NEUROPACE RNS SYSTEM

P100026 · NeuroPace, Inc. · PFN · Nov 14, 2013 · Neurology

Device Facts

• Record ID: P100026
• Device Name: NEUROPACE RNS SYSTEM
• Applicant: NeuroPace, Inc.
• Product Code: PFN · Neurology
• Decision Date: Nov 14, 2013
• Decision: APPR
• Device Class: Class 3
• Attributes: Therapeutic

Indications for Use

The RNS® System is an adjunctive therapy in reducing the frequency of seizures in individuals 18 years of age or older with partial onset seizures who have undergone diagnostic testing that localized no more than 2 epileptogenic foci, are refractory to two or more antiepileptic medications, and currently have frequent and disabling seizures (motor partial seizures, complex partial seizures and/ or secondarily generalized seizures). The RNS® System has demonstrated safety and effectiveness in patients who average 3 or more disabling seizures per month over the three most recent months (with no month with fewer than two seizures), and has not been evaluated in patients with less frequent seizures.

Device Story

Cranially implanted neurostimulator (RNS-300M) monitors ECoG activity via cortical/depth leads; detects abnormal patterns using programmable tools (area, line-length, bandpass); delivers responsive electrical stimulation to interrupt seizure patterns. Used in epilepsy centers by neurosurgeons/neurologists. Patient uses remote monitor at home to upload data to PDMS. Output allows clinicians to adjust detection/stimulation settings. Benefits include reduced seizure frequency for refractory patients.

Clinical Evidence

Pivotal study (N=191) was a randomized, double-blind, sham-controlled trial. Primary endpoint: reduction in disabling seizures during 12-week Blinded Evaluation Period (BEP) vs. baseline. Modified GEE analysis showed Treatment group had 37.9% reduction vs. 17.3% in Sham (p=0.012). Safety data from 256 subjects (903 patient-years) showed SAE rate comparable to historical controls (epilepsy surgery/DBS). No unanticipated device-related SAEs. 55% of subjects reaching 2 years post-implant achieved ≥50% seizure reduction.

Technological Characteristics

Titanium-enclosed neurostimulator with Li-CFx/SVO battery. Platinum/Iridium electrodes on silicone leads. Sensing: ECoG. Stimulation: constant current, rectangular biphasic charge-balanced pulses (1-333Hz, 40-1000μs). Connectivity: inductive telemetry via wand to programmer/remote monitor. Sterilization: 100% ethylene oxide. Software: proprietary detection algorithms (area, line-length, bandpass).

Indications for Use

Indicated for adults 18+ with partial onset seizures, refractory to 2+ antiepileptic medications, with 1-2 localized epileptogenic foci and frequent disabling seizures (average 3+/month, no month <2). Contraindicated for patients at high risk for surgical complications (infection, coagulation disorders, low platelet count), those with other brain-stimulating implants, or those unable to operate the remote monitor/magnet.

Regulatory Classification

Identification

For the treatment of epilepsy.

Submission Summary (Full Text)

{0}

SUMMARY OF SAFETY AND EFFECTIVENESS DATA (SSED)

I. GENERAL INFORMATION

Device Generic Name: Implanted brain stimulator

Device Trade Name: RNS® System

Device Procode: PFN

Applicant’s Name and Address: NeuroPace, Inc.
1375 Shorebird Way
Mountain View, California 94043

Date of Panel Recommendation: February 22, 2013

Premarket Approval Application (PMA) Number: P100026

Date of FDA Notice of Approval: November 14, 2013

Expedited: Not applicable

II. INDICATIONS FOR USE

The RNS® System is an adjunctive therapy in reducing the frequency of seizures in individuals 18 years of age or older with partial onset seizures who have undergone diagnostic testing that localized no more than 2 epileptogenic foci, are refractory to two or more antiepileptic medications, and currently have frequent and disabling seizures (motor partial seizures, complex partial seizures and/ or secondarily generalized seizures). The RNS® System has demonstrated safety and effectiveness in patients who average 3 or more disabling seizures per month over the three most recent months (with no month with fewer than two seizures), and has not been evaluated in patients with less frequent seizures.

