivery System, shown in Figure 2, includes a rotating wheel within the handle for articulation of flex catheter, a tapered tip at the distal end of the delivery system to facilitate crossing the native valve, a balloon for deployment of the bioprosthesis, and radiopaque markers.

![img-1.jpeg](img-1.jpeg)
Figure 2 - RetroFlex 3 Delivery System

## 5.3 The RetroFlex Balloon Catheter

The RetroFlex Balloon Catheter, shown in Figure 3, is used to pre-dilate stenotic cardiac valves. The device consists of a shaft and balloon with radiopaque markers indicating working length of the balloon. At the proximal end of the device, there is a standard “Y-connector” for balloon inflation and guidewire insertion.

![img-2.jpeg](img-2.jpeg)
Figure 3 - RetroFlex Balloon Catheter

## 5.4 The Ascendra Balloon Catheter

The Ascendra Balloon Catheter, shown in Figure 4, includes a handle with deflection mechanism, a pusher tip to aid in valve placement, a balloon for deployment of the bioprosthesis, and radiopaque markers indicating working length of the balloon. At the proximal end of the device, there is a standard “Y-connector” for balloon inflation and guidewire insertion. The Ascendra Balloon Catheter is supplied with a loader that covers

P110021: FDA Summary of Safety and Effectiveness Data

the crimped valve as it is inserted into the Ascendra sheath, and optional extension tubing.

![img-3.jpeg](img-3.jpeg)
Figure 4 - Ascendra Balloon Catheter

## 5.5 The Ascendra Balloon Aortic Valvuloplasty Catheter

The Ascendra Balloon Aortic Valvuloplasty Catheter, shown in Figure 5, is a coaxial catheter with a distal inflatable balloon intended to pre-dilate the stenotic aortic valve prior to implantation of the bioprosthesis. Two radiopaque marker bands indicate the dilating section of the balloon and aid in balloon placement. At the proximal end of the catheter, there is a standard “Y” connector for balloon inflation and a guidewire lumen. The Ascendra Balloon Catheter is supplied with optional extension tubing.

![img-4.jpeg](img-4.jpeg)
Figure 5 - Ascendra Balloon Aortic Valvuloplasty Catheter

## 5.6 The Ascendra Introducer Sheath Set

The Ascendra Introducer Sheath Set, shown in Figure 6, is used for the introduction and removal of devices used with the Edwards SAPIEN transcatheter heart valve. It consists of two components, sheath and introducer. The introducer is inserted in the sheath during device preparation. A radiopaque marker is located at the sheath tip for visualization when inserting the sheath. There are printed non-radiopaque depth markings on the distal end of the body that can be used to gauge the depth of distal end of the sheath within the ventricle.

![img-5.jpeg](img-5.jpeg)
Figure 6 - Ascendra Introducer Sheath Set

P110021: FDA Summary of Safety and Effectiveness Data

P110021: FDA Summary of Safety and Effectiveness Data
Page 5 of 55

# 5.7 The Crimper

The Crimper, shown in Figure 4, is comprised of a housing and a compression mechanism, creating an aperture that is opened and closed by means of a handle located on the housing. The crimper includes a balloon gauge to verify diameter of an inflated balloon catheter and a crimp gauge to verify collapsed diameter of the device.

![img-6.jpeg](img-6.jpeg)
Figure 7 - Crimper

# 6. ALTERNATIVE PRACTICES AND PROCEDURES

Alternatives for patients with severe symptomatic native aortic valve stenosis deemed to be at excessive risk for surgery, or non-operable (non-surgical) include temporary relief using a percutaneous technique called balloon aortic valvuloplasty (BAV) or medical therapy (no obstruction-relieving intervention). For patients who are operable, surgical aortic valve replacement (AVR) is an alternative. Each alternative has its own advantages and disadvantages. A patient should fully discuss these alternatives with his/her physician to select the method that best meets expectations and lifestyle.

# 7. MARKETING HISTORY

Commercial distribution of the SAPIEN Transcatheter heart valve Model 9000TFX and accessories outside the United States (U.S.) began in October 2007. The SAPIEN Transcatheter Heart Valve has been marketed in the U.S. for inoperable patients since November 2, 2011. Currently, the device is approved for distribution in the 27 member states under the European Union, Croatia, Iran, Israel, Jordan, Kuwait, Monaco, Norway, Russia, Saudi Arabia, Singapore, South Africa, Switzerland, Thailand and Turkey. The SAPIEN valve and accessories have not been withdrawn from the market in any country for any reason related to the safety and effectiveness of the device.

# 8. POTENTIAL ADVERSE EFFECTS OF THE DEVICE ON HEALTH

The adverse events listed below are associated with access complications associated with catheterization or valvuloplasty, and events associated with local and/or general anesthesia.

- Death
- Stroke/transient ischemic attack or neurological deficit
- Paralysis
- Permanent disability

- Respiratory insufficiency or respiratory failure
- Hemorrhage requiring transfusion or intervention
- Cardiovascular injury including perforation or dissection of vessels, ventricle, myocardium or valvular structures that may require intervention
- Pericardial effusion or cardiac tamponade
- Embolization including air, calcific valve material or thrombus
- Infection including septicemia and endocarditis
- Heart failure
- Myocardial infarction
- Renal insufficiency or renal failure
- Conduction system injury (defect) which may require a permanent pacemaker
- Arrhythmia
- Retroperitoneal bleed
- Femoral AV fistula or pseudoaneurysm
- Reoperation
- Peripheral ischemia or nerve injury
- Restenosis
- Pulmonary edema
- Pleural effusion
- Bleeding
- Anemia
- Abnormal lab values (including electrolyte imbalance)
- Hypertension or hypotension;
- Allergic reaction to anesthesia or to contrast media
- Hematoma
- Syncope
- Pain or changes at the access site
- Exercise intolerance or weakness
- Inflammation
- Angina
- Heart murmur
- Fever

Additional potential risks specifically associated with the use of the bioprosthesis include, but may not be limited to the following:

- Stroke
- Vascular injury necessitating graft placement
- Cardiac arrest
- Cardiogenic shock
- Emergency cardiac surgery
- Cardiac failure or low cardiac output
- Coronary flow obstruction/transvalvular flow disturbance
- Device thrombosis requiring intervention
- Valve thrombosis
- Device embolization
- Device migration or malposition requiring intervention

P110021: FDA Summary of Safety and Effectiveness Data

- Valve deployment in unintended location
- Valve stenosis
- Structural valve deterioration (wear, fracture, calcification, leaflet tear/tearing from the stent posts, leaflets retraction, stent creep, suture line disruption of components of a prosthetic valve, thickening, stenosis, or other)
- Device degeneration
- Paravalvular or transvalvular leak
- Valve regurgitation
- Hemolysis
- Device explants
- Nonstructural dysfunction
- Non-emergent reoperation

For the specific adverse events that occurred in the clinical study, please see Section 10 below.

## 9. SUMMARY OF PRECLINICAL STUDIES

*In vitro* studies were performed for the Edwards SAPIEN Transcatheter Heart Valve Model 9000TFX and non-implantable accessories as recommended in the ISO 5840: *Cardiovascular Implants-Cardiac Valve Prostheses*. (2005) standard.

### 9.1 Biocompatibility Studies

Toxicology and biocompatibility testing for the SAPIEN Transcatheter Heart Valve Model 9000TFX and accessories was conducted in accordance with Good Laboratory Practices (21 CFR §58) and ISO 10993-1: 2003 *Biological Evaluation of Medical Devices Part 1: Evaluation and Testing*.

Summaries of the test results for the SAPIEN Transcatheter Heart Valve Model 9000TFX are provided in Table 1. Summaries of the test results for the RetroFlex 3 Delivery System, RetroFlex Balloon Catheter, Ascendra Balloon Catheter, Ascendra Introducer Sheath Set, Ascendra Balloon Aortic Valvuloplasty Catheter are provided in Table 2. Results for the Crimper are provided in Table 3. Test samples for the studies consisted of all patient-contacting portions of the devices (direct and indirect contact) after all manufacturing processes including sterilant exposure. All results were found to be acceptable.

