RELAY THORACIC STENT-GRAFT WITH PLUS DELIVERY SYSTEM

P110038 · Bolton Medical, Inc. · MIH · Sep 21, 2012 · Cardiovascular

Device Facts

• Record ID: P110038
• Device Name: RELAY THORACIC STENT-GRAFT WITH PLUS DELIVERY SYSTEM
• Applicant: Bolton Medical, Inc.
• Product Code: MIH · Cardiovascular
• Decision Date: Sep 21, 2012
• Decision: APPR
• Device Class: Class 3
• Attributes: Therapeutic

Intended Use

The Relay® Thoracic Stent-Graft with Plus Delivery System is indicated for use in the endovascular repair of fusiform aneurysms and saccular aneurysms/penetrating atherosclerotic ulcers in the descending thoracic aorta in patients having appropriate anatomy, including Iliac or femoral access vessel morphology that is compatible with vascular access techniques, devices and/or accessories Non-aneurysmal aortic neck diameter in the range of 19 – 42 mm Non-aneurysmal proximal aortic neck lengths between 15 and 25 mm and distal aortic neck lengths between 25 and 30 mm, depending on the diameter stent-graft required

Device Story

The Relay® Thoracic Stent-Graft with Plus Delivery System is a modular, endovascular device for treating descending thoracic aortic aneurysms and penetrating atherosclerotic ulcers. The system consists of a self-expanding nitinol stent-graft and a two-stage delivery system. The stent-graft is delivered endoluminally via femoral or iliac artery access over a 0.035" guidewire. The delivery system features a primary and secondary sheath to facilitate tracking through tortuous anatomy and an apex release mechanism for controlled deployment and potential repositioning. Upon deployment, the nitinol stents expand to conform to the aortic wall, providing a seal. The device is used in a clinical setting by physicians. The output is the exclusion of the aneurysm from systemic circulation, which reduces the risk of rupture and provides a less invasive alternative to open surgical repair.

Clinical Evidence

Pivotal prospective, non-randomized, multi-center study (n=120 Relay; n=60 surgical control). Primary safety endpoint: major adverse events (MAE) within 1 year (26.7% Relay vs 50.0% surgical; p<0.001). Primary effectiveness endpoint: freedom from major device-related adverse events at 1 year (97% free; lower limit of 97.5% CI 0.93). Secondary data from feasibility study, continued access arm, and European registry (RESTORE) support findings. Bench testing verified structural integrity, fatigue life (400 million cycles), and biocompatibility.

Technological Characteristics

Materials: Nitinol stents, polyester graft fabric, platinum-iridium radiopaque markers. Design: Self-expanding, modular, straight or tapered configurations. Delivery: Two-stage system (22-26 Fr), 90 cm working length, hydrophilic coating. Sterilization: Gamma irradiation (SAL 10^-6). Connectivity: None. Software: None.

Indications for Use

Indicated for endovascular repair of fusiform and saccular aneurysms and penetrating atherosclerotic ulcers in the descending thoracic aorta in patients ≥18 years with appropriate iliac/femoral access and aortic neck anatomy (19-42 mm diameter; 15-25 mm proximal neck length; 25-30 mm distal neck length). Contraindicated in patients with conditions threatening graft infection or known allergies to device materials.

Submission Summary (Full Text)

{0}

# Summary of Safety and Effectiveness Data

## I. GENERAL INFORMATION

| Device Generic Name: | Endovascular Graft |
| --- | --- |
| Device Trade name: | Relay® Thoracic Stent-Graft with Plus Delivery System |
| Device Procode: | MIH |
| Applicant’s Name and Address: | Bolton Medical, Inc., 799 International Parkway, Sunrise, Florida 33325 |
| Premarket Approval Application Number: | P110038 |
| Date of Panel Recommendation: | None |
| Date of FDA Notice of Approval: | September 21, 2012 |
| Expedited: | Not Applicable |

## II. INDICATIONS FOR USE

The Relay® Thoracic Stent-Graft with Plus Delivery System is indicated for use in the endovascular repair of fusiform aneurysms and saccular aneurysms/penetrating atherosclerotic ulcers in the descending thoracic aorta in patients having appropriate anatomy, including

- Iliac or femoral access vessel morphology that is compatible with vascular access techniques, devices and/or accessories
- Non-aneurysmal aortic neck diameter in the range of 19 – 42 mm
- Non-aneurysmal proximal aortic neck lengths between 15 and 25 mm and distal aortic neck lengths between 25 and 30 mm, depending on the diameter stent-graft required

## III. CONTRAINDICATIONS

The Relay® Thoracic Stent-Graft with Plus Delivery System is contraindicated in the following clinical scenarios:

- Patients who have a condition that threatens to infect the graft

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 1 of 98

{1}

- Patients who are sensitive to, or have known allergies to, the device materials

## IV. WARNINGS AND PRECAUTIONS

The warnings and precautions can be found in the Instructions for Use for the Relay® Thoracic Stent-Graft with Plus Delivery System.

## V. DEVICE DESCRIPTION

### A. Relay® Thoracic Stent-Graft with Plus Delivery System

The Relay® Thoracic Stent-Graft with Plus Delivery System is comprised of two components:

- Relay® Thoracic Stent-Graft
- Plus Delivery System

The Relay® Thoracic Stent-Graft is intended to be delivered endoluminally via access through the femoral or iliac artery to the site of the lesion using the Plus Delivery System. The stent-graft is constrained within the secondary sheath (2nd stage), which was further constrained within the primary sheath (1st stage) until deployed at the intended site of treatment. The pre-loaded system is advanced to the diseased location over a guidewire. Upon deployment, the stent graft self-expands due to the superelastic properties of the nitinol stent. Following expansion of the device within the aorta, the proximal and distal ends of the stent-graft are intended to conform to the shape and size of the proximal and distal seal zones of the targeted lesion due to the radial force of the stents.

### B. Relay® Thoracic Stent-Graft

Relay® Thoracic Stent-Grafts are composed of self-expanding nitinol stents sutured to polyester graft fabric. The skeleton of these devices is a series of sinusoidal nitinol stents sewn along the length of the graft fabric with surgical suture. The Relay® Thoracic Stent-Graft features a bare proximal terminating stent and a covered distal terminating stent. Relay® stent-grafts are constructed with four different types of stents, each having a specific function at their location. Figure 5-1 shows the Relay® graft with associated stent types. In addition, there is nitinol spiral support strut for longitudinal support. Platinum-iridium, radiopaque, dumb-bell shaped markers are strategically placed on the graft to facilitate radiographic visualization of the graft material edge.

Relay® Thoracic Stent-Grafts are offered with diameters ranging from 22 mm to 46 mm and covered lengths ranging from approximately 90 mm to 250 mm. Additionally, Relay® Thoracic Stent-Grafts are available in straight configurations, where the diameter is uniform over the length of the stent-graft, or tapered configurations where the diameter decreases over the length of the implant.

During manufacturing, the Relay® Thoracic Stent-Grafts are preloaded into a delivery system.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 2 of 98

{2}

Figure 5-1: Relay® Stent-Graft Configuration with Types of Stents

## C. Relay® Thoracic Stent-Graft Configuration and Placement

The Relay® Thoracic Stent-Graft is a modular device that accommodates the use of additional stent-graft sections depending on the configuration of the anatomy where single or multiple components may be required to achieve sufficient coverage of the diseased aorta. If the vessel diameter and condition require variable proximal and distal diameter devices, the smallest diameter stent-graft should be placed first, either at the proximal or distal end of the lesion, as appropriate. The additional section is to be deployed within the primary piece following the oversizing requirements as detailed in the Instructions for Use (IFU) manual. If the vessel diameter and condition require the same proximal and distal diameter devices, the primary section should be placed first at the proximal end of the lesion. To achieve the same final diameter with the proximal and distal sections, a tapered configuration is required for the distal section. The flare of the tapered graft permits the oversizing requirements between components.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 3 of 98

{3}

Figure 5-2: Combining Relay® Stent-Grafts

## D. Plus Delivery System

The Plus delivery system is a single-use, disposable, two-stage delivery device consisting of sheaths and catheters (primary introduction sheath, secondary delivery sheath, through lumen), as well as a handle and apex release mechanism. It is available in outer diameters ranging from 22 to 26 Fr depending on stent-graft size, and has a working length of 90 cm. The distal end of the system features a pre-formed curve, intended to

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 4 of 98

{4}

facilitate alignment of the stent-graft upon deployment. The stent-graft is constrained within the secondary sheath (2nd stage), which was further constrained within the primary sheath (1st stage). The delivery system is designed to be tracked over a 0.035" guide wire to facilitate introduction of the device through the femoral and iliac arteries. Once the system reaches the placement location, the proximal handle of the delivery system is advanced to exit the secondary sheath from the primary sheath in preparation for deployment. The secondary sheath, composed of thin wall, flexible polyester, enables the thoracic stent-graft to be more easily advanced and deployed in curved and tortuous portions of the anatomy than a polymeric sheath would allow. The secondary sheath, which was connected to the delivery catheter and the delivery handle (deployment grip), is retractable to deploy the constrained stent-graft in a controlled fashion.