III. CONTRAINDICATIONS

The RNS® System is contraindicated for:

- Patients at high risk for surgical complications such as active systemic infection, coagulation disorders (such as the use of anti-thrombotic therapies) or platelet count below 50,000.
- Patients who have medical devices implanted that deliver electrical energy to the brain.
- Patients who are unable, or do not have the necessary assistance, to properly operate the NeuroPace® Remote Monitor or Magnet.

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 1

{1}

- The following medical procedures are contraindicated for patients with an implanted RNS® System. Energy from these procedures can be sent through the implanted brain stimulation system and cause permanent brain damage which may cause severe injury, coma, or death. Brain damage can occur from any of the listed procedures even if the RNS® Neurostimulator is turned off or if the Leads are not connected to the Neurostimulator, and can occur even if the Neurostimulator has been removed, if any Leads (or any part of a Lead), or the cranial prosthesis remain.

- MR imaging is contraindicated for patients with an implanted RNS® System. Do not perform an MRI on a patient with any implanted RNS® Neurostimulator or Lead (or any portion of a Lead). Even if the Neurostimulator has been removed, the patient should not have an MRI if any part of a Lead or the Cranial Prosthesis is still implanted.

The RNS® System is MR Unsafe. Testing has not been performed to define conditions of use to ensure safety of the RNS® System in an MR environment.

- Diathermy procedures are contraindicated in patients implanted with an RNS® Neurostimulator and associated Leads. (Diathermy is any treatment that uses high-frequency electromagnetic radiation, electric currents, or ultrasonic waves to produce heat in body tissues.) Patients absolutely CANNOT be treated with any type of shortwave, microwave, or therapeutic ultrasound diathermy device whether or not it is used to produce heat. These treatments should not be applied anywhere on the body.

- Electroconvulsive Therapy (ECT) is contraindicated for patients with an implanted RNS® System.

- Transcranial Magnetic Stimulation (TMS) is contraindicated for patients with an implanted RNS® System.

IV. WARNINGS AND PRECAUTIONS

The warnings and precautions can be found in the RNS® System labeling.

V. DEVICE DESCRIPTION

The NeuroPace RNS® System includes a cranially implantable programmable neurostimulator that senses and records brain electrical activity. In response to the detection of previously identified patterns the neurostimulator is designed to deliver electrical stimulation to the brain to interrupt those patterns before the patient experiences clinical seizures. It is not a seizure detection device.

A. Implanted Components

The following are the implanted components of the RNS® System:

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 2

{2}

- RNS® Neurostimulator (model RNS-300M)

The RNS® Neurostimulator (model RNS-300M) contains electronic circuitry and a Lithium-carbon monofluoride/silver vanadium oxide (Li-CFx/SVO) battery that are hermetically sealed within a flat curved titanium enclosure. It is implanted within the cranium coplanar with the skull surface and is covered by the scalp. A Ferrule mechanically supports and secures the Neurostimulator in the skull. The Neurostimulator is connected to one or two Leads that are surgically placed in or near the epileptic seizure foci in the brain.

The neurostimulator monitors electrocorticographic (ECoG) activity and can be programmed to detect abnormal electrical activity. Three programmable detection tools (area, line-length, and bandpass) are provided. The detection tools are highly configurable and can be adjusted by the physician to optimize the detection for each individual patient. Up to two independent detectors can be programmed for any two sensing channels.

When detection criteria are met, the Neurostimulator delivers short trains of constant current, rectangular biphasic charge balanced pulses. Stimulation parameters can be programmed as follows:

Table 1: Stimulation Output Parameters

| Maximum Current Amplitude @ 500 Ω | 11.5 mA ± 10% |
| --- | --- |
| Maximum Voltage Amplitude @ 500 Ω | 6V ± 10% |
| Pulse Width | 40 – 1000 μs |
| Frequency | 1 to 333Hz |
| Pulses Per Burst | 1 to 1666 |
| Maximum Charge Density | 25 μC/cm²/phase |
| Current Path Options | Bipolar or Multipolar |

- NeuroPace® Cortical and Depth Leads

The NeuroPace® Leads provide an interface through which electrical activity of the brain can be sensed and recorded by the RNS® Neurostimulator and through which electrical stimulation can be delivered. Cortical Strip Leads are placed on the surface of the brain near the epileptic foci and Depth Leads are stereotactically introduced into epileptic foci in the brain. The lead specifications are provided in Table 2 below.