Table 1 - Summary of Biocompatibility Testing - SAPIEN Transcatheter Heart Valve Model 9000TFX

|  Test | Purpose | Results  |
| --- | --- | --- |
|  Cytotoxicity: Percent Inhibition of Cell Growth | Determine whether test article extract would inhibit cell growth | Test article found to be non-inhibitory to cell growth at a sample concentration representative of the device’s clinical application. Inhibitory to cell growth at elevated sample concentrations.  |
|  Cytotoxicity: Medium Eluate Method (MEM) | Determine whether test article extracts would cause cytotoxicity and cell lysis | Test article sample was non-cytotoxic. 0% cell lysis was observed with equivalent results to the negative control.  |

P110021: FDA Summary of Safety and Effectiveness Data
Page 7 of 55

P110021: FDA Summary of Safety and Effectiveness Data
Page 8 of 55

|  Test | Purpose | Results  |
| --- | --- | --- |
|  Cytotoxicity: Agar Overlay Test | Determine whether solid samples of test article would cause cytotoxicity and cell lysis. | Solid samples of the stent frame were non-cytotoxic. 0% cell lysis was observed with equivalent results to the negative control.
Cytotoxicity was observed in solid samples of the cloth, suture, and tissue material due to glutaraldehyde and formaldehyde residuals present in the solid sample.  |
|  Sensitization: Guinea Pig Maximization | Investigate the potential for delayed dermal contact sensitization. | No irritation was present on any of the test or control animals at 24 or 48 hour readings using saline and vegetable oil extracts. Non-sensitizing.  |
|  Irritation/Intracutaneous Toxicity: Rabbit Intracutaneous Reactivity | Determine whether test article extracts would cause local dermal irritation or toxic effects | No evidence of irritation or abnormal effects over a 72 hour period as compared to negative controls.  |
|  Systemic Toxicity: USP Mouse Systemic Injection | Determine whether test article extracts would cause acute systemic toxicity | No weight differences or observed systemic effects as compared to negative controls over 72 hour test period.  |
|  Systemic Toxicity: Material Mediated (Rabbit) Pyrogen Test | Determine the presence of chemical pyrogens in test article extracts by measuring temperature rise in intravenously injected rabbits. | No temperature rise or abnormalities in any test or control animals.  |
|  Implantation
Subacute/Subchronic Toxicity
Chronic Toxicity | Determine whether the test article would cause systemic toxicity affects after 7, 30, and 90 days intramuscular implantation in rabbits. | No microscopic evidence of cytotoxicity.  |
|  Genotoxicity: Ames Test – Plate Incorporation | Determine whether test article extracts would cause mutagenic changes in five S. typhimurium strains | Test article extracts demonstrated no mutagenic potential under both the activated and non-activated conditions.  |
|  Genotoxicity: Chromosomal Aberration Assay | Determine whether test article extracts would cause genotoxicity in Chinese Hamster ovary cells | Test article extracts demonstrated no mutagenic potential under both the activated and non-activated conditions.  |
|  Genotoxicity: Mouse Micronucleus | Determine whether test article extracts would cause genotoxic changes as determined by induced micronucleated polychromatic erythrocytes | Test article extracts were determined to be non-mutagenic.  |
|  Hemocompatibility: Hemolysis | Determine whether the test article would cause hemolysis in vitro and determine the degree of inhibition or promotion of clotting time | No hemolytic effects observed under static conditions for both extract and solid samples. Material’s extract did not adversely affect the clotting time and was determined to be compatible with plasma.  |

Table 2 - Summary of Biocompatibility Testing for RetroFlex 3 Delivery System, RetroFlex Balloon Catheter, Ascendra Balloon Catheter, Ascendra Balloon Aortic Valvuloplasty Catheter, and Ascendra Introducer Sheath Set

|  Test | Purpose | Results  |
| --- | --- | --- |
|  Cytotoxicity: Medium Eluate Method (MEM) | Determine whether test article extracts would cause cytotoxicity and cell lysis | Test article sample was non-cytotoxic. 0% cell lysis was observed with equivalent results to the negative control.  |
|  Cytotoxicity: Agar Overlay Test | Determine whether solid samples of test article would cause cytotoxicity and cell lysis | Solid samples of test articles were non-cytotoxic. 0% cell lysis was observed with equivalent results to the negative control.  |
|  Sensitization: Guinea Pig Maximization | Investigate the potential for delayed dermal contact sensitization | No irritation was present on any of the test or control animals at 24 or 48 hour readings using saline and vegetable oil extracts. Non-sensitizing.  |
|  Irritation/Intracutaneous Toxicity: Rabbit Intracutaneous Reactivity | Determine whether test article extracts would cause local dermal irritation or toxic effects | No evidence of irritation over a 72 hour period as compared to negative controls.  |
|  Systemic Toxicity: USP Mouse Systemic Injection | Determine whether test article extracts would cause acute systemic toxicity | No weight differences or observed systemic effects as compared to negative controls over 72 hour test period.  |
|  Systemic Toxicity: Material Mediated (Rabbit) Pyrogen Test | Determine the presence of chemical pyrogens in test article extracts by measuring temperature rise in intravenously injected rabbits. | Temperature rise of ≤0.1°C and no abnormalities in any test or control animals.  |
|  Hemocompatibility: Hemolysis | Determine whether the test article would cause hemolysis in vitro and determine the degree of inhibition or promotion of clotting time | No hemolytic effects observed under static conditions for both extract and solid samples. Material’s extract did not adversely affect the clotting time and was determined to be compatible with plasma.  |
|  Hemocompatibility: Complement Activation | Evaluate the test article’s potential to activate the C3 and C5 complement system | Test article was determined to be hemocompatible and not at risk to activate complement at a level of concern in a clinical application. Results equivalent to negative control.  |

Table 3 - Summary of Biocompatibility Testing for Crimper

|  Test | Purpose | Results  |
| --- | --- | --- |
|  Cytotoxicity: Medium Eluate Method (MEM) | Determine whether test article extracts would cause cytotoxicity and cell lysis | Test article sample was non-cytotoxic. 0% cell lysis was observed with equivalent results to the negative control.  |
|  Sensitization: Guinea Pig Maximization | Investigate the potential for delayed dermal contact sensitization. | No irritation was present on any of the test or control animals at 24 or 48 hour readings using saline and vegetable oil extracts. Non-sensitizing.  |

P110021: FDA Summary of Safety and Effectiveness Data

|  Test | Purpose | Results  |
| --- | --- | --- |
|  Irritation/Intracutaneous Toxicity: Rabbit Intracutaneous Reactivity | Determine whether test article extracts would cause local dermal irritation or toxic effects | No evidence of irritation or abnormal effects over a 72 hour period as compared to negative controls.  |

## 9.2 SAPIEN Valve Hydrodynamic Performance

*In vitro* hydrodynamic performance studies of the SAPIEN Model 9000TFX bioprosthesis (test valve) were completed to evaluate performance under steady and pulsatile flow testing conditions. Valves were evaluated after nominal deployment and after deployment into irregular shapes (under deployed, oval deployed, and over deployed). The studies were conducted in accordance with the ISO 5840: Cardiovascular Implants-Cardiac Valve Prostheses (2005) standard. Reference articles for the nominally deployed SAPIEN valve studies consisted of commercially available aortic valves; reference articles for the irregular studies consisted of nominally deployed SAPIEN valves. A matrix of the tests performed and corresponding results is provided in Table 4.

|  Test | Purpose/Objective | Test/Reference Articles | Results  |
| --- | --- | --- | --- |
|  Steady Forward Flow | To determine the pressure drop at various steady forward flow rates. | Nominal
Test: Size 23mm & 26mm
Reference: Size 23mm & 27mm
Irregular
Test: Irregular SAPIEN size 23mm & 26mm
Reference: Nominal SAPIEN size 23mm &, 26mm | The SAPIEN valve offers acceptable hemodynamics with pressure gradients and effective orifice areas that are comparable to those offered by the reference valves.  |
|  Steady Backflow Leakage | To determine the leakage rate at various steady back flow pressures. | Nominal
Test: Size 23mm & 26mm
Reference: Size 23mm & 27mm
Irregular
Test: Irregular SAPIEN size 23mm & 26mm
Reference: Nominal SAPIEN size 23mm &, 26mm | The SAPIEN valve offers satisfactory performance in terms of its competency to prevent significant transvalvular aortic back-flow during the diastolic phase.  |

P110021: FDA Summary of Safety and Effectiveness Data

P110021: FDA Summary of Safety and Effectiveness Data
Page 11 of 55
17

|  Test | Purpose/Objective | Test/Reference Articles | Results  |
| --- | --- | --- | --- |
|  Pulsatile Flow Pressure Drop | To determine pressure drop and effective orifice area performance under pulsatile flow conditions. | Nominal
Test: Size 23mm & 26mm
Reference: Size 23mm & 27mm

Irregular
Test: Irregular SAPIEN size 23mm & 26mm

Reference: Nominal SAPIEN size 23mm &, 26mm | The SAPIEN valve offers acceptable hydrodynamics with a larger effective orifice area than those required by the ISO 5840:2005 acceptance criteria for aortic valves, and similar pressure drop to the reference valves.  |
|  Pulsatile Flow Regurgitation | To determine regurgitation performance under pulsatile flow conditions. | Nominal
Test: Size 23mm & 26mm
Reference: Size 23mm & 27mm

Irregular
Test: Irregular SAPIEN size 23mm & 26mm

Reference: Nominal SAPIEN size 23mm &, 26mm | The SAPIEN valve offers acceptable hydrodynamics with regurgitant fractions that were lower than those required by the ISO 5840:2005 acceptance criteria.  |
|  Flow Visualization | To qualitatively investigate flow characteristics in the vicinity of the valve. | Nominal
Test: Size 23mm & 26mm
Reference: Size 23mm & 27mm

Irregular
Test: Irregular SAPIEN size 23mm & 26mm

Reference: Nominal SAPIEN size 23mm &, 26mm | The SAPIEN valve offers acceptable aortic flow patterns throughout the entire cardiac cycle.