The system features an apex release mechanism which constrains the bare stent. This mechanism is controlled by sliding the outer control tube over the guide wire lumen after the deployment from the secondary sheath. This feature provides the ability to reposition the device in a partially deployed state. In addition, this feature provides a controlled apposition of the bare stent to the vessel wall. Figure 5-3 shows the complete delivery system.

Figure 5-3: Delivery System Schematic

| Legend | | | |
| --- | --- | --- | --- |
| 1. Delivery System Tip | 5. Radiopaque Marker | 9. Flush Port / Stopcock | 13. Apex Release Retainer |
| 2. Apex Holder | 6. Stationary Grip | 10. Deployment Grip | 14. Apex Release Grip |
| 3. Inner Secondary Sheath | 7. Shipping Retainer | 11. Controller | 15. Guidewire Luer |
| 4. Outer Primary Sheath | 8. Main Body | 12. Stainless Steel Rod | 16. Flushport |
| | | | 17. Handle Body Mark |

PMA P110038: FDA Summary of Safety and Effectiveness Data

{5}

# VI. ALTERNATIVE PRACTICES AND PROCEDURES

There are several other alternatives for the treatment of thoracic aortic aneurysms (TAA) including endovascular repair using another endovascular grafting system, surgical implantation of a synthetic graft within the aneurysmal vessel, and medical management. Each alternative has its own advantages and disadvantages. The physician should fully discuss these alternatives with his/her patient to select the method that best meets expectations and lifestyle.

# VII. MARKETING HISTORY

The Relay® Thoracic Stent-Graft has been commercially available for distribution in Europe since April 2005. Approval of the Plus Delivery System was granted in March 2009. The Relay® Thoracic Stent-Graft with Plus Delivery System is currently sold in several other nations including Australia, Argentina, Brazil, and Mexico.

The Relay® Thoracic Stent-Graft has not been withdrawn from the market for any reason related to safety or effectiveness.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 6 of 98
13

{6}

# VIII. POTENTIAL ADVERSE EFFECTS OF THE DEVICE ON HEALTH

Adverse events or complications associated with the use of the Relay® Thoracic Stent-Graft with Plus Delivery System that may occur and that may require intervention include, but are not limited to, those listed in Table 8-1.

Table 8-1: Potential Adverse Effects

| Access Failure | Dysphagia | Reaction to Anesthesia |
| --- | --- | --- |
| Adynamic Ileus | Edema (e.g., leg, foot) | Reaction / Pain at Catheter Insertion Site |
| Allergic Reaction (to contrast, antiplatelet therapy, stent-graft materials) | Embolism (with transient or permanent ischemia or infarction) | Renal Complications (failure, insufficiency) |
| Amputation | Endoleak | Reoperation |
| Anaphylaxis | Excessive / Inappropriate Radiation Exposure | Seizure |
| Anesthetic Complications | Extrusion / Erosion | Seroma |
| Aneurysm Expansion | Fever / Localized Inflammation | Shock |
| Aneurysm / Lesion Rupture | Fistulas (aorto-bronchial, aorto-enteric, aorto-esophageal, arteriovenous, lymph) | Stent-Graft Dilatation / Rupture |
| Angina | Gastrointestinal Complications (bleeding, diarrhea, nausea, vomiting) | Stent-Graft Failure |
| Bleeding Complications (hemorrhage, hematoma, coagulopathy, procedural bleeding, post-procedural bleeding) | Genitourinary Complications (urinary incontinence, hematuria) | Stent-Graft Infection |
| Blindness | Hepatic Failure | Stent-Graft Migration |
| Bowel Ischemia | Impotence | Stent-Graft Misplacement |
| Bowel Necrosis | Incision Site Complications | Stent-Graft Tearing/Wear |
| Bowel Obstruction | Infection/Sepsis (including wound infection) | Stent-Graft Twisting/Kinking |
| Cardiac events (arrhythmia, tachyarrhythmia, cardiac tamponade, myocardial infarction, congestive heart failure, hypertension, hypotension, tachycardia, bradycardia) | Intramural Hematoma | Suture Fracture |
| Catheter Breakage | Ischemia (spinal cord, perfusion pathways, peripheral, limb, vascular) | Tissue Necrosis |
| Cerebral Vascular Accident – CVA (stroke) | Lymphocele | Transient Ischemic Attack |
| Change In Mental Status | Neuropathy (e.g., femoral) | Vascular Access Complications |
| Claudication (buttock, lower limb) | Pain (e.g., intercostals pain, general pain, etc.) | Vascular Spasm/Trauma |
| Compartment Syndrome | Paralysis/Paraplegia/Paresthesia/Paraparesis/Spinal Neurological Deficit | Vessel Damage /Trauma/Rupture |
| Contrast Toxicity | Perforation (vessel / device) | Vessel Dissection |
| Conversion To Open Repair | Peripheral Nerve Injury | Vessel (arterial or venous) or Device (Stent-Graft) Occlusion/Thrombosis |
| Death | Post Implantation Syndrome | Vessel or Stent-Graft Stenosis |
| Deployment Difficulties/Failures | Pseudoaneurysm | Wire Form Fractures |
| Device Dehiscence | Pulmonary Complications (atelectasis respiratory failure, respiratory depression, pneumonia, pulmonary edema, pulmonary embolism) | Wound Dehiscence |
| Device Insertion Or Removal Difficulty | Radiation Overexposure or Reaction | Wound Healing Complications |

For adverse events that occurred during the clinical studies, please see Section X, below.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 7 of 98

{7}

# IX. SUMMARY OF PRECLINICAL STUDIES

## A. Laboratory Studies

### Biocompatibility

Biocompatibility testing was conducted on the materials that comprise the Relay® Thoracic Stent-Graft with Plus Delivery System in accordance with ISO 10993-1 and Good Laboratory Practices (21 CFR Part 58). Biocompatibility studies for the Relay® Thoracic Stent Graft were conducted based on the principles of an implant device that is in permanent contact with blood (&gt;30 days), whereas the studies for the Plus Delivery System were based on the principles of an externally-communicating device with limited contact with circulating blood (&lt;24 hr). Bolton Medical utilizes two separate suppliers for the graft fabric (Supplier 1 and Supplier 2). Since the materials from the two suppliers has been confirmed to be equivalent, this testing supports both suppliers.

Table 9-1 and Table 9-2 provide a summary of the biocompatibility test results for the Relay® Thoracic Stent-Graft and the Plus Delivery System, respectively.

Table 9-1: Stent-Graft Tests

| Test | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Cytotoxicity ISO Elution Method | To evaluate if the device has the potential to induce cytotoxic effects | Test article must not show evidence of toxicity induction | Pass. No evidence of cell lysis or toxicity induction; test article graded less than 2 (mild reactivity) |
| Hemocompatibility | | | |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 8 of 98

{8}

Table 9-1: Stent-Graft Tests

| • In vitro
Hemolysis study | To assess if the device could cause red blood cell hemolysis | Test article must be non-hemolytic. | Pass. Hemolytic index for the test article in direct contact with blood was 1.1%. Hemolytic index for test article extract was 0.1%. As a result the test article was considered non-hemolytic |
| --- | --- | --- | --- |
| • C3a Complement Activation Study | To assess if the device activates the complement system | Test article must not exhibit significant activation of the complement system | Pass. C3a concentration for the test article was not higher than for the activated NHS control or the negative control. As such the test article was not considered an activator of the complement system |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 9 of 98

{9}

Table 9-1: Stent-Graft Tests

| • SC5b-9 Complement Activation Study | To assess if the device activates the complement system | Test article must not exhibit significant activation of the complement system | Pass. The concentration of SC5b-9 in the test article was statistically higher than both the activated NHS and negative controls and was 17.1% of the positive reference control. However, the SC5b-9 result was within the historical range of the activated NHS and negative controls. As such, the test article was considered to be a low potential activator of the complement system |
| --- | --- | --- | --- |
| • Partial Thromboplastin Time | To determine the potential of the test article to cause an effect on the coagulation cascade via the intrinsic coagulation pathway | Test article must not exhibit significant effects on the coagulation pathway | Pass. Plasma exposed to the test article had an average clotting time of 300 seconds and was 100% of the negative control. As such, the test article was considered a non-activator. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 10 of 98