Table 2: Cortical and Depth Lead Specifications

| | Cortical Strip Leads | Depth Leads |
| --- | --- | --- |
| Lead Length | 15 cm, 25 cm, and 35 cm | 30 cm and 44 cm |
| Lead Diameter | 1.27 mm | 1.27 mm |
| Number of Electrodes | 4 | 4 |
| Electrode Arrangement | 1 x 4 array | 1 x 4 array |
| Electrode Material | Platinum/Iridium | Platinum/Iridium |
| Electrode Spacing | 10 mm | 3.5 mm and 10 mm |
| Electrode Surface Area | 0.079 cm² | 0.079 cm² |

PMA P100026: FDA Summary of Safety and Effectiveness Data

{3}

| | Cortical Strip Leads | Depth Leads |
| --- | --- | --- |
| Impedance | 15 cm: 15 Ω (+/- 10%) | 30 cm: 30 Ω (+/- 10%) |
| | 25 cm: 25 Ω (+/- 10%) | 44 cm: 44 Ω (+/- 10%) |
| | 35 cm: 35 Ω (+/- 10%) | |
| Lead Body Material | Silicone | Silicone |
| Electrode Material | Platinum/Iridium | Platinum/Iridium |

Figure 1: RNS® Neurostimulator and Cortical and Depth Lead

# B. External Components

The following are the external components of the RNS® System:

- NeuroPace® Programmer (model PGM-300)

The NeuroPace® Programmer is a laptop computer that runs proprietary NeuroPace® Programmer Application Software (model 3302 version 1.6.0.2) and utilizes a Wand (model W-02) to communicate with an RNS® Neurostimulator. The Programmer provides the clinician with a user interface to select and download detection and responsive stimulation settings to the neurostimulator, to view real-time ECoG signals, to test the RNS® System integrity, and to upload data and diagnostic information from the RNS® Neurostimulator for viewing.

- NeuroPace® Remote Monitor (model DTR-300)

The NeuroPace® Remote Monitor is a home-use monitoring device that utilizes Wand (model W-02) to communicate with an implanted RNS® Neurostimulator. The Remote Monitor is provided to a patient or caregiver to collect data from the implanted RNS® Neurostimulator and to upload these data using analog telephone lines by way of a secure connection to the Patient Data Management System (PDMS).

PMA P100026: FDA Summary of Safety and Effectiveness Data

{4}

- NeuroPace® Patient Data Management System (model 4340)
The NeuroPace® Patient Data Management System (PDMS) is used for storage and access to historical Neurostimulator and patient data. During synchronization of the Programmer or Remote Monitor with the PDMS, Neurostimulator information regarding detections and stimulations, as well as stored ECoG recordings and Neurostimulator self-diagnostic information are uploaded automatically to the PDMS and combined with previously uploaded information.

C. Accessories

The following accessories are provided with the RNS® System:

- Cranial Prosthesis - Occupies a vacant Ferrule if the Neurostimulator has been explanted and not replaced.
- Connector Cover (Model CC-01) - Secures the proximal lead contacts to the neurostimulator.
- Connector Plug (Model CP-01) - Are used to fill all vacant ports in the Connector Cover.
- Craniectomy Template - May be used as a pattern to mark and delineate the shape of the Ferrule on the skull prior to making a craniectomy.
- Ferrule and Ferrule Clamp (Model F-01 and Model FC-01) - Installed in a craniectomy to secure and mechanically support the RNS® Neurostimulator in the skull. The Ferrule Clamp is used to secure the neurostimulator to the ferrule.
- Lead Strain Relief (Model LSR-01) - Supports the proximal end(s) of the lead(s) at their exit from the Neurostimulator Connector Cover, protecting the Lead from stress near the connector.
- Magnet (Model M-01) - When placed over the implanted RNS® Neurostimulator, suppresses therapy as long as the Magnet is in position and, if the neurostimulator is programmed to do so, triggers ECoG storage.
- Torque Driver (Model TD-01) - Used to tighten the screw that secures the Connector Cover to the Neurostimulator and to tighten the Ferrule Clamp that secures the Neurostimulator to the Ferrule.
- Tunneling Tool (Model TT-01), Tunneling Tool Tip (Model TTT-01), Tunneling Straw (Model TTS-01) - Used to tunnel an implanted Lead from its cranial exit point, through a sub-galeal pathway, to the implanted RNS® Neurostimulator location.