Broad central jet-like flows and no flow stasis during opening were observed in all SAPIEN valves, with no retrograde jet-like flow.  |
|  Verification of Bernoulli Relationship | To determine whether the Bernoulli relationship applies to clinical pressure drop measurements. | Nominal
Test: Size 23mm & 26mm
Reference: Size 23mm & 27mm

Irregular
Test: Irregular SAPIEN size 23mm & 26mm

Reference: Nominal SAPIEN size 23mm &, 26mm | Pressure drop results for the SAPIEN valve demonstrated correlation with the Bernoulli relationship.  |

# 9.3 SAPIEN Valve Structural Performance

In vitro structural performance studies of the SAPIEN Model 9000TFX were performed. Commercially available aortic valve replacements (AVR) and Cordis Palmaz Genesis stents were used as control articles in studies requiring concurrent testing of devices marketed in the U.S. A matrix of tests performed and corresponding results are provided in Table 5.

Table 5 - Structural Performance Evaluation

|  Test | Purpose/Objective | Test/Reference Articles | Results  |
| --- | --- | --- | --- |
|  Accelerated Wear | To assess long-term performance of the valve though accelerated wear. | Nominal
Test: Size 23mm & 26mm
Reference: Size 23mm & 27mm
Irregular
Test: Irregular SAPIEN size 23mm & 26mm
Reference: Nominal SAPIEN size 23mm &, 26mm | All valves survived durability testing to 200 million cycles in accelerated wear testers without excessive structural damage and/or functional impairment.

After testing to 200 million cycles, all valves met the minimum EOA and Total Regurgitation Fraction requirements of ISO 5840:2005.  |
|  Dynamic Failure Mode | To obtain information about the failure modes affecting the durability of the valve. | Test: Size 23mm & 26mm
Reference: Size 23mm & 27mm | All of the failures for both the test and reference valves occurred at pressures well beyond what would be experienced in vivo.  |
|  Frame Crush Resistance | To evaluate the resistance of the valve to lateral compressive loads. | SAPIEN frames size 23mm, SAPIEN frames size 26mm | Minimum force required to compress the frame was acceptable.  |
|  Frame Corrosion Resistance | To characterize the corrosion resistance of the valve frames and 5-hole bars in accordance with ASTMF2129-08 | Test: SAPIEN frames size 23mm, SAPIEN frames size 26mm, SAPIEN 5-hole bars
Reference: Cordis Palmaz Genesis stents | Corrosion resistance of SAPIEN frames and 5-Hole Bars are equivalent to the commercially available stent.  |
|  Frame Fatigue | To determine frame fatigue resistance to 600 million cycles. | SAPIEN frames size 23mm, SAPIEN frames size 26mm | No frame cracks or fractures observed at completion of 600 million cycles under 60x magnification.  |
|  Stress Analysis (FEA) | To characterize mechanical behavior of the frame during deployment and operation. | Modeling based on in vitro and clinical data of 23mm and 26mm SAPIEN frames. | Results indicate that the worst-case 26mm SAPIEN frame should not fracture for 600 million cycles, even under the unlikely simultaneous combination of all the worst-case conditions.  |

The following additional structural performance studies were completed with acceptable results: grain structure analysis, open circuit potential, material mechanical properties, fatigue life determination (i.e., Goodman diagram), force on commissure.

P110021: FDA Summary of Safety and Effectiveness Data

19

# 9.4 SAPIEN Valve Design Specific Performance Studies

Design specific *in vitro* performance studies of the SAPIEN Model 9000TFX were completed. The following studies were completed with acceptable results: percent surface area, frame overexpansion safety factor investigation, frame foreshortening and recoil, frame radial strength, valve migration force, pulsatile flow migration, radiopacity.

# 9.5 SAPIEN Valve Magnetic Resonance Imaging (MRI) Compatibility

Testing of this device in magnetic fields of 1.5 and 3.0 Tesla has shown that this device is MR Conditional. It can be scanned safely under the following conditions:

- Static magnetic field of 1.5 Tesla or 3 Tesla
- Spatial gradient field of 2500 Gauss/cm or less.
- Maximum whole-body-averaged specific absorption rate (SAR) of 2 W/kg for 15 minutes of scanning.
- Normal mode operation, as defined in IEC 60601-2-33, of the MR system.

# 9.6 Delivery System and Accessory Performance Testing

The following tests were performed for the RetroFlex 3 Delivery System and showed acceptable results: dimensional verification, visual inspection, simulated use, balloon characterization, bond strength, hemostasis, and migration.

The following tests were performed for the RetroFlex Balloon Catheter and showed acceptable results: dimensional verification, visual inspection, simulated use, balloon characterization, bond strength, and balloon compliance.

The following tests were performed for the Ascendra Balloon Catheter and showed acceptable results: dimensional verification, visual inspection, simulated use, balloon characterization, bond strength, hemostasis, deployed valve diameter, and migration.

The following tests were performed for the Ascendra Balloon Aortic Valvuloplasty Catheter and showed acceptable results: dimensional verification, visual inspection, simulated use, balloon characterization, and bond strength.

The following tests were performed for the Ascendra Introducer Sheath Set and showed acceptable results: dimensional verification, visual inspection, simulated use, bond strength, and hemostasis.

The following tests were performed for the Crimper and showed acceptable results: dimensional verification, visual inspection, and simulated use.

# 9.7 SAPIEN Valve Animal Studies

Feasibility studies were conducted in over 100 animals (porcine, bovine, canine, and ovine) in an attempt to identify a suitable animal model and study feasibility of percutaneous delivery of the valve. The valves used in these studies were either early prototypes (equine and bovine) or the Cribier-Edwards™ Aortic Bioprosthesis, Model 9000. A chronic study was performed on this model of the valve in which 19 juvenile sheep with induced aortic

P110021: FDA Summary of Safety and Effectiveness Data Page 13 of 55

insufficiency were treated. Fourteen (14) percutaneous implants of the 23mm Model 9000 valve were attempted in the proximal descending aorta and 5 sheep were treated surgically with the control article, a commercially available pericardial bioprosthesis. An overview of this study is provided in Table 6.

Table 6 - GLP Chronic Study Overview

|  Sample Size/Animal Model | 19 sheep with induced aortic insufficiency (Hufnagel Model)  |
| --- | --- |
|  Test Articles | Cribier-Edwards™ Aortic Bioprosthesis, Model 9000  |
|  Control Articles | Commercially available pericardial bioprosthesis  |
|  Technique | Percutaneous implant of valve and surgical implantation of control articles in the proximal descending aorta.  |
|  Results | 14 percutaneous implants attempted
10 successful animals (sacrificed between 10 - 21 weeks)
3 procedure related deaths
1 non-related early death
5 surgical implants – Control
3 procedural deaths
2 sacrificed within 48 hours due to valve issues  |
|  Conclusion | 6 animals survived to 21 weeks. The gross findings and histopathology results suggest that the valve is capable of long-term implant.  |

A chronic in vivo animal implantation study was conducted using the SAPIEN Valve, Model 9000TFX in the adult ovine model. A total of eighteen test article Model 9000TFX valves were implanted in the aortic position of 18 adult male sheep for a 10 week (n=9) and 20 week (n=9) evaluation study; 3 of 9 animals survived to at least 10 weeks and 6 of 9 survived to at least 20 weeks. Three (3) control articles were implanted in the aortic position of 3 adult male sheep; 2 control animals survived to at least 20 weeks and were clinically normal prior to explant; 1 animal survived to less than 14 days. No control valves were evaluated at 10 weeks. The results of this study indicate that the 9000TFX valve model has acceptable hemodynamic performance. Normal healing with pliable leaflets and no thrombus were observed, with no evidence of infection or calcification when implanted for 20 weeks. The two valve models were comparable for all parameters evaluated. A summary of the study results is provided in Table 7.

Table 7 - GLP Chronic Study Summary

|  Evaluation Parameter | Summary of Results  |
| --- | --- |
|  Clinical History and Hematology | All 10-week and six 20-week sheep were clinically normal prior to explants. At implant and explant, hematology was within normal limits for both groups. Clinical chemistry and complete blood count results were within normal limits for the majority of animals. Among the remaining animals, some values were either slightly above or below the reported normal range but none was considered to be clinically significant. Findings were comparable between both groups. Three test animals had elevated plasma free hemoglobin; this may have been due to red cell damage during sample collection as no clinical signs of hemolysis were observed.  |
|  Hemodynamic Performance | At 20 weeks, there were no differences from the average pre-explant peak gradients between the two groups for both normotensive and hypertensive readings, and no differences from the average post-implant and pre-explant cardiac outputs between the two groups.  |

P110021: FDA Summary of Safety and Effectiveness Data

|   | The six 20-week test valves had evidence of mild to moderate aortic valve insufficiency by echocardiography exams of paravalvular origin. One of two control valves had mild insufficiency.