{10}

Table 9-1: Stent-Graft Tests

| • In vivo
Thromboresistance | To evaluate the potential of the device to resist thrombus formation when placed in the vasculature | Test article should demonstrate thrombus resistance | Pass. No significant thrombus detected on the stent-graft at 1 and 2 weeks. Maximum thrombus score < 1. Results suggest the test article is resistant to thrombus formation |
| --- | --- | --- | --- |
| ISO Acute Systemic Toxicity | To evaluate the potential for the device to elicit acute systemic toxic events | Test article must not induce mortality or show evidence of systemic toxicity | Pass. No mortality or evidence of systemic toxicity |
| Pyrogenicity (USP Materials Mediated Pyrogen Test) | To evaluate if the device has the ability to induce a pyrogenic response | Test article must be non-pyrogenic; | Pass. No temperature increase > 0.5°C detected; therefore, non-pyrogenic. |
| Irritation (ISO Intracutaneous Study) | To evaluate if the device has the potential to induce irritation | Test article extract must not significantly induce irritation. | Pass. The difference between the overall mean erythema/edema scores for both the sodium chloride and sesame oil test article extracts and control scores was 0.0 |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 11 of 98

{11}

Table 9-1: Stent-Graft Tests

| Sensitization (ISO Guinea Pig Maximization Test) | To evaluate the potential of the device to cause dermal irritation | Test article must not be a dermal sensitizer. | Pass. No evidence of inducing delayed dermal contact sensitization (not considered a sensitizer) for either the sodium chloride or sesame oil test article extracts. All reaction grades were 0. |
| --- | --- | --- | --- |
| Subchronic Toxicity (ISO) | To evaluate the potential of the device to cause systemic toxic effects following repeated exposures | Test article must not show evidence of systemic toxicity. | Pass. No evidence of systemic toxicity.
Daily clinical observations were within limits and similar between test and controls. No changes in histopathology, hematology, or clinical chemistry. |
| Implantation | | | |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 12 of 98

{12}

Table 9-1: Stent-Graft Tests

| ISO Muscle Implantation Test (4 weeks) | To evaluate the potential for the device to elicit irritation or toxic responses after implantation | Test article must not elicit significant irritation or toxic responses after implantation. | Pass. Macroscopic reaction was not significant compared to the negative control. Microscopically, the test article was classified as a slight irritant compared to the control |
| --- | --- | --- | --- |
| ISO Muscle Implantation Test (12 weeks) | To evaluate the potential for the device to elicit irritation or toxic responses after implantation | Test article must not elicit significant irritation or toxic responses after implantation. | Pass. Macroscopic reaction was not significant compared to the negative control. Microscopically, the test article was classified as a moderate irritant compared to the control |
| Genotoxicity
ISO Reverse Mutation Study | To evaluate if the device could induce mutagenic changes in selected bacterial test strains | Test article must be non-mutagenic | Pass. In no case was there a 2-fold or greater increase in the mean number of revertent tester strains (TA98, TA100, TA1535, TA 1537, and WP2uvrA) in the presence of the test article extract. Therefore considered non-mutagenic. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 13 of 98

{13}

Table 9-1: Stent-Graft Tests

| • In vivo Mouse Lymphoma Assay | To determine the ability of the device to induce forward mutations | Test article must be non-mutagenic | Pass. Mutant frequencies and cloning efficiencies of the test article preparations were well within the limits defined for a negative control response. Therefore considered non-mutagenic. |
| --- | --- | --- | --- |
| • In vivo Mouse Micronucleus Assay | To determine the ability of the device to induce in vivo clastogenic events or to damage the mitotic spindle | Test article must be non-mutagenic | Pass. No apparent gross manifestations of toxicity or significant erythropoietic disturbances resulting in delayed mutagenesis. Also no increases in mPCE production as compared to controls. Therefore considered non-mutagenic. |

PMA P110038: FDA Summary of Safety and Effectiveness Data

{14}

Table 9-2: Delivery System Tests

| Test | Purpose | Results | Pass/Fail |
| --- | --- | --- | --- |
| Cytotoxicity (ISO Elution Method) | To evaluate if the device has the potential to induce cytotoxic effects | Test article must not show evidence of toxicity induction | Pass No evidence of cell lysis or toxicity induction; test article graded less than 2 (mild reactivity). |
| Hemocompatibility
• In vitro Hemolysis Study | To assess if the device could cause red blood cell hemolysis | Test article must be non-hemolytic. | Pass. Hemolytic index for test article extract and for test article direct contract was 0%. As such, the test article was considered non-hemolytic |
| • Partial Thromboplastin Time | To determine the potential of the test article to cause an effect on the coagulation cascade via the intrinsic coagulation pathway | Test article must not exhibit significant effects on the coagulation pathway | Pass. Plasma exposed to the test article had an average clotting time of 303.7 seconds and was 77% of the negative control. As such, the test article was considered a minimal activator. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 15 of 98

{15}

Table 9-2: Delivery System Tests

| Test | Purpose | Results | Pass/Fail |
| --- | --- | --- | --- |
| • C3a Complement Activation Study | To assess if the device activates the complement system | Test article must not exhibit significant activation of the complement system | Pass. C3a concentration for the test article was not higher than for the activated NHS control or the negative control. As such the test article was not considered an activator of the complement system |
| • SC5b-9 Complement Activation Study | To assess if the device activates the complement system | Test article must not exhibit significant activation of the complement system | Pass. SC5b-9 concentration for the test article was not higher than for the activated NHS control or the negative control. As such the test article was not considered an activator of the complement system |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 16 of 98

{16}

Table 9-2: Delivery System Tests

| Test | Purpose | Results | Pass/Fail |
| --- | --- | --- | --- |
| • In vivo
Thromboresistance | To evaluate the potential of the device to resist thrombus formation when placed in the vasculature | Test article should demonstrate thrombus resistance | Pass. Thrombus was detected only on the flexible inner delivery sheath. This was believed to be due to flow disturbance rather than biomaterial effect. Overall, the device was relatively resistant to thrombus formation |
| Acute Systemic Toxicity (ISO) | To evaluate the potential for the device to elicit acute systemic toxic events | Test article must not induce mortality or show evidence of systemic toxicity | Pass. No mortality or evidence of systemic toxicity from the test article |
| Pyrogenicity (USP Pyrogen Test) | To evaluate if the device has the ability to induce a pyrogenic response | Test article must be non-pyrogenic | Pass; no temperature increases > 0.5°C, therefore, non-pyrogenic |
| Irritation (ISO Intracutaneous Study) | To evaluate if the device has the potential to induce irritation | Test article must not induce significant irritation. | Pass. The difference between test article extracts and control was 1.0 or less. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 17 of 98

{17}

Table 9-2: Delivery System Tests

| Test | Purpose | Results | Pass/Fail |
| --- | --- | --- | --- |
| Sensitization (ISO Maximization Test) | To evaluate the potential of the device to cause dermal irritation | Test article must not be a dermal sensitizer. | Pass. No evidence of inducing delayed dermal contact sensitization for either the sodium chloride or sesame oil test article extracts. All reaction grades were 0. |

## Bench Testing

Bolton Medical conducted comprehensive pre-clinical, bench and analytical testing on the Relay® Thoracic Stent-Graft with Plus Delivery System. The *in vitro* testing was intended to verify that the performance attributes of the Relay® Thoracic Stent-Graft with Plus Delivery System are sufficient to minimize adverse events under anticipated clinical conditions. Testing was conducted in accordance with ISO 25539-1, Cardiovascular implants --- Endovascular devices --- Part 1: Endovascular prostheses. This testing included both the stent-graft and the delivery system. The testing details include results from T=0 (baseline) as well as results using samples accelerated aged to 3 years (T=3). An asterisk (*) indicates testing was performed at both T=0 and T=3. Testing verified that the Relay® Thoracic Stent-Graft with Plus Delivery System met its product performance and design specifications. Table 9-3 outlines the tests performed.