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 5

{5}

- Lead Cap (Model LC-01) - The Lead Cap protects the proximal end of a lead when it is not connected to the RNS® Neurostimulator.
- Stop Gauge (Model SG-01) - Placed on a Depth Lead prior to implantation to indicate the appropriate depth of its insertion.
- Suture Sleeve (Model SS-01) - Protects the lead body when sutures are used to secure a lead.

## VI. ALTERNATIVE PRACTICES AND PROCEDURES

There are currently three alternative modalities available for the treatment of epilepsy: antiepileptic drugs (AEDs), vagus nerve stimulation, and resective epilepsy surgery. Antiepileptic medications are tried first, usually in monotherapy. If the first AED is not effective, alternative AEDs are tried alone or in polytherapy. In people with epilepsy for whom medications are not effective or who have unacceptable medication related side effects, vagus nerve stimulation or resective neurosurgery may be an option. Vagus nerve stimulation therapy is adjunctive to AED therapy. Neurosurgery for the treatment of epilepsy usually requires removal of some portion of the brain.

## VII. MARKETING HISTORY

The RNS® System has not been marketed in the United States or any foreign country.

## VIII. POTENTIAL ADVERSE EFFECTS OF THE DEVICE ON HEALTH

Potential adverse effects associated with the RNS® System include those related to the implantation procedure, those related to performance of the Neurostimulator and Leads and those related to long-term patient tolerance of the implant. Adverse effects which may potentially occur, but were not reported in the clinical trials for the RNS® System, include the following:

- Allergic reaction to the implanted material
- Brain abscess

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

## IX. SUMMARY OF PRECLINICAL STUDIES

### A. Laboratory Studies

1. RNS® Neurostimulator:

The RNS® Neurostimulator underwent numerous testing for electrical safety, output characterization, dimensional verification, hermeticity, environmental conditions, mechanical verification, battery safety and validation, and x-ray interaction. Key testing on the neurostimulator is summarized in Table 3 below. Testing demonstrated the RNS® Neurostimulator operated according to specifications after exposure to the tested conditions (i.e., passed testing).

PMA P100026: FDA Summary of Safety and Effectiveness Data

{6}

Table 3: Summary of key testing performed and passed on the RNS® Neurostimulator

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| Output Characterization | Verify proper output (amplitude, pulse width, frequency, etc.) and detection parameters of the IPG function are within specified tolerances | Device output is within specifications under expected temperature (35-40°C) and loads (500 to 1200 Ω). |
| DC Leakage Current | Verify the leakage current is in an acceptable range | Per ISO 14708-1:2000, part 16 (modified, requirement limit is stricter than standard (<0.5 μA) |
| Integrated Circuits (IC) | Verify the proper functioning of the ICs including | ICs function per specifications. |
| Dimensional | Verify that device meets dimensional requirements | Physical Inspection per ISO 14708-1:2000, 15.2 - Device meets geometric requirements for thickness and external features. |
| Helium leak | Verify feedthroughs remain hermetic after mechanical loading | Helium leak per Mil Std 202, Method 112, Condition C, Procedure 1 |
| | Verify neurostimulator hermetic seal | Neurostimulator hermetic seal has a leak rate no greater than 5.0 x 10-9 cc-atm/s of helium per MIL-STD-883. |
| Environmental | Verify device conforms to functional requirements and is not damaged by temperature change and thermal shock | Testing per EN 45502-1: 1997, 26.2; IEC 60601-1; IEC 60068-2-14 test Nb, and ASTM D 4169-99, 15.2. Confirm devices continue to meet visual, hermeticity, fine leak and functional requirements after stress. |
| | Verify device conforms to functional requirements and is not damaged by mechanical loads (shock, vibration and atmospheric pressure change) | Testing per ISO 14708-1:1997, 23.2 (Note that test was performed with lower frequency limit changed from 5Hz to 10Hz), CEI/IEC 60068-2-47 & 60068-2-64, ASTM D 3332-99 and ASTM D 4169-99. Confirm devices continue to meet visual, hermeticity fine leak and functional requirements after stress. |
| Drop Test | Verify the device performance is not affected by being dropped | Expose to a mechanical shock equivalent to a 19.5" drop. Confirm devices continue to meet visual, hermeticity fine leak and functional requirements after stress. |
| Temperature rise limit during single fault condition | Temperature rise should not cause burns during single fault conditions | Temperature rise is less than or equal to 2°C limit during single fault conditions per EN 45502-1: 1997 17.1. |
| Battery | Battery Capacity Verification (Longevity). | Battery longevity testing using maximum and medium use parameters from clinical study to estimate battery longevity. |