Angiography evaluation at 20 weeks indicated that 4 of 6 test valves had Grade 1-2 regurgitation of undetermined origin. Two test valves had Grade 3-4 regurgitation with at least one for paravalvular origin. One control valve had Grade 3-4 regurgitation from undetermined location.  |
| --- | --- |
|  Histopathology | Histopathology results showed no apparent differences in tissue reactions (general healing, calcification, or morphology of the tissue/valve interface) between the test device and the control device. Tissue reactions towards the test and control devices were generally of low severity and were considered to be typical of this type of device implant.  |
|  Gross Observations | General healing results were comparable among the two valve models at 20 weeks. There were no differences between the gross observation valve findings for calcific deposits, thrombus formation, vegetative growths, leaflet damage, material wear, suture integrity, right dehiscence or frame fracture. Both groups presented with minimal to moderate valve leaflet host tissue overgrowth. Individual sheep from the test group had minimal leaflet retraction and minimal to moderate paravalvular spaces was observed for both groups.  |

## 9.8 Sterilization

The SAPIEN Valve Model 9000TFX is sterilized by terminal liquid sterilization (TLS) in buffered glutaraldehyde solution. The RetroFlex 3 Delivery System, RetroFlex Balloon Catheter, and Crimper are sterilized by ethylene oxide (EO). After sterilization, the devices are held in quarantine until sterility is verified per process specifications. The TLS and EO processes have demonstrated Sterility Assurance Levels (SAL) exceeding the industry standard of $10^{-6}$ in validation studies.

## 9.9 Shelf Life

Packaging and product integrity studies were conducted to ensure that the shelf life for each package and product is maintained for a minimum of two (2) years for the SAPIEN Valve, RetroFlex 3 Delivery System, RetroFlex Balloon Catheter, Ascendra Balloon Catheter, Ascendra Balloon Aortic Valvuloplasty Catheter, Ascendra Introducer Sheath Set, and Crimper.

## 9.10 Package Integrity

The packaging for the SAPIEN valve consists of a 3.8 oz jar, a lid and gasket closure system, and shelf and shipping containers. This system has been evaluated via physical testing and microbial challenge and was shown to maintain its sterile barrier following four years of real-time aging and exposure to temperature variations and simulated shipping conditions.

The RetroFlex 3 Delivery System, RetroFlex Balloon Catheter, Ascendra Balloon Catheter, Ascendra Balloon Aortic Valvuloplasty Catheter, Ascendra Introducer Sheath Set, and Crimper are packaged in Tyvek pouches and shelf and shipping cartons. These systems

P110021: FDA Summary of Safety and Effectiveness Data Page 15 of 55

have been evaluated and shown to maintain sterile barrier following two years of accelerated aging and exposure to temperature variations and simulated shipping conditions.

## 9.11 Product Integrity

### 9.11.1 SAPIEN Valve Biological Tissue

Edwards ThermaFix-processed bovine pericardial tissue has previously been validated and approved under PMA application P860057 regarding the Carpentier-Edwards® PERIMOUNT® Pericardial Bioprosthesis product family. The tissue used for the SAPIEN valve is identical to the tissue used on the PERIMOUNT valve. Biochemical evaluation was conducted on tissue stored in glutaraldehyde solution for four years real time. All device specifications were met for moisture content, ninhydrin content, shrinkage temperature, and enzymatic digestion of tissue.

Histological examination of leaflets was conducted on leaflet samples from whole valves at zero-time and after two years of real-time aging. Results demonstrated that aging of tissue does not appear to impact the microstructure of bovine pericardial tissue used in the SAPIEN valve. A stress relaxation study was completed to compare cyclic load decay for tissue leaflet samples at zero-time to tissue leaflets at zero-time and after three years of real-time aging. No statistically significant difference was observed between groups.

### 9.11.2 SAPIEN Valve Non-biological Components and Whole Valve Testing

Functionality of the SAPIEN valve’s non-biologic components (polymers: valve holder, skirt, sleeve, and sutures; and metallics: frame and frame samples) and whole-valve hydrodynamic and wear testing were completed after 2 years real-time aging.

Tensile testing of the frame met acceptance criteria. Corrosion resistance of the frame demonstrated higher resistance than the zero-time reference. Tensile testing of all polymer components met acceptance criteria relative to zero-time reference strengths. All valves passed the minimum hydrodynamic performance requirements for EOA and Regurgitant Fraction per ISO 5840:2005. The 2 year real-time aged SAPIEN valves survived durability testing out to 200 million cycles in accelerated wear testers under aortic pressure test conditions without failure, significant tissue wear or frame deformation and fracture. These valves offered a larger EOA and lower regurgitant fractions than those required per the minimum performance requirements of ISO 5840:2005 after 200 million cycles.

### 9.11.3 Delivery System and Accessories

Functionality and product integrity of the RetroFlex 3 Delivery System, RetroFlex Balloon Catheter, Ascendra Balloon Catheter, Ascendra Balloon Aortic Valvuloplasty Catheter, Ascendra Introducer Sheath Set and Crimper was demonstrated after following two years of accelerated aging and exposure to temperature variations and simulated shipping conditions.

## 10. SUMMARY OF PRIMARY CLINICAL STUDY (G030069)

The applicant conducted a clinical study to establish a reasonable assurance of safety and effectiveness of transcatheter aortic valve replacement with the SAPIEN Transcatheter

P110021: FDA Summary of Safety and Effectiveness Data
Page 16 of 55

Heart Valve for transfemoral or transapical delivery in patients with severe symptomatic native aortic valve stenosis who have been determined by a cardiac surgeon to be at high risk for open aortic valve replacement and in whom existing co-morbidities would not preclude the expected benefit from correction of the aortic stenosis in the U.S., Canada and Germany under IDE # G030069. Data from this clinical study were the basis for the PMA approval decision. A summary of the clinical study is presented below.

## 10.1 Study Design

The PARTNER trial was a prospective, unblinded, randomized, controlled, multi-center pivotal trial evaluating the safety and effectiveness of the Edwards SAPIEN THV, via transfemoral (Cohort A and Cohort B) or transapical (Cohort A only) delivery. Because the study enrolled two distinct populations, the two Cohorts were separately-powered and analyzed. As depicted in the diagram below, an initial stratification based on operability for aortic valve replacement (AVR) surgery was used to assign the patients to Cohort A or B. Assignment to Cohorts was followed by determination of the possibility of vascular access for transfemoral delivery. Patients who were considered high surgical risk and eligible for transfemoral access were stratified into Cohort A and randomized to treatment (transfemoral AVR) or control (surgical AVR). Cohort A patients who were not eligible for transfemoral access were evaluated as candidates for transapical delivery and, if appropriate, randomized to treatment (transapical AVR) or control (surgical AVR). Those patients who were considered non-surgical candidates were stratified into Cohort B and randomized to treatment (transfemoral AVR) or control ("standard" therapy). Those assigned to Cohort B who did not meet the criteria for transfemoral delivery were not enrolled in the study because the sponsor declined to have a transapical arm in Cohort B. This PMA relates to only the Cohort A study.

Figure 8 - PARTNER Trial Enrollment
![img-7.jpeg](img-7.jpeg)
AVR=aortic valve replacement surgery, TA=transapical. TAVR=transcatheter aortic valve replacement, TF=transfemoral.

P110021: FDA Summary of Safety and Effectiveness Data

A total of 1057 subjects were enrolled at 27 sites in the PARTNER study in the two arms – 699 patients in Cohort A (transfemoral or transapical insertion of the SAPIEN compared to surgical AVR); 358 patients in Cohort B (transfemoral insertion of the SAPIEN versus “standard” therapy in an inoperable population).

The protocol was fully approved in March 2009 (Version 3.2), a few months before enrollment in Cohort A was complete (August 2009). In order to allow for continued access to a device when there may be a gap between trial completion and final regulatory review, additional patients who are still subject to the same patient protection measures as the IDE trial are enrolled under the “Continued Access Protocol (CAP).” These patients are enrolled into an “extension” of the initially approved sample size. The CAP was approved on August 13, 2009 for enrollment of Cohort A subjects in a non-randomized study.

The statistical analysis plan (SAP) included in Protocol Version 3.2 was finalized in March 2009. For this Cohort A study, FDA reviewed and assessed a dataset of events through September 21, 2011.

## 10.2 Patient Selection Process and Enrollment Criteria

The methodology for assessing patient risk incorporated the Society of Thoracic Surgeons (STS) risk calculator, and in addition, incorporated a minimum of two experienced surgeons and a cardiologist to make the initial high risk decision, taking into account risk factors not evaluated by the STS risk calculator. This decision was then peer reviewed on routine case review conference calls.

The major inclusion and exclusion criteria for the Cohort A study are summarized below.

## 10.2.1 Inclusion Criteria

The major inclusion criteria for patient entry into the study included the following:

1) Patient has senile degenerative aortic valve stenosis with echocardiographically derived criteria: mean gradient >40 mmHg or jet velocity greater than 4.0 m/s or an initial aortic valve area (AVA) of < 0.8 cm² (indexed effective orifice area [EOA] < 0.5 cm²/m²). (Qualifying AVA baseline measurement must be within 45 days prior to randomization).

2) Patient is symptomatic from his/her aortic valve stenosis, as demonstrated by New York Heart Association (NYHA) Functional Class II or greater.

3) Patients must have co-morbidities such that the surgeon and cardiologist Co-PIs concur that the predicted risk of operative mortality is ≥15% and/or a minimum STS score of 10. A candidate who does not meet the STS score criteria of ≥10 can be included in the study if a peer review by at least two surgeon investigators (not including the enrolling surgeon) concludes and documents that the patient’s predicted risk of operative mortality is ≥15%. The surgeon's assessment of operative comorbidities not captured by the STS score must be documented in the study case report form as well as in the patient medical record.