Results obtained from these *in vitro* studies support the safety and effectiveness of the Relay® Thoracic Stent-Graft with Plus Delivery System.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 18 of 98

{18}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Delivery System Verification Tests | | | |
| General Appearance and Physical Checks (*) | To verify that the system appearance characteristics are acceptable to the end-use, and to verify that the components of the system are working properly | The device must not exhibit signs of objectionable discoloration or damage, and all aspects of the system should be in proper working order and in the intended positioning as required by the IFU | All samples met specification |
| Bond strength (*) | To determine the bond strength of the joints and/or fixed connections of the delivery system | Sub-assemblies tested must meet pre-determined pull forces depending on the bond. Acceptance criteria ranged from 5 lbs to 25 lbs (22.24 N to 111.2N) | All samples met specification |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 19 of 98

{19}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Component dimensional compatibility (includes dimensional verification) (*) | To determine the system dimensions for verification to design specification, and to evaluate the dimensional compatibility between the system and its accessory devices listed in the Instructions for Use (IFU) | • System must be compatible with 0.035” guide wire and 0.036” mandrel
• Delivery system sheath O.D. must meet pre-determined tolerances.
• Useable length must meet predetermined specifications: 600 mm +/- 5 mm (non-deployed); 895 mm min (deployed) | All samples met specification |
| Simulated Use (includes pushability, trackability and torqueability) (*) | To evaluate the performance of the delivery system using an aortic model that simulates the intended use conditions. This test includes a qualitative assessment of simulated use, flex/kink, pushability, torqueability, and trackability of the thoracic system | Characterization study | Guide wire acceptance, pushability, trackability, torqueability, kink resistance, and flushing of the guide wire lumen were all evaluated and determined to be acceptable. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 20 of 98

{20}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Profile/diameter Test (*) | To determine the system maximum diameter at the loaded stent-graft section (largest profile) in order to evaluate the dimensional component compatibility between the delivery system and the vasculature | The device must pass through a Go Gauge over the loaded section of the device | All dimensions for the test samples met the acceptance criteria. |
| Torsional Bond Strength | To determine the torque required to cause failure of the bonded joints of the delivery system components | The delivery system sheath introducer must be torqued at 180° without any damage to the sheath bond. | All samples were torqued 180° without any sheath attachment damage. |
| Tubing tensile strength | To determine the strength of the tubing used in the delivery system. | Introducer Sheath: 7% maximum elongation at 25 lbs (111.2 N) | All test samples were within 7% elongation at 25 lbs.(111.2N) |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 21 of 98

{21}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Force to deploy (*) | To determine the force to advance and deploy the stent-graft from the delivery system as well as testing for all relevant characteristics pertaining to deployment (e.g., accuracy, re-seating, system removal, etc.). | - Advancement force: ≤25 lbs. (111.2N)
- Deployment force: ≤25 lbs. (111.2 N)*
- Clasp Release ≤10lbs (44.5N) | All samples met acceptance criteria. |
| Flex / Kink | To determine the minimum radius of curvature that the system can accommodate without kinking | The loaded delivery system must permit deployment around the tested radii arches without kinking that would prevent deployment or cause damage to the stent-graft or delivery system | All samples met acceptance criteria. |
| Assessment of hemostasis (*)^{a} | To evaluate the system’s ability of any seals or valves to maintain adequate hemostasis | Amount of water obtained through leaking in 1 minute should be ≤15 g. | Samples met the acceptance criteria. The maximum amount of water lost was 6.7g |
| Visibility | To evaluate the ability to visualize the system using the imaging techniques specified in the IFU. | Test units must be visible under fluoroscopy | All samples met acceptance criteria. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 22 of 98

{22}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Lubricity test (*)^{b} | To determine the lubricity of the Plus delivery system sheath | Force of the coated sheath must be lower than an non-coated sheath | There was a significant difference between coated and non-coated sheaths/tips. |
| Manual alignment | To evaluate the ability of the Plus system to manually align while still in the secondary sheath | Characterization study | All samples were able to be manually rotated 360° without difficulty |
| Tracking through tortuous vessel | To evaluate the ability of the Plus delivery system to tract through extreme tortuous aortas | Characterization study | All samples were evaluated for pushability, tracking, kinking and torqueability. |
| Vessel Wall Rigidity | To evaluate the ability of the Plus delivery system to track through an extremely tortuous aorta | Characterization study | No excessive force against the vessel wall was noted. |
| Particulate Test | To determine the amount of particulate matter associated with the hydrophilic coating of the introducer and tip (as compared to original uncoated system) | Characterization Study | No statistically significant difference between the uncoated sheaths and coated sheaths, thus confirming that hydrophilic coating does not create a greater incidence of loose particulates. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 23 of 98
30

{23}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Stent-Graft Design Verification Tests | | | |
| Stent-graft Dimensional verification (*) | To determine the dimensions of the stent-graft in the deployed state for verification to design specifications | Straight Configuration: Length must be within +/- 2 mm of the drawing dimension
Tapered Configuration: Inner diameter must be within +/- of drawing dimensions. | All dimensions met the acceptance criteria |
| Visibility | To evaluate the ability to visualize the system using the imaging techniques specified in the IFU. | Test units must be visible under fluoroscopy | All test units were visible under fluoroscopy |
| Implant length to diameter relationship | To determine the relationship between implant length and expanded implant diameter | The length of the stent-graft must be within specification (+/- 2 mm of the assembly drawing value) while compressed in the minimum and maximum simulated vessel sized tubes. | All samples met the acceptance criteria. |
| Strength of stent/attachment to graft bond (*)^{c} | To determine the strength of the fixations or bond between the graft material and the stent/attachment | 5 lbs (22.2 N) per apex | All samples met the acceptance criteria. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 24 of 98

{24}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| | system. | | |
| Recoil (*) | To determine the outer diameter of the stent-graft in the deployed state for verification to design specifications. The purpose of this test was to show that the implant can withstand the strains experienced in radial compression during loading and unloading without any significant change to dimensions or geometry | Stent-graft recoil outer diameter must be within –0 mm and +2 mm of the nominal diameter at the proximal and distal ends. (-1 / +2 for Relay® Plus Shelf Life) | All dimensions met the acceptance criteria |
| Flex/Kink | To determine the minimum radius of curvature that the stent-graft can accommodate without kinking | The stent-graft must bend into various radii arches without kinking, which was defined as 25% or more of the graft lumen not being patent | All samples met specification |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 25 of 98
32

{25}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Flex / Kink -- Apposition Test | To determine if the stent-graft can be deployed in a straight section of the simulated vessel to verify the pre-curve of the nitinol inner control tube does not affect graft apposition | The graft loaded in the delivery system must be deployed with complete apposition. | All stent-grafts had complete apposition up to the first covered stent |
| Stent-graft Integrity (post-deployment) (*) | To demonstrate that the stent-graft retains its physical integrity after the deployment process | The sample must not exhibit physical damage that will negatively impact the performance of the device. Any observed damage will be analyzed on an individual basis. | There were no negative observations noted. |
| Crush resistance | To determine the force required to permanently radially deform or fully collapse the stent-graft as measured perpendicular to the longitudinal axis | Observations were documented as pass/fail along with the forces used to crush the stent-graft and the deflection observed. Any deformation to the stent-graft was considered a failure. | All samples were crushed to collapse without damage. |
| Local compression | To determine the deformation of the stent-graft in response to localized compressive forces, perpendicularly applied to the | Observations were documented as pass/fail along with the forces used to compress the stent-graft and the deflection observed. Any deformation to the stent-graft was | All samples were compressed to collapse without damage. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 26 of 98
83

{26}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| | longitudinal axis of the stent-graft. | considered a failure. | |
| Migration resistance and sealing | To determine the force required to displace the stent-graft in a mock artery. This test provides an indication of the resistance to migration provided by the fixation mechanisms of the stent-graft. In addition, it determined if the fixation points were against the mock artery completely in order to address sealing characteristics. | Safety coefficient (representing the stability of the device and based on the ratio of ultimate contact shear to actual contact shear) must be > 1 | Coefficient for 34mm size = 2.29; coefficient for 46 mm size = 2.00 |
| Radial outward force (hoop strength) (*) | To determine the force exerted by a self-expanding implant as a function of the implant diameter | Characterization study / Positive outward force expected | The Relay® stent-graft demonstrated positive outward radial force. Proximal seal zone radial force ranged from 2.8 N to 3.9 N; distal seal zone ranged from 3.4 N to 5.3 N. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 27 of 98
34

{27}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Burst/circumferential strength | To determine the pressurized burst strength or circumferential strength of the stent-graft if used with an accessory balloon. | The stent-graft must withstand 1.5 ATM of pressure without damage | All samples withstood the 1.5 ATM without damage |
| Longitudinal tensile strength | To determine the longitudinal tensile strength of the stent-graft | 25 lbs (111.2N) | All samples met the acceptance criteria. |
| Pull test for modular components | To determine the force required to separate the modular components of a stent-graft in the deployed state. | Characterization study / 4 N minimum | All samples exhibited forces ranging from 6N to 10 N on average. Average separation force was 8.7N |
| Factory anastomotic strength | To determine the tensile strength of any manufactured anastomosis (in this case, graft seam) | 16 lbs (71.2 N) per 2 cm section | Average force was 37.8N. |
| Porosity/water permeability | To determine the rate of fluid flow through the wall of the stent-graft as virgin material and with sutures | Characterization study | Seamed Relay® material had a higher permeability than the non-seamed material (413 mL/min/cm² versus 232 mL/min/cm²). |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 28 of 98

{28}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| Integral water permeability | To determine the rate of water leakage through the entire stent-graft, incorporating all modular components and extension devices. | Characterization study | The integral water permeability for the stent-graft was 168 ml / min / cm². All of the samples were tested at hypertensive blood pressure (150mmHg) and one was at (140mmHg).