P110021: FDA Summary of Safety and Effectiveness Data

P110021: FDA Summary of Safety and Effectiveness Data
Page 19 of 55

## 10.2.2 Exclusion Criteria

The major exclusion criteria for patient entry into the study included the following:

1) Evidence of an acute myocardial infarction (MI) ≤ 1 month before the intended treatment (defined as: Q wave MI, or non-Q wave MI with total CK elevation of CK-MB ≥ twice normal in the presence of MB elevation and/or troponin level elevation (WHO definition).

2) Mixed aortic valve disease (aortic stenosis and aortic regurgitation with predominant aortic regurgitation >3+).

3) Any therapeutic invasive cardiac procedure performed within 30 days of the index procedure, (or 6 months if the procedure was a drug eluting coronary stent implantation).

4) Pre-existing prosthetic heart valve in any position, prosthetic ring, severe mitral annular calcification (MAC), severe (greater than 3+) mitral insufficiency, or Gorlin syndrome

5) Need for emergency surgery for any reason.

6) Native aortic annulus size < 18mm or >25mm as measured by echocardiogram.

7) Patient has been offered surgery but has refused surgery.

8) Recent (within 6 months) cerebrovascular accident (CVA) or a transient ischemic attack (TIA).

There was not a specific exclusion criterion for patients with critical aortic stenosis who had co-morbid conditions limiting the length or quality of their life. This was an abbreviated listing of the main inclusion and exclusion criteria; there were a total of 29 entrance criteria for the subjects in this study.

## 10.2.3 Primary Safety and Effectiveness Endpoint

The primary effectiveness and safety endpoint for Cohort A was freedom from all cause mortality at exactly day 365, analyzed in the ITT population.

The hypotheses for the primary endpoint are:

$$
H _ {0}: S _ {T} (T) - S _ {C} (T) \leq - 0. 0 7 5
$$

$$
H _ {A}: S _ {T} (T) - S _ {C} (T) > - 0. 0 7 5
$$

where $S_{T}(T)$ is the freedom from all cause mortality at exactly day 365 for treatment arm and $S_{C}(T)$ is that for control arm.

The test statistic is $\frac{\hat{S}_{(T)}(T) - \hat{S}_{(C)}(T) + 0.075}{\sqrt{\hat{V}[\hat{S}_{(T)}(T)] + \hat{V}[\hat{S}_{(C)}(T)]}}$, where $\hat{S}_{(T)}(T)$ and $\hat{S}_{(C)}(T)$ are the survival rates estimated by the Kaplan-Meier algorithm, and $\hat{V}[\hat{S}_{(T)}(T)]$ and $\hat{V}[\hat{S}_{(C)}(T)]$ are the

Page 20 of 55

variances estimated by Greenwood's formula. The null hypothesis will be rejected, and non-inferiority concluded, if the test statistic is greater than 1.645.

In addition to formal analysis of non-inferiority endpoints, the Kaplan-Meier (KM) curves will be presented for each group in the analysis, and a 95% two-sided confidence interval for the difference of the curves will be shown.

## 10.2.4 Secondary Safety and Effectiveness Endpoints

This section discusses pre-specified secondary safety and effectiveness endpoints as well as endpoints that are more clinically relevant. Both are described below.

## 10.2.4.1 FDA Secondary Endpoints

The following serious adverse event endpoints are considered to be clinically important, and should be interpreted in the context of the totality of data demonstrating safety and effectiveness of SAPIEN. Specifically the following adverse events were assessed at 1 year and will be presented in this summary:

- Deaths;
- Neurological Events;
- Aortic Regurgitation;
- Bleeding;
- Vascular Complications; and
- Atrial Fibrillation.

## 10.2.4.2 Pre-Specified Secondary Endpoints

The following, selected, pre-specified secondary endpoints are also interesting to note.

- Time from randomization to the first occurrence of a Major Adverse Cardiac and Cerebrovascular Event (MACCE) within 1 year. The MACCE definition included:
- Death
- Myocardial infarction (MI)
- Stroke
- Renal failure
- Total hospital days from the index procedure to one year post procedure
- NYHA functional classification at 1 year
- Six minute walk test at 1 year

## 10.2.4.3 Other Secondary Endpoints of Interest:

1) Separate analyses of the primary endpoint in the transapical and transfemoral groups.
2) Functional improvement from baseline as measured per a) NYHA functional classification, b) effective orifice area (EOA) and c) six minute walk test at 30 days, six months and one year.
3) Freedom from MACCE at 30 days, 6 and 12 months. MACCE definition includes death, MI, stroke and renal failure.

P110021: FDA Summary of Safety and Effectiveness Data

4) Evidence of prosthetic valve dysfunction (hemolysis, infection, thrombosis, severe paravalvular leak or migration) at 30 days, 6 and 12 months.
5) Length of index hospital stay (ITT).
6) Total hospital days from the index procedure to one year post procedure (ITT).
7) Improved Quality of Life (QOL) from baseline at 30 days, 6 and 12 months (ITT).
8) Improved valve function demonstrated by a responder analysis showing the percentage of patients in each treatment group who have a greater than 50% improvement in AVA/EOA at 30 days, 6 and 12 months.

### 10.2.4.4 Other Adjunctive Analyses

In addition to the pre-specified primary endpoint at one year and several secondary endpoints evaluated at 30 days, 6 months, and/or 1 year, longer-term data are also presented. As part of the additional adjunctive analyses, 2 year data for mortality and major adverse cardiac and cerebrovascular events and findings related to the CAP cohort are included in this summary.

### 10.2.4.5 Comparison of Results to Sample Size Estimation

In calculating the sample size needed for the study, it was assumed that 65% of the patients would have the transfemoral approach (actual 70.4%). It was further estimated that the transfemoral patients would have improved 12-month mortality for the SAPIEN (25%) versus open AVR (30%). The study indicates a 12-month mortality of 22.24% for SAPIEN and 26.36% for open AVR on the transfemoral approach. It was assumed that for the transapical approach, the 12-month mortality would be 35% for both transapical TAVI and open AVR. The study indicates a 12-month mortality of 29.04% for transapical TAVI and 27.86% for open AVR.

### 10.3 Accountability of PMA Cohort

The study results are presented based on two populations: Intent-To-Treat (ITT) and As Treated (AT). There is also a third population, the Valve Implant population, consisting of those patients who received the valve. A summary of the patient populations is provided in the table below.

|   | Intent-to-Treat (ITT) | As Treated (AT) | Valve Implant  |
| --- | --- | --- | --- |
|  Treatment TAVI | n=348 | n=344 | n=326  |
|  Control AVR | n=351 | n=313 | n=311  |
|  Total | n=699 | n=657 | n=637  |

### 10.3.1 Intent-To-Treat Population

Of the 699 patients in the high risk, Cohort A, 348 were assigned to TAVI (SAPIEN) treatment group (244 of whom were implanted via the transfemoral route, and 104 of whom were implanted via the transapical approach), 351 were randomized into the AVR (control) group (248 of whom were eligible for transfemoral and 103 of whom were

P110021: FDA Summary of Safety and Effectiveness Data
Page 21 of 55

eligible for transapical), forming the Intent To Treat (ITT) population, defined as all randomized patients.

## 10.3.2 As Treated Population

The As Treated (AT) population was based on the treatment actually received. Therefore, the As Treated population is defined as follows:

- **AT SAPIEN**: This population consists of the Cohort A patients randomized to the treatment arm for whom the study valve implant procedure is begun, and the day of implant is considered day 0 for these patients. The definition of “procedure is begun” is “the time the study catheter is placed in the patient in the catheterization laboratory.”

Four patients did not have an attempt at the procedure (i.e. ITT=348; AT=344)

If a treatment patient in Cohort A was assigned to the transfemoral approach, and it was determined during further access evaluation that the transapical approach was needed, that patient was considered to be a transapical patient for the as treated analyses of implant subgroups. This did not impact the combined Cohort A analysis.

- **AT Control**: This population consists of the Cohort A patients randomized to the Control arm for whom the valve implant procedure was begun. The day of implant was considered day 0 for these patients. The definition of “procedure is begun” was “the induction of general anesthesia for the open operation.” A total of n=38 patients were to have received a control valve, but did not (i.e., ITT = 351; AT = 313)

## 10.3.3 Valve Implant Population

The valve implant population is defined as the subset of the As Treated population consisting of those patients (Treatment or Control) for whom the valve was implanted and remained in position.

Among the AT patients, 18 TAVI patients did not have the valve in position at the end of 1 year. Thus, the valve implant population includes 326 patients in TAVI arm. Two AVR patients did not have the valve implanted.

This summary presents the data using the most appropriate treatment population for each particular analysis.

## 10.4 Study Population Demographics and Baseline Parameters

### 10.4.1 Baseline Demographics

The table below summarizes the baseline demographics for each group.