The integral water permeability was similar to the non-seamed fabric (205 mL/min/cm² vs. 168 mL/min/cm²) |
| Corrosion | To evaluate corrosion resistance of the stent-grafts metal components | Characterization study | The test was conducted per ASTM F2129 and evaluated the general resistance to pitting corrosion. Results indicated resistance to localized corrosion. Average breakdown potentials on pre and post fatigued stents were greater than 600 mV vs. SCE |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 29 of 98

{29}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| MR compatibility
Conducted per ASTM F 2052 and ASTM F 2182 | To characterize the stent-graft’s performance in the magnetic resonance environment | 1) The presence of the stent-graft must not pose an additional unacceptable risk to patients when subjected to 1.5T and 3.0T magnetic fields.
2) To characterize image artifact | No observed magnetic field interactions (e.g., translational attraction, migration, or torque) and no MR-related heating at levels to present risk.
Image artifact was characterized. |
| Durability
---Stress/strain analysis (Finite Element Method Analysis) | Finite element method analysis was used to determine the maximum strains in compression when subjected to catheter loading and an in vivo pulsatile loading environment. | Characterization study | The worst case stent design was identified.
Information was used as a reference in appropriate in vitro testing including pulsatile fatigue testing. |
| ---Fatigue (stent apex) | To evaluate the durability of the stents | Test samples must demonstrate a stent fatigue life in excess of 400 million cycles (10 years in vivo simulation). | No fractures after 400 million cycles |
| ---Fatigue (fabric seam) | To evaluate the durability of the fabric | Test samples must demonstrate a seam fatigue life in excess of 400 million cycles. (10 years in vivo simulation). | No suture breaks, fabric tears, suture hole elongation, or seam separations after 400 million cycles |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 30 of 98

{30}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| --Fatigue (pulsatile whole device; single section) | To evaluate the durability of the stent-graft in a simulated in vivo environment | Devices must demonstrate structural and lumen integrity over 400 million cycles. (10 years in vivo simulation). | Devices retained basic structural and lumen integrity |
| --Fatigue (pulsatile whole device, overlap) | To evaluate the durability of overlapped stent-grafts in a simulated in vivo environment | Devices must demonstrate structural and lumen integrity over 400 million cycles. (10 years in vivo simulation). | Devices retained basic structural and lumen integrity |
| --Fatigue (pulsatile, whole device, bending, overlap) | To evaluate the durability of overlapped stent-grafts in a bent configuration in an in vivo environment | Devices must demonstrate structural and lumen integrity over 400 million cycles. (10 years in vivo simulation). | Devices retained basic structural and lumen integrity |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 31 of 98
38

{31}

Table 9-3: Relay® Thoracic Stent-Graft with Plus Delivery System Bench Test Results

| Tests | Purpose | Acceptance Criteria | Results |
| --- | --- | --- | --- |
| --Fatigue (longitudinal / orbital) | To evaluate the durability of the stent-graft in a simulated in vivo environment | Devices must demonstrate structural and lumen integrity over 400 million cycles (10 years in vivo simulation). | Devices retained basic structural and lumen integrity, no fractures of longitudinal support system detected |
| (*) indicates testing done at both T = 0 and T = 3.
a Shelf-life testing on original Relay® system only since there was no change to the delivery system with regard to the hemostatic control mechanism in the Relay® with Plus Delivery System
b Shelf-life testing on the Relay® with Plus Delivery System only since the original system did not have a hydrophilic coating
c Shelf-life testing on original Relay® system only since there was no change to the stent-graft with the introduction of the Plus Delivery System
d Package integrity testing was conducted as part of the shelf-life studies of the original Relay® system and successfully demonstrated that the package remains integral after 3 years simulated aging. Since the packaging configuration was not changed with the introduction of the Plus delivery system, this test was not repeated | | | |

## B. Animal Studies

Preclinical, in vivo animal testing, using full-scale devices, manufactured under the same conditions as product to be commercialized was conducted for up to 26 weeks (6 months) in 15 ovine test systems to evaluate the delivery/deployment, functionality (e.g., patency, integrity, visibility, etc.), and healing associated with the Relay® Thoracic Stent-Graft when placed in the thoracic aorta of sheep to ensure the effects of implantation on this part of the vascular anatomy were adequately evaluated. This study was conducted in accordance with applicable portions of Good Laboratory Practice Regulations (21 CFR Part 58).

The results demonstrated adequate performance of the Relay® Thoracic Stent-Graft with it original delivery system, Transport® as assessed by adequate access, advancement, deployment, deployment accuracy, visibility and other related features. Although the in vivo animal testing was conducted with the prior delivery system (i.e. Transport®), the vast majority of the testing evaluated the stent-graft which remained unchanged. Stent-graft patency, integrity and histopathological responses were acceptable. A summary of the in vivo animal testing is provided in Table 9-4.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 32 of 98

{32}

Table 9-4: Summary of Relay® Ovine Implant Study

| Study | Number / Type of Animal* | Test Article | Objectives | Success Criteria | Results |
| --- | --- | --- | --- | --- | --- |
| Preclinical Evaluation of the Bolton Medical Thoracic Stent Graft in an Ovine Model (UA-04-BOL1) – 4 and 26 weeks | 6 sheep – 4 week arm
7 sheep – 26 week (6 month) arm | Relay® Thoracic Stent-Graft | To evaluate the delivery/deployment, functionality (e.g., patency, integrity, visibility, etc.), and healing associated with the Relay® Stent-Graft when placed in the thoracic aorta of sheep | To ensure the effects of implantation on this part of the vascular anatomy are adequately evaluated; including adequate device handling, integrity and healing (histopathological) response | The Relay® device was easily deployed and the implant was well-tolerated by the sheep. The devices remained intact and widely patent without evidence of migration through explant at 4 and 26 weeks.

Histological evaluation verified cellular incorporation of all implants.

The development of a stable, anti-thrombogenic luminal cellular lining was observed in all grafts by 26 weeks. There was evidence mild-to-moderate inflammatory response but no histological evidence of infection. |

*Fifteen (15) sheep entered the study. Two (2) were withdrawn early due to complications. One sheep had abnormal intestinal distention at the time of the procedure, making isolation of the aorta difficult and prolonging procedure time. It was not possible to extubate this animal and it was euthanized. Another sheep suffered post-operative paraplegia, a known complication for this species, which did not resolve. The animal was euthanized.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 33 of 98

{33}

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 34 of 98

# C. Additional Studies

## Packaging, Shelf Life Testing and Sterilization

The Relay® Thoracic Stent-Graft with Plus Delivery System is a single-use device provided sterile to the end user. The Relay® Thoracic Stent-Graft with Plus Delivery System is sterilized by gamma irradiation and is validated to demonstrate a Sterility Assurance Level (SAL) of $10^{-6}$.

Packaging performance testing demonstrates that the packaging design for the Relay® Thoracic Stent-Graft with Plus Delivery System is sufficient to adequately protect the device and maintain the integrity of the device package throughout its three-year shelf life claim.

Shelf-life testing results are presented within the *in vitro* bench test results as part of Table 9-3. Accelerated shelf-life product testing conducted on the Relay® Thoracic Stent-Graft with Plus Delivery System supports a 3-year shelf-life claim.

{34}

# X. SUMMARY OF PRIMARY CLINICAL STUDIES

The safety and effectiveness data supporting the Relay® Thoracic Stent-Graft with Plus Delivery System included data from a multi-center pivotal study across the United States, a multi-center feasibility study conducted across the United States, data from a Continued Access arm of the pivotal trial, and a post-market European Registry. These sources of data are summarized in Table 10-1.

Table 10-1: Summary of Clinical Studies

| Study | Study Design | Objective | # Sites | # Enrolled |
| --- | --- | --- | --- | --- |
| Relay® Phase II study G040175 | Prospective, non-randomized, multi-center with comparison to a combination concurrent/historical control | To evaluate the safety and effectiveness of the Relay® Thoracic Stent-Graft with Plus Delivery System | 27 | 180 |
| Relay® Feasibility study (Phase I) | Prospective, single-arm multi-center | To evaluate the safety and preliminary performance | 6 | 30 |
| Continued Access | Prospective, non-randomized, multi-center | To continue gathering safety and effectiveness data of the device | 16 (20 permitted) | 12 subjects presented; enrollment ongoing |
| European registry (RESTORE) | Post-market, multi-center single-arm | Evaluate clinical performance post-market | 22 | 304 |

## Relay® Phase II Study

The applicant performed a clinical study to establish a reasonable assurance of the safety and effectiveness of the Relay® Thoracic Stent-Graft for treating descending thoracic aortic aneurysms (fusiform aneurysm and saccular aneurysms/penetrating ulcers) in the U.S. under IDE number G040175. Data from this clinical study were the basis for the PMA approval decision. A summary of the clinical study is presented below.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 35 of 98

{35}

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 36 of 98

# A. Study Design

The study was an open-label, non-randomized, prospective, multicenter, two-arm clinical study. There was no masking. This study was designed to evaluate the safety and effectiveness of the Relay® Thoracic Stent-Graft in subjects with a diagnosed thoracic aortic aneurysm or penetrating atherosclerotic ulcer compared with subjects who underwent open surgical repair for the same pathologies. The study included 120 subjects treated with the Relay® Thoracic Stent-Graft (Relay® cohort) and 60 surgical control subjects (surgical control cohort). The study included 29 investigational sites, 27 of which enrolled subjects. The surgical control cohort was a combination of prospectively- and retrospectively-treated subjects.

During the course of the study, two changes were implemented. The delivery system was modified from the original system to the Plus Delivery System. At the same time, inclusion criteria of the protocol was modified to permit enrollment of subjects with isolated penetrating ulcers (PAUs). Due to these changes, two subgroup analyses were performed for subjects who were treated with the Relay® Stent-Graft.

# 1. Clinical Inclusion and Exclusion Criteria

## a) Type of Controls

The surgical control group consisted of 60 subjects and was a combination of prospectively-treated (n=7) and retrospectively-treated (n=53) subjects. Prospective subjects were those who underwent open surgical repair after the date of the institutional review board (IRB) approval at the institution in which they were treated. The number of prospective controls was augmented by data from retrospective subjects. The necessary number of retrospective controls to achieve a total of 60 controls were identified by collecting consecutive lists of subjects who had undergone surgical repair within the 10 years prior to IRB approval for the institution in which they were treated and enrolling the 60 most recently treated surgical subjects among all the participating institutions who meet the eligibility criteria and consent to participate.

It was not feasible to randomize subjects to implantation of the Relay® Thoracic Stent-Graft or surgical repair due to physician preferences, subject preferences, and ethical concerns. Subjects were enrolled on a first come/first serve basis; however, a single investigational site was not permitted to enroll more than 30% of the enrollment total. Subjects were enrolled in the Relay® cohort if they met all of the inclusion criteria and none of the exclusion criteria. In order to minimize selection bias, similar inclusion/exclusion criteria applied to both the endovascular and surgical cohorts. Subjects in the surgical control cohort did not have to meet the anatomical criteria required for placement of the Relay® device. In addition, enrollment of surgical subjects at the same sites that were enrolling endovascular subjects was encouraged to minimize differences in subject care between the two groups.

The surgical control group consisted of both prospectively-treated and retrospectively-treated subjects. Surgical subjects were considered prospective subjects if they

{36}

underwent surgical repair after the date of IRB approval at the institution. Subjects were considered retrospective subjects if they already had surgical repair prior to the date of IRB approval. The retrospective portion of the cohort was assembled based on medical record review. Subjects were selected for screening from a master list of all patients who underwent open surgical repair in the 10 years prior to IRB approval across all participating hospitals. The most recently treated were screened first.

## b) Treatment Arms

Subject meeting eligibility criteria for the Relay® Stent-Graft were enrolled into the Relay® cohort. Subjects who did not meet the criteria for the Relay® cohort and who underwent surgical repair were enrolled into the surgical control cohort. Similarly, surgically-treated subjects retrospectively identified were enrolled into the surgical control cohort.

## c) Clinical Inclusion/Exclusion

Enrollment in the Relay® cohort was limited to patients who met the following selection criteria as shown in Table 10-2.

Table 10-2: Inclusion and Exclusion Criteria for Relay® Cohort

| Inclusion Criteria | Exclusion Criteria |
| --- | --- |
| a.) Subject was ≥18 years of age. | a) Subject had any of the following conditions in his/her descending thoracic aorta:
1. dissections – acute or chronic, in ascending or descending aorta
2. intramural hematoma (current or previous)
3. acute transection or acute traumatic injury
4. pseudoaneurysm (false aneurysm)
5. symptomatic aneurysm, including ruptured lesions. |
| b.) Subject must have met at least one of the following:
1. descending thoracic fusiform aneurysm, 5 cm in diameter or greater
2. descending thoracic aneurysm that was 4 cm or more in diameter that had increased in size by 0.5 cm in last 6 months
3. descending thoracic aneurysm with a maximum diameter that exceeded 2 times the diameter of the | b) Subject’s proximal neck diameter, measured outer-wall to outer-wall on a sectional image or multiplanar reconstruction CT was <18 or >42 mm. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 37 of 98

{37}

Table 10-2: Inclusion and Exclusion Criteria for Relay® Cohort

| Inclusion Criteria | Exclusion Criteria |
| --- | --- |
| nonaneurysmal, adjacent aorta
4. saccular aneurysm in the descending thoracic aorta or PAU. | |
| c.) Subject had proximal and distal aortic neck suitable for stent-graft placement, with diameter ranging between 18 mm and 42 mm. | c) Subject’s distal neck diameter, measured outer-wall to outer-wall on a sectional image or multiplanar reconstruction CT was <18 or >42 mm. |
| d.) Subject had a proximal attachment zone distal to the left common carotid and a distal attachment zone proximal to the origin of the celiac artery. The length of the attachment zones depended on the intended stent-graft diameter. The proximal attachment zone was 15 mm for 22 to 28 mm grafts, 20 mm for 30 to 38 mm grafts, and 25 mm for 40 to 46 mm grafts. The distal attachment zone was 25 mm for 22 to 38 mm grafts and 30 mm for 40 to 46 mm grafts. Note that coverage of the left subclavian artery was permitted. Additionally, coverage of the celiac artery was permitted but only if this artery was already occluded at the time of the procedure. | d) Subject had prohibitive calcification, occlusive disease, or tortuosity of intended fixation sites. |
| e.) Subject’s vascular dimensions (e.g., aortic diameters, length from left subclavian to celiac artery) were in the range that could safely be treated with the Relay® Delivery System. | e) Subject had circumferential thrombus in region of intended fixation sites. |
| f.) Subject had adequate vascular access (e.g., patent iliac or femoral arteries) for introduction of the delivery system (26 Fr maximum outer diameter [8.7 mm]). Alternatively, subject may have had femoral or iliac arteries that were extended via an access conduit. | f) Subject had an increasing tapered proximal neck with ≥3 mm increase in diameter from proximal fixation site to the aneurysm. |
| g.) Subject agreed to comply with 1-month, 6-month, and 1-year follow-ups, in addition to, an annual visit out to 5 years. | g) Subject had a decreasing tapered distal neck with ≥3 mm increase in diameter from distal fixation site to the aneurysm. |
| h.) Subject (or legally authorized representative) agreed to sign an ICF prior to treatment. | h) Subject’s aneurysm or distal thoracic aortic neck angle precluded advancement of the introduction system. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 38 of 98