P110021: FDA Summary of Safety and Effectiveness Data

Table 9 - Patient Baseline Demographics

|  Characteristic | TAVI (SAPIEN)
N=348 | AVR (Control)
N=351 | P-value  |
| --- | --- | --- | --- |
|  Age (yr), mean±SD | 83.6±6.8 | 84.5±6.4 | 0.07  |
|  Male sex, n (%) | 201/348 (57.8) | 198/351 (56.4) | 0.82  |
|  STS score^{b}, mean±SD | 11.8±3.3 | 11.7±3.5 | 0.61  |
|  NYHA (New York Heart Association) class, n/total n (%): |  |  | 0.79  |
|  II | 20/348 (5.7) | 21/349 (6.0) |   |
|  III or IV | 328/348 (94.3) | 328/349 (94.0) |   |
|  Coronary artery disease, n/total n (%) | 260/347 (74.9) | 266/346 (76.2) | 0.66  |
|  Previous myocardial infarction, n/total n (%) | 92/347 (26.5) | 103/346 (29.8) | 0.35  |
|  Previous intervention, n/total n/total n (%) |  |  |   |
|  CABG (coronary artery bypass grafting) | 148/348 (42.5) | 152/349 (43.6) | 0.82  |
|  PCI (percutaneous coronary intervention) | 116/346 (33.5) | 110/348 (31.6) | 0.63  |
|  Balloon aortic valvuloplasty | 46/348 (13.2) | 35/349 (10.0) | 0.20  |
|  Cerebral vascular disease, n/total n (%) | 96/327 (29.4) | 87/325 (26.8) | 0.49  |
|  Peripheral vascular disease, n/total n (%) | 149/345 (43.2) | 142/341 (41.6) | 0.70  |
|  COPD (chronic obstructive pulmonary disease), n (%): |  |  |   |
|  Any | 152/348 (43.7) | 151/351 (43.0) | 0.88  |
|  Oxygen-dependent | 38/220 (17.3) | 38/229 (16.6) | 0.90  |
|  Creatinine >2 mg/dl (177 μmol/liter), n/total n (%)^{&} | 37/343 (10.8) | 22/344 (6.4) | 0.04  |
|  Atrial fibrillation, n/total n (%) | 81/199 (40.7) | 75/172 (43.6) | 0.60  |
|  Permanent pacemaker, n/total n (%) | 69/348 (19.8) | 76/349 (21.8) | 0.58  |
|  Pulmonary hypertension, n/total n (%) | 125/295 (42.7) | 111/302 (36.8) | 0.15  |
|  Extensively calcified aorta, n (%) | 2/348 (0.60) | 4/351 (1.1) | 0.69  |
|  Deleterious effects of chest-wall irradiation, n (%) | 3/348 (0.9) | 3/351 (0.9) | 1.00  |
|  Chest-wall deformity, n (%) | 0/348 (0.0) | 1/351 (0.3) | 1.00  |
|  Frailty** | 46/295 (15.6) | 53/301 (17.6) | 0.58  |
|  Liver Disease, n/total n (%) | 8/348 (2.3) | 11/349 (3.2) | 0.64  |
|  Echocardiographic findings |  |  |   |
|  Aortic valve area – cm^{2} | 0.7±0.2 | 0.6±0.2 | 0.11  |
|  Mean aortic valve gradient – mm Hg | 42.6±14.6 | 43.5±14.3 | 0.42  |
|  Mean LVEF - % | 52.5±13.5 | 53.3±12.8 | 0.59  |
|  Moderate or severe MR – n/total n (%)^{a} | 66/337 (19.6) | 71/338 (21.0) | 0.70  |

** Frailty was subjectively determined by surgeons using prespecified criteria for purposes of frailty score validation; see Section 10.8.14 for additional details
& To convert creatinine to micromoles/liter, multiply by 88.4.
Moderate to severe regurgitation was defined as regurgitation of grade 3+ or higher
The Society of Thoracic Surgeons (STS) score measures patient risk at the time of cardiovascular surgery on a scale that ranges from 0% to 100%, with higher numbers indicating greater risk. An STS score higher than 10% indicates very high surgical risk.

Note that approximately 43% of the patients had a prior CABG, 10-13% had a prior balloon aortic valvuloplasty, 20% had a pacemaker, and 41-43% of patients had atrial fibrillation. The majority of the patients had been hospitalized for aortic stenosis in the past.

## 10.4.2 Operative Risk

The STS score predicted 11.7% for the 30-day mortality for the average surgeon at the average hospital. The Kaplan Meier (KM) 30-day mortality for the As Treated surgical AVR control was 8.0%. Therefore, the observed/expected ratio for the surgeons in this trial was 0.68 – indicating much better than average surgeons.

P110021: FDA Summary of Safety and Effectiveness Data

P110021: FDA Summary of Safety and Effectiveness Data
Page 24 of 55

## 10.4.3 Patient Selection

### 10.4.3.1 Variations in patient selection

An enrollment screening log was kept at each investigative site. Overall, 27% of the patients screened in the trial were enrolled. The ratio of the number of patients screened to those excluded varied among the sites, as described in the table below.

|  Site | Ratio Screen failure/total screened | %  |
| --- | --- | --- |
|  01 | 36/266 | 14%  |
|  18 | 58/146 | 40%  |
|  20 | 84/191 | 44%  |
|  09 | 251/403 | 62%  |
|  15 | 181/289 | 63%  |

### 10.4.3.2 Variations in site enrollment ratios of inoperable to high risk

There was a 3.4 fold variation in the enrollment ratio of transapical (TA) to transfemoral (TF), and a 4.3 fold variation in the ratio of “high risk” cohort A to “inoperable” Cohort B subjects between the sites, as depicted in the table below, which tabulates the ratios for the 6 highest enrolling sites.

|  Site Number | Cohort A Randomized TA Patients | Cohort A Randomized TF Patients | TA/TF Ratio | Cohort A Randomized PMA Patients | Cohort B Randomized PMA Patients | Cohort A/ Cohort B Ratio  |
| --- | --- | --- | --- | --- | --- | --- |
|  01 | 40 | 55 | 0.73 | 95 | 21 | 4.52  |
|  02 | 25 | 72 | 0.35 | 97 | 33 | 2.94  |
|  04 | 22 | 25 | 0.88 | 47 | 45 | 1.04  |
|  08 | 29 | 38 | 0.76 | 67 | 43 | 1.56  |
|  09 | 23 | 29 | 0.79 | 52 | 21 | 2.48  |
|  10 | 24 | 92 | 0.26 | 116 | 36 | 3.22  |

## 10.5 Primary Safety and Effectiveness Endpoint Results

The following section focuses on an analysis of the primary safety and effectiveness endpoint which evaluates freedom from mortality at one year. There are also discussions of the gender analysis for the primary endpoint, and a review of the differences in mortality between the transfemoral and transapical groups.

### 10.5.1 Results of Primary Endpoint - Freedom from All Cause Mortality at One Year

At the end of 1 year, there were 84 (out of 348) and 89 (out of 351) deaths in the TAVI and AVR arm in the ITT population, respectively. The Kaplan-Meier estimates of the all-cause mortality rate at 1 year are stated to be 24.27% and 26.80% for the TAVI

(treatment) and AVR (control) arm, respectively. The survival difference (TAVI-AVR) was 0.0253, and the 95% one-sided lower confidence limit (CL) for the difference was -0.0299, which is greater than the pre-defined non-inferiority margin (-0.075). The p-value for the non-inferiority test is 0.0014, indicating that the primary endpoint is met with a 0.075 non-inferiority margin.

In addition to the 1 year data, patient outcomes at 2 years are also presented, allowing for an assessment of longer-term results of SAPIEN THV implantation. The Kaplan-Meier cumulative incidence curve for the all-cause mortality to two years is shown below for the ITT population. The Kaplan-Meier estimates of the all-cause mortality rate at 2 years are stated to be 41.3% and 35.5% for the TAVI (treatment) and AVR (control) arm, respectively.

![img-8.jpeg](img-8.jpeg)
Figure 9 - Kaplan-Meier Cumulative Incidence Curve for All-Cause Mortality (ITT Population)

Based upon these data, there is no significant difference in mortality between the treatment and control groups at the 1 year endpoint. A careful review of the death narratives for this study did not raise any specific concerns regarding the causes of death in this study.

## 10.5.2 Analysis of AT Population

For the AT population, at the end of 1 year, there were 81 (out of 344) and 78 (out of 313) deaths in the TAVI and AVR arm, respectively. The Kaplan-Meier estimates of the all cause mortality rate at 1 year are stated to be 23.7% and 25.2% for the TAVI (treatment) and AVR (control) arm, respectively. The survival difference (TAVI-AVR) was -1.5, and the 95% one-sided lower CL for the difference was -0.004, which is greater than the pre-defined non-inferiority margin of -7.5%. The p-value for the non-inferiority test is 0.0037, indicating that the primary endpoint is met with a 7.5% non-inferiority margin on the AT population.

P110021: FDA Summary of Safety and Effectiveness Data

10.5.3 Site Poolability for the Primary Endpoint

Center effect on the primary endpoint was evaluated by the sponsor using Cox regression. Using Site 01 as the reference group, hazard ratios of different sites over the reference group were reported for ITT population and for AT population. Except for one center (Site 15), all other 95% CI of center hazard ratios include 1. Site 15 contributes 25/699 = 3.58% of the ITT subjects in the database and 20/657 = 3.0% of the AT subjects.