{38}

Table 10-2: Inclusion and Exclusion Criteria for Relay® Cohort

| Inclusion Criteria | Exclusion Criteria |
| --- | --- |
| | i) Subject had an anatomical variance that would compromise circulation to the carotid, vertebral, or innominate arteries after device placement that was not amenable to subclavian revascularization. This did not apply to subjects with occluded celiac arteries. |
| | j) Subject was pregnant. |
| | k) Subject was morbidly obese preventing adequate x-ray visualization of the aorta. |
| | l) Subject had known or suspected connective tissue disorder (e.g., Marfan’s syndrome, Ehlers-Danlos syndrome). |
| | m) Subject had a blood coagulation disorder or bleeding diathesis for which treatment could not be suspended for 1 week pre and post repair. |
| | n) Subject had coronary artery disease (CAD) with unstable angina and had not received coronary revascularization within the last 3 months. |
| | o) Subject had chronic obstructive pulmonary disease (COPD) requiring the routine need for oxygen therapy outside the hospital setting (e.g., daily or nightly home use). |
| | p) Subject had acute renal failure or renal insufficiency with a creatinine value ≥2.5 mg/dL and was not on renal replacement therapy or dialysis. |
| | q) Subject had active systemic infection and/or mycotic aneurysms. |
| | r) Subject had a stroke within 3 months of the treatment date. |
| | s) Subject had less than 1-year life expectancy as evidenced by factors prohibiting major medical intervention (e.g., presence of malignant tumor, advanced age). |
| | t) Subject was participating in another research study or had received an investigational research study drug or device within 30 days of screening. |

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 39 of 98
46

{39}

Table 10-2: Inclusion and Exclusion Criteria for Relay® Cohort

| Inclusion Criteria | Exclusion Criteria |
| --- | --- |
| | u) Subject was confronted with other medical, social, or psychological issues that the investigator believed might have interfered with treatment and/or follow-up. These reasons were documented. For example, adherence to a theological or personal doctrine with aversion or opposition to blood transfusion, etc. |
| | v) Subject had a coexisting abdominal aortic aneurysm (AAA), which the investigator believed required concomitant treatment within 45 days. |
| | w) Subject had a prior AAA repair (endovascular or surgical) that was performed less than 6 months prior to treatment. |
| | x) Subject had a prior endovascular repair (e.g., stent, stent-graft) in the descending thoracic aorta. Device could not have been placed within any prior surgical graft. |
| | y) Subject had an untreatable allergy or sensitivity to contrast media or device components. |
| | z) Subject had been admitted to the hospital for a major surgical or medical procedure within 45 days of the planned procedure or was planning to undergo other major surgical or medical procedure within 45 days post implantation (e.g., coronary artery bypass graft, organ transplantation). This excluded any planned procedures for the prospective stent-graft placement (e.g., common carotid to left subclavian transposition/bypass, left carotid to axillary bypass, were acceptable. Carotid to carotid bypasses were not permitted). |

A comparison of the inclusion and exclusion criteria for the Relay® cohort and the surgical control cohort is described in Table 10-3. Differences in inclusion and exclusion criteria were intended to accomplish the following:

- reflect that the surgical cohort did not need to meet specific anatomical requirements necessary for implantation of the Relay® device;

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 40 of 98

{40}

- accommodate the different clinical follow-up practices for subjects undergoing surgical repair; and
- address the variances in clinical procedures typical for endovascular versus surgical operations.

Table 10-3: Comparison of Clinical Inclusion and Exclusion Criteria – Relay® and Surgical Control Cohorts

| Relay® | | Surgical Control |
| --- | --- | --- |
| Inclusion Criteria | | |
| | Criteria c – h | Not required for the surgical cohort since the anatomic requirements for proper implantation of the device are not essential |
| | Subject agreed to comply with 1-month, 6-month, and 1-year follow-ups, in addition to, an annual visit out to 5 years. | Prospective Subjects Only: Subject (or legally authorized representative) agreed to provide data from 1-month, 6-month, and 1-year follow-up visits and was encouraged to return annually out to 5 years.
Retrospective Subjects Only: Subject (or legally authorized representative) agreed to provide/release all available data surrounding subject’s surgical repair and/or use of historical data as permitted by institutional policies. |
| | Subject (or legally authorized representative) agreed to sign an ICF prior to treatment | Institutional requirements regarding informed consent of retrospective subjects were observed |
| Exclusion criteria | | |
| | Criteria b – i | Not required for the surgical cohort since the anatomic constraint for successful implantation of the device are not essential |
| | Subject was morbidly obese preventing adequate x-ray visualization of the aorta. | Not required for surgical subjects since x-ray visualization is not required. |
| | Subject had an untreatable allergy or sensitivity to contrast media or device components | Surgical subjects will not receive the Relay® device; therefore allergy to these components did not need to be assessed. |
| | Not applicable for endovascular procedures | Subject required hypothermic arrest, great vessel revascularization, or visceral debranching. |

PMA P110038: FDA Summary of Safety and Effectiveness Data

{41}

# 2. Follow-up Schedule

The first Relay® subject was enrolled in the study on January 23, 2007 (treatment date for first subject) and the last Relay® subject was enrolled on May 5, 2010. The 1-year visit for the last subject enrolled was on 25 April 2011. The prospective controls were enrolled between May 2007 and July 2009. Retrospective controls were treated between October 1998 and April 2007 and enrolled on the basis of chart review.

The study follow-up schedule for the Relay® cohort included clinical assessments at hospital discharge, 1, 6, and 12 months post-procedure and annual visits thereafter. The protocol-required imaging was provided to the core laboratory for assessment. For prospective surgical subjects, study assessments were similar. Data collected included baseline and demographic information, procedural information, and follow-up data at 1 month, 6 months, and 1-year. Annual follow-up visits (including a spiral CT scan at the 1-year follow-up visit) out to 5 years were strongly encouraged but not required. Follow-up data (e.g., 1-month, 6-month, and 1-year and annual data) for retrospective subjects was derived via medical chart review, using the hospital visits closest in time to date of the protocol prescribed follow-up regimen. Where possible, the imaging used to determine subject eligibility as well as a 1-year spiral CT scan was collected.

# 3. Clinical Endpoints

The analysis included clinically-relevant endpoints for patients with thoracic aortic pathologies. The endpoints used by Bolton Medical to demonstrate the safety of the device were adequate to describe the adverse events resulting from using the Relay® Thoracic Stent-Graft with Plus Delivery System. Similarly, the endpoints used by Bolton Medical to demonstrate the effectiveness of the device were adequate to demonstrate the treatment effect.

## a) Safety

The primary safety analysis compared the distribution of Relay® and surgical control subjects experiencing major adverse events (MAEs) within 1 year post-procedure. MAEs included aneurysm-related mortality, stroke, paralysis / paraplegia, myocardial infarction, procedural bleeding, respiratory failure, renal failure, and wound healing complications. The distribution of subjects experiencing at least 1 event in 1 year was compared for each group using the Kaplan-Meier method. The Kaplan Meier method estimates the probability of experiencing events over time. Since the probability of not having an event plus the probability of experiencing 1 or more events will sum to 1, then the probability of experiencing at least 1 event is calculated as 1 minus the probability of surviving (not having an event) within 1 year. The null hypothesis was that the probability of patients experiencing at least 1 major adverse event within 1 year is equivalent between both treatments using a two-sided alpha level of 0.05. Rejection of the null hypothesis would provide evidence that the probability of experiencing at least 1 major adverse event is not the same between the two treatments. The 1-year time-to-event distribution was compared between the Relay® and surgical control groups using the log-rank test.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 42 of 98

{42}

Several sensitivity analyses were conducted for the primary safety endpoint. An unadjusted Cox proportional hazards model was used to calculate the hazard ratio, its 95% confidence interval, and the p-value for treatment effect. A hazard ratio (Relay:Surgical) &lt; 1 and an associated p-value &lt; 0.05 were intended to provide evidence of superiority of the Relay treatment.

Additionally, adjustment for potentially confounding variables was based on the propensity score. This is a method of adjusting a comparative analysis for the biases caused by non-random treatment assignments, using logistic regression to assign a score to each individual based on the probability of being classified as part of the Relay group. The propensity score model included the following important covariates, which are known to affect patient outcome: age, gender, smoking status, maximum lesion diameter, coronary artery disease, renal function, chronic obstructive pulmonary disease, diabetes and history of stroke. In addition, any other baseline variables that were significantly different (p &lt; 0.10) between the groups were considered as possible covariates in a stepwise selection, forcing the variables above into the model. If more than 10% of the values are missing for any covariate, they were excluded from the model. In order to include all subjects in the analysis, missing values of any covariate in the final model were replaced by the treatment-group mean for that covariate. A Cox proportional hazards model adjusting for quintiles of the propensity score was used to calculate the hazard ratio, its 95% confidence interval, and the p-value for treatment effect.

## b) Effectiveness

The primary effectiveness endpoint was freedom from major device-related adverse events: endoleak (Types I, III and IV), stent migration (&gt; 10mm as compared to the 1 month visit), lumen occlusion, aneurysm rupture, and deployment failure/conversion to surgical repair occurring through 1-year post-procedure. The proportion of subjects in the Effectiveness sample who were free from major device-related AEs at 1-year post-procedure was compared against a performance goal of 0.80 using a 1-sided z-test (normal approximation to the binomial) at an alpha level of 0.025. Rejection of the null hypothesis would provide evidence that this performance goal (proportion-free greater than 0.80) was met.