A logistic regression model containing treatment, site, and treatment by site interaction is performed on all-cause mortality as well as on MACCE. No significant interaction is detected on either endpoint (p-value > 0.15).

10.5.4 Gender Differences for Primary Endpoint

A post hoc analysis was conducted to compare mortality between genders. The study was not powered for each gender separately.

In the ITT population, males composed 57.8% (201/348) of TAVI arm and 56.7% (198/351) of AVR. In the AT population, males were 57.6% (198/344) of TAVI arm and 57.2% (179/313) of AVR.

In both ITT and AT populations, males performed better with AVR. All-cause mortality was numerically higher in the TAVI arm than that in the AVR arm. The mortality rates at 1 year are 28.52% and 25.21% for TAVI and AVR, respectively, in the ITT population. The mortality rates at 1 year are 27.44% and 22.67% for TAVI and AVR, respectively, in the AT population. The 95% one-sided lower confidence limits of survival difference (TAVI-AVR) are -10.69% and -12.14% for ITT and AT, respectively. Both are less than the pre-specified non-inferiority margin (-7.5%).

In both ITT and AT populations, females perform better with TAVI. All-cause mortality was numerically higher in the AVR arm than that in the TAVI arm. The mortality rates at 1 year are 18.45% and 29.03% for TAVI and AVR, respectively, in the ITT population. The mortality rates at 1 year are 18.58% and 28.56% for TAVI and AVR, respectively, in the AT population. The 95% one-sided lower confidence limits of survival difference (TAVI-AVR) are 2.36% and 1.64% for ITT and AT, respectively. Both are greater than the pre-specified non-inferiority margin (-7.5%).

Table 12 - One-Year All-Cause Mortality in Males Vs. Females

|  Group | Intent to Treat |   |   | As Treated  |   |   |
| --- | --- | --- | --- | --- | --- | --- |
|   |  TAVI | AVR | 95% LCL* | TAVI | AVR | 95% LCL*  |
|  Male | 28.52% | 25.21% | -10.69% | 27.44% | 22.67% | -12.14%  |
|  Female | 18.45% | 29.03% | 2.36% | 18.58% | 28.56% | 1.64%  |

* Pre-specified non-inferiority margin -7.5%

In the continued access protocol (CAP) cohort, 1588 patients were enrolled in the TAVI registry (since randomization was eliminated for the CAP cohort) and 770 of them are female (48.5%). At one year, the K-M estimated event rates in ITT population are

P110021: FDA Summary of Safety and Effectiveness Data

18.54% for females and 25.94% for males, respectively. Those numbers are numerically close to those observed in the randomized study (18.45% and 28.52%, respectively).

Although this study was not powered for each gender separately, it appears that treatment effects are in the opposite direction for males versus females.

## 10.5.5 Transfemoral and Transapical Approaches

Though the interaction of treatment and approach (transfemoral versus transapical) is tested and is found to be not significant (p-value > 0.15), separate analyses of the primary endpoint in the transapical and transfemoral groups are of interest and are presented here for both the ITT and AT groups.

## 10.5.5.1 Transfemoral Approach

In the ITT population, for the transfemoral approach, there were 244 patients and 248 patients in the treatment and control groups, respectively. For the all cause mortality, the KM event rates at 1 year are 22.24% and 26.36% for the transfemoral treatment group and control group, respectively. The survival difference is 4.12% (Transfemoral-Control). The 95% one-sided lower CL for the survival difference is -2.34%.

In the AT population, for the transfemoral approach, there are 240 patients and 221 patients in the treatment and control groups, respectively. For the all cause mortality, the KM event rates at 1 year are 21.35% and 25.18% for the transfemoral treatment group and control group, respectively. The survival difference is 3.83% (Transfemoral-Control). The 95% one-sided lower CL for the difference is -2.68%.

## 10.5.5.2 Transapical Approach

In the ITT population, for the transapical approach, there were 104 patients and 103 patients in the treatment and control group, respectively. For the all cause mortality, the KM event rates at 1 year are 29.04% and 27.86% for the transapical treatment group and control group, respectively. The survival difference is -1.18% (Transapical-Control). The 95% one-sided lower CL for the difference is -11.69%.

In the AT population, for the transapical approach, there are 104 patients and 92 patients in the treatment and control group, respectively. For the all cause mortality, the KM event rates at 1 year are 29.07% and 25.28% for the transapical treatment group and control group, respectively. The survival difference is -3.79% (Transapical-Control). The 95% one-sided lower CL for the difference is -14.29%.

The mortality rates are numerically higher in the treatment group for the transapical approach.

There were limited patients in the transapical arm of the IDE randomized clinical trial (RCT). The final assessment of transapical delivery is presented later where the totality of the IDE RCT and CAP data are assessed.

P110021: FDA Summary of Safety and Effectiveness Data
Page 27 of 55

Table 13 - One-Year All-Cause Mortality in Transfemoral Vs. Transapical

|  Group | Intent to Treat |   |   | As Treated  |   |   |
| --- | --- | --- | --- | --- | --- | --- |
|   |  TAVI | AVR | 95% LCL | TAVI | AVR | 95% LCL  |
|  Transfemoral | 22.24% | 26.36% | -2.34% | 21.35% | 25.18% | -2.68%  |
|  Transapical | 29.04% | 27.86% | -11.69% | 29.07% | 25.28% | -14.29%  |

## 10.6 Limitations of Interpretation of Study Results

The interpretation of the results of this study is not without limitations. The following sections discuss special considerations to note when interpreting the data. The impact of these limitations on overall data interpretation is unknown.

Sensitivity Analyses for ITT Population

The Sponsor performed a worst case analysis to assess the robustness of the mortality results. The assumption used in the worst case analysis was that AVR patients who were censored prior to 1 year were considered alive at 1 year, and AVR patients who did not receive the procedure were also considered alive at 1 year; and that TAVI patients who were censored prior to 1 year were considered dead as of the censoring date, and TAVI patients who did not receive the procedure were also considered dead at 1 year.

The primary endpoint of all-cause mortality is still met with a 0.075 non-inferiority margin on the worst case analysis.

Although the primary endpoint was met, issues related to potential selection bias and other study limitations described below should be considered when interpreting these results.

## Patient Treatment

The following section highlights FDA’s interpretation of data related to patient treatment.

### 10.6.1 Heterogeneity of Treatment

This trial was designed to compare isolated AVR to TAVI. However, a review of the data resulted in a comparison that includes a heterogeneous group of patients and a combination of therapies as shown in the figure below.

|  CONTROL | TREATMENT  |
| --- | --- |
|  • AVR | • TAVI  |
|  • AVR + concomitant operations | • Attempted TAVI then AVR  |
|  • No AVR – no intervention | • No TAVI - nothing  |
|  • Delayed AVR | • TAVI - no intervention for concomitant conditions  |
|  • TAVI |   |

Figure 10 - Heterogeneity of Treatment

P110021: FDA Summary of Safety and Effectiveness Data

Due to the heterogeneity of treatment control group, interpretability of any differences between patient groups is limited. The following sections more fully described the following:

1. Failure to attempt to treat;
2. Delay in treatment; and
3. 3.Concomitant operations.

Results are confounded by the issue of AVR patients not receiving AVR, TAVI patients receiving AVR, and AVR patients undergoing concomitant operations.

## 10.6.2 Failure to Attempt to Treat

### 10.6.2.1 Imbalance in Failure to Attempt to Treat

There was an imbalance between the control and treatment groups as to those patients who refused/withdrew (28 fold increase in AVR group), patients who died before the procedure (2.3 fold increase in AVR group), and those judged to have deteriorated (5 fold increase in AVR group). The table below demonstrates the issue of differential numbers of failure to treat.

|  Reason | TAVI (N=348) | AVR (N=351)  |
| --- | --- | --- |
|  Died before the procedure | 2 (0.6%) | 5 (1.4%)  |
|  Refusal | 1 (0.3%) | 17 (4.8%)  |
|  Withdrawal | 0 (0%) | 11 (3.1%)  |
|  Pre-treatment deterioration | 1 (0.3%) | 5 (1.4%)  |
|  Total | 4 (1.1%) | 38 (10.8%)  |

Only 3/28 of the patients who refused AVR or withdrew from the study were known to be dead at one year (one patient refused AVR because she “started feeling better”).

It cannot be assumed that the sickness of the patients who chose to withdraw from the study was the same as those who were treated. If less sick patients differentially withdrew from the AVR arm, this could possibly bias results towards the treatment group in an Intent-to-Treat analysis. The imbalance between reasons for failing to treat patients has the potential of introducing selection bias in both the ITT and As Treated (AT) analyses.

This phenomenon of failure to treat in the control group occurred differently at various sites, as is shown in the table below.

|  Site | 10 | 02 | 08 | 04 | 15  |
| --- | --- | --- | --- | --- | --- |
|  Total control pts enrolled at site | 116 | 97 | 67 | 47 | 25  |
|  # control pts not getting AVR | 1 | 5 | 6 | 5 | 5  |
|  % pts not getting control AVR | 1% | 5.1% | 9.0% | 10.6% | 20%  |

P110021: FDA Summary of Safety and Effectiveness Data

Almost 11% of the patients did not get the assigned treatment in the AVR group.

Because these patients had critical aortic stenosis, it was expected that they would be treated per the group to which they were assigned and be indicated for isolated valve replacement. The trial results are confounded as a result of failing to treat these patients, possibly indicating a biased result towards worse outcomes in the ITT AVR group, because some patients did not receive the recommended treatment for their disease and the issue of concomitant operations in the AVR group. This could also bias results of the As Treated analysis against the AVR group if those patients not treated were “less sick” and therefore were excluded from the As Treated analysis, or if those patients had concomitant operations. These data need to be interpreted carefully since patient treatment across sites was not homogenous due to the large variation in the rates of Failure to Treat among sites.

## 10.6.2.2 Imbalance in Failed Treatment

In the TAVI group, a total of n=18 SAPIEN patients were excluded because either the SAPIEN was never implanted or did not remain *in situ* at the end of the index procedure, as detailed in the following table.

|  Reasons for Unsuccessful TAVI | Status | n  |
| --- | --- | --- |
|  Valve embolization | Did not remain *in situ* | 5  |
|  TEE findings | Not implanted | 5  |
|  Access problems | Not implanted | 4  |
|  Died prior to valve deployment | Not implanted | 2  |
|  Femoral artery tear | Not implanted | 1  |
|  Large sigmoid septum | Not implanted | 1  |
|  Total |   | 18  |

In the control AVR group two patients were operated on but did not receive a valve. One patient had a severely calcified aorta and subsequently underwent TAVI (alive at one year) and another needed a reoperation and died during the procedure.

There is an imbalance in patients who had attempted treatment that did not result in an implanted valve. Similar to earlier comments, the impact of these events on overall data interpretability is unknown.

## 10.6.2.3 Cross-Over - Use of AVR in TAVI arm

There were a total of 11 AVR procedures performed in patients randomized to TAVI, several of which were emergency procedures. These patients are summarized below:

i. Not implanted because of congenital septal condition
ii. Annulus 26mm, converted to AVR
iii. Annulus >25mm, converted to AVR done
iv. SAPIEN embolized to LV, emergency AVR (multiple complications)

P110021: FDA Summary of Safety and Effectiveness Data Page 30 of 55

v. SAPIEN embolized to LV, emergency AVR (multiple complications)
vi. SAPIEN embolized to LV, emergency AVR (multiple complications)
vii. Annulus 27mm, converted to AVR done
viii. SAPIEN embolized to LV, emergency AVR (patient died)
ix. SAPIEN embolized to descending aorta, emergency AVR (multiple complications)
x. Aortic dissection during attempted TAVI, AVR 3 mos later abandoned due to access procedures, AVR 3 mos later

These patients were included in TAVI arm for both the ITT and AT analyses. The impact of these patients on the overall results is unknown since the beneficial effects of AVR could possibly introduce bias in favor of the TAVI arm. FDA also notes that in the TAVI arm, these patients would have remained untreated for their critical aortic stenosis without the use of open AVR as a bailout procedure. It should also be noted that converting a patient from an elective TAVI to an emergency AVR is known to increase the risk of mortality.

### 10.6.2.4 Delay in Treatment

In the TAVI group, there was a mean 11-day delay from randomization to the procedure, whereas in the AVR group the mean delay was 16 days. The data also show that more patients in the control group had a considerable delay between randomization and treatment than in the treatment group. For instance, one patient did not have AVR because of “worsening lab values” – however, this occurred 14 months after randomization.

The impact of delay in treatment on results is difficult to interpret, but could have confounding effects on the assessment of overall safety and effectiveness.

### 10.6.2.5 Concomitant Operations in the AVR group

This trial was intended to compare isolated open AVR to isolated TAVI. The inclusion/exclusion criteria specifically excluded patients from the study with “Untreated coronary artery disease (CAD) requiring revascularization.” However, 21 patients (6.7%) in the AVR group had a concomitant coronary artery bypass (CAB) procedure. Patients with CAD remained untreated in the SAPIEN group.

In addition, multiple exclusion criteria were intended to exclude the need for operation for associated conditions. However, concomitant operations for associated conditions were performed in 12.8% (40/313 AT) of the control patients. These data are provided in the table below.

P110021: FDA Summary of Safety and Effectiveness Data

Table 17 - Concomitant Operations

|  CABG | 20  |
| --- | --- |
|  CABG + aortic endarterectomy | 1  |
|  MV repair | 4  |
|  MV replacement | 1  |
|  MV repair, annular enlargement | 1  |
|  MV repair, root enlargement | 1  |
|  TV repair | 1  |
|  TV annuloplasty, Root replacement | 1  |
|  Root/arch replacement | 3  |
|  Aortoplasty | 2  |
|  Ascending Aortic endarterectomy | 3  |
|  Ablation for afib | 1  |
|  Excision Left Atrial Appendage | 1  |
|  TOTAL Patients with concomitant operations (% total 40/313) (As Treated) | 40 (12.8%)  |

Of the 40 patients who underwent concomitant operations in the control group, 42.5% (17/40) had died by 1 year.

The operative risk of combination operations (AVR+CAB, AVR+ other valves, ablation, etc.) is known to be higher than for isolated valve procedures. This higher operative risk could bias safety results in a short-term study. Patients randomized to the SAPIEN group who were untreated for these concomitant conditions could affect long-term results for TAVI, but might not be captured in this shorter term study. This could introduce bias in favor of the treatment group in both the ITT and AT analyses because of the short-term increased risk of concomitant operations and because the long-term effectiveness of treating these conditions were not captured by the short-term (1 year) primary effectiveness endpoint.

## 10.6.2.6 Lack of Standardized Antithrombotic Treatment in the AVR population

There were no pre-specified antithrombotic regimens in the control group in the protocol for this study. The following regimen for antithrombotic drugs was provided for the TAVI arm.

Table 18 - Recommended Medication Regimen

|  Medication | Pre-Procedure | During Catheterization | Post-Procedure | 30-Day Follow-up | 6 month follow-up  |
| --- | --- | --- | --- | --- | --- |
|  IV Heparin | PRN | 5000 IU bolus, then as needed to achieve/maintain ACT ≥250 sec |  |  |   |
|  Aspirin | 75-100 mg QD |  | 75-100 mg QD | 75-100 mg QD | 75-100 mg QD  |
|  Clopidogrel* | 300 mg (if not on long-term therapy) | 75 mg QD | 75 mg QD | 75 mg QD for 6 months |   |

* Ticlopidine could be used instead of clopidogrel at the investigator's discretion

P110021: FDA Summary of Safety and Effectiveness Data

The non-protocolized antithrombotic regimen resulted in important variations between the two arms of the trial, especially in the use of clopidogrel in the larger transfemoral arm. The following table presents the actual antithrombotic use over the first year.

Table 19 - Actual Medication Regimen

|  Medication | Randomized Patients (% pts)  |   |   |   |   |
| --- | --- | --- | --- | --- | --- |
|   |   | Transapical |   | Transfemoral  |   |
|   |  Visit | AVR (N=103) | SAPIEN (N=104) | AVR (N=248) | SAPIEN (N=244)  |
|  Aspirin | Baseline | 64/103 (62.1%) | 64/104 (61.5%) | 150/248 (60.5%) | 166/244 (68.0%)  |
|   | 1 yr | 52/103 (50.5%) | 62/104 (59.6%) | 143/248 (57.7%) | 171/244 (70.1%)  |
|  |   |   |   |   |   |
|  Clopidogrel | Baseline | 29/103 (28.2%) | 25/104 (24.0%) | 42/248 (17.0%) | 52/244 (21.3%)  |
|   | 1 yr | 19/103 (18.4%) | 22/104 (21.2%) | 26/248 (10.5%) | 72/244 (29.5%)  |
|  |   |   |   |   |   |
|  Warfarin | Baseline | 21/103 (20.4%) | 19/104 (18.3%) | 50/248 (20.2%) | 49/244 (20.1%)  |
|   | 1 yr | 8/103 (7.8%) | 11/104 (10.6%) | 17/248 (6.9%) | 28/244 (11.5%)  |

The lack of a standardized antithrombotic protocol in the AVR arm makes evaluation of the post-procedural stroke rate difficult to interpret. There are currently no approv…

---

**Source:** [https://fda.innolitics.com/device/P110021](https://fda.innolitics.com/device/P110021)

**Published by [Innolitics](https://innolitics.com)** — a medical-device software consultancy. We help companies design, build, and clear FDA-regulated software and AI/ML devices. If you're preparing [a PMA](https://innolitics.com/services/regulatory/), [a 510(k)](https://innolitics.com/services/510ks/), [a SaMD](https://innolitics.com/services/end-to-end-samd/), [an AI/ML medical device](https://innolitics.com/services/medical-imaging-ai-development/), or [an FDA regulatory strategy](https://innolitics.com/services/regulatory/), [get in touch](https://innolitics.com/contact).

**Cite:** Innolitics at https://innolitics.com