## c) Secondary Endpoints

The secondary effectiveness analyses for major device-related AEs [endoleak (excluding Type II), stent migration (migration ≥ 10 mm as compared to the 1-month visit), lumen occlusion, aneurysm rupture, conversion to surgery] at the 1-month and 6-month follow-up visits were analyzed. The individual components of the primary endpoint are presented descriptively as event rates. Other secondary effectiveness endpoints included lesion measurement changes from the 1-month visit as compared with the 6-month and 1-year visits, device integrity failures, and vascular access complications.

The secondary safety analyses for the composite endpoint of MAEs (stroke, paraplegia, myocardial infarction, respiratory failure, renal failure, and aneurysm related mortality) at time points other than the 1-year follow-up (i.e., at the 1-month and 6-month follow-up

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 43 of 98

{43}

visits), as well as individual components of the composite endpoint, were compared between Relay® and surgical cohorts using Cox models. All-cause mortality was also analyzed.

In addition, clinical utility parameters (duration of procedure, transfusions required, length of hospital stay, time in the ICU) were compared between Relay® implantation and surgical procedures using $t$ tests for continuous data and chi-square tests for binary (yes/no) or categorical data.

## d) Success / Failure Criteria

The Relay® Thoracic Stent-Graft clinical study was considered successful if the null hypotheses for primary effectiveness and the primary safety endpoints (as described in sections X.A.3.a. and X.A.3.b.) were rejected.

## e) Pre-Specified Statistical Analysis Plan

### (1) Study Hypothesis

Analysis of the Relay® clinical trial results included hypothesis testing of both safety and effectiveness endpoints. Secondary effectiveness endpoints were presented as descriptive statistics. Secondary safety endpoints were also presented descriptively, and in addition, these data were subjected to a Cox proportional hazards analysis (as described in sections X.A.3.a and X.A.3.e.(7)) to evaluate differences in treatment effects. The Relay® Thoracic Stent-Graft clinical study would be considered a success if the null hypotheses of both the primary effectiveness and the primary safety endpoints (as described in sections X.A.3.a. and X.A.3.b.) were rejected.

### (2) Comparator

Safety data for the Relay® cohort was compared to a cohort of surgical control subjects treated at trial institutions within the past 10 years of site initiation. The surgical control cohort was a combination of prospectively- (n=7) and retrospectively-treated (n=53) subjects. In addition to covariate analysis, propensity score analysis was used to assess comparability of the groups. The control group was analyzed to justify the use of both retrospectively- and prospectively-enrolled patients.

### (3) Methodology

This study was designed as a non-adaptive frequentist trial. The sample size was fixed by design and not adapted as a function of preliminary results.

### (4) Sample Size Justification

The sample size for the Relay® cohort was driven by the primary effectiveness analysis. Assuming that the proportion of subjects remaining free from major

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 44 of 98

{44}

device-related adverse events is 0.90, 108 endovascular subjects (with one year follow-up) was intended to provide 80% power for a one-sided z-test (normal approximation to the binomial) at an alpha level of 0.025 against an alternative of 0.80. Accounting for an expected 10% loss to follow-up, 120 subjects yield 80% power.

The sample size for the surgical control group was based on the primary safety analysis. A log-rank test was proposed to compare the one-year time-to-event distributions of the Relay® and surgical control groups. A total of 108 subjects in the Relay® (endovascular) arm and 50 surgical subjects (both with one year follow-up) was intended to provide 90% power at a two-sided alpha level of 0.05 to detect a difference in the distribution of subjects experiencing major adverse events if the one-year event probabilities are 0.25 in the Relay® cohort and 0.50 in the surgical control cohort. Assuming withdrawal and loss to follow-up of 20%, approximately 60 subjects were required for the surgical control cohort.

## (5) Statistical Test

Hypothesis testing and other statistical testing was conducted as described in sections X.A.3.e.

## (6) Method for Accommodating Missing Data

In general, missing data were not imputed and analyses were based on available data. However, missing or incomplete adverse event start dates were imputed based on a predefined algorithm. Sensitivity analyses were conducted as part of the primary safety and effectiveness analyses.

In addition, tipping point analyses were conducted on the two primary endpoints to account for the impact of non-evaluable subjects. The start of the 1-year visit window (337 days) was used as the cutoff point for converting censored subjects to subjects with events. The lower bound of the 97.5% confidence interval (CI) for freedom from event was displayed using the following methods:

- Greenwood’s variance, loglog transformation
- Peto (Lao, 1995)

The CIs for event probabilities were calculated as 1-“lower bound” found for the event-free, which employs the failure variance as the survival variance. For Greenwood’s 97.5% one-sided lower bound, the lower bound of the two-sided 95% CI was used. For Peto, the za is the ath quantile of the standard normal cumulative distribution function (a=0.975).

The tipping point analysis of the primary effectiveness endpoint was performed utilizing the one-sided z-test and point estimates and the lower limit of the one-sided 97.5% CI. Primary effectiveness was based on all 120 subjects in the Relay® treatment. The tipping point analysis kept the denominator at 120 subjects. Starting with the 4 subjects identified who actually experienced a major

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 45 of 98

{45}

device-related adverse event (AE), subjects who did not have 1 year follow-up (21 subjects) were added one by one as having experienced an event at each increment. There were 21 stages provided that converted each of the 21 subjects without 1 year follow-up.

The inclusion of 12 additional subjects with a major device-related AE still provided a proportion free estimate of 0.867 and lower bound of the one-sided 97.5% CI of 0.806, meaning a total of 16 subjects out of the 120 subjects could have experienced a device-related AE before seeing failing results.

Tipping point analyses of the effectiveness endpoint via the Kaplan-Meier analyses were also performed, through repeatedly treating all censored subjects as having experienced a device-related event at the time of their leaving the study for any reason. Two approaches were taken – one with the tipping point analysis starting with the earliest (smallest) censored time and the other with the tipping point using the latest (largest) censored time. In each approach, subjects were added into the tipping point analysis based on their censored time and were converted from event-free to experiencing a major device-related AE. Subjects were imputed based on the days since procedure (censored time) and if more than one subject had the same day, the subject entered based on subject number.

The original Kaplan-Meier analysis conducted per the study protocol data showed the probability of remaining event-free through 1-year post-procedure as 0.96 with the two-sided 97.5% CI (Greenwood’s, loglog) of (0.89, 0.99). The analysis was repeated with the lower bound of the one-sided 97.5% CI using the two techniques described above. All results on these lower bounds are above 0.80 (Greenwood=0.902, Peto=0.925). Starting with the earliest (smallest) censored time yielded all CI’s to be in agreement and to be above 0.80 until a total of 14 subjects (the original 4 who actually experienced an event, plus 10 additional subjects) were considered to have experienced a major device-related AE (Greenwood=0.809, Peto=0.821). In contrast, starting with the latest censored time all CI’s were in agreement until a total of 13 subjects experienced a major device-related AE. At this stage, days since procedure ranged from 166 to 333 days and the CI’s were Greenwood=0.802 and Peto=0.819.

Tipping-point analyses of the primary safety endpoint were performed. These Kaplan-Meier analyses were repeated by progressively converting censored subjects’ times to first major adverse event times. As with the effectiveness analysis, the analyses were done starting with the earliest (smallest) censored time and also starting with the latest (largest) censored time. In addition, since the safety analysis considered the control group as well, an analyses was done with the Relay® treatment starting at the earliest (smallest) censored time and Surgical treatment starting at the latest (largest) time as well as with the tipping point starting at latest (largest) censored time within the Relay® treatment and earliest (smallest) censored time for Surgical treatment.

PMA P110038: FDA Summary of Safety and Effectiveness Data
Page 46 of 98

{46}

In each approach, subjects were added into the tipping point analysis based on their censored time and were converted from event-free to experiencing a major adverse event. Subjects were imputed based on the days since procedure (censored time) and if more than one subject had the same day, the subject entered based on subject number.

The original Kaplan-Meier analysis of the study data showed the probability of experiencing at least 1 major adverse event through 1-year post-procedure with the two-sided 97.5% CI (Greenwood’s, loglog) for Relay® treatment as 0.27 (0.19, 0.38) and for Surgical treatment as 0.51 (0.38, 0.67). The p-value for the log rank test was &lt;0.001, showing the distribution of Relay® subjects experiencing at least 1 major adverse event within 1 year post-procedure was significantly lower than the distribution of surgical control subjec…