PMA P100026: FDA Summary of Safety and Effectiveness Data

{7}

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| | Electrical, Visual, Dimensional, Hermeticity, Short Circuit Testing, Environmental, and Forced Discharge Tests | Testing fulfills the requirements of UN Recommendations on Transport of Dangerous Goods, Manual of Tests and Criteria, 4th revised edition, section 38.3. |
| X-ray Imaging | Evaluate the safety and functionality after x-ray imaging and identification of radiographic markings. | Device remains functional after exposure to x-ray; radiographic marker is visible in x-ray; and minimal to no distortion of anatomical features adjacent to device. |
| Particulate Matter | Verify there is no unacceptable release of particulate matter when the device is used as intended | Per EN 45502-1, 14.2. |

# 2. Depth and Cortical Leads:

The depth and cortical leads underwent numerous testing for dimensional verification, electrical safety, environmental conditions, mechanical verification, and x-ray interaction. Key testing on the leads is summarized in Table 4 below. Testing demonstrated the depth and cortical leads operated according to specifications after exposure to the tested conditions (i.e., passed testing).

Table 4: Summary of key testing performed and passed on the depth and cortical leads

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| Dimensional | Verify that the lead geometry (diameter, length, distal lead configuration) meet design specifications | Depth lead geometry (diameter and length) meet design specifications |
| DC Resistance and Electrical Isolation | Verify protection due to electricity | Testing per ISO 14708-3:2008, 6.102 and 16.3 |
| Environmental | Verify lead remains electrically functional after exposure to pressure changes. | Testing per ISO 14708-1:2000, section 25.1. Verify lead is electrically functional (measure continuity and leakage) and performs per specifications. |
| | Verify lead remains electrically functional after exposure to temperature changes | Per ISO 14708-1:2000, section 26.2. Verify lead is electrically functional (measure continuity and leakage) and performs per specifications. |
| Lead Tensile Strength | Lead remains electrically functional after exposure to tensile stressors | Testing per ISO 14708-1:2000, section 23.3. Verify lead is electrically functional (measure continuity and leakage) and performs per specifications. |
| Flexural Fatigue | Lead conductors do not | Testing per EN45502-1, clause 23.3 and 23.4. |

PMA P100026: FDA Summary of Safety and Effectiveness Data

{8}

| Test | Purpose | Acceptance Criteria |
| --- | --- | --- |
| Testing | fatigue after flexural stressors | Verify lead is electrically functional (measure continuity and leakage) and performs per specifications. |
| Connector Assembly Lead Insertion/Extraction Force Test | Confirm insertion/extraction forces meet specifications | Insertion and extraction forces before and after preconditioning shall not exceed limits as follows:
• connector insertion force ≤ 1N
• extraction force ≤ 2N |
| Connector Assembly Lead Fixation Test | Connector lead fixation (retention) force | The assembly shall be subjected to straight separating pulls of 4 N +/- 0.5 N, for 60 sec without lead slippage. The assemblies are then tested to failure. |
| Connector Electrical Isolation | Verify protection due to electricity | Testing per ISO 14708-3:2008, 6.102 and 16.3 |
| Particulate Matter | No unacceptable release of particulate matter when the lead is used as intended | Test per EN 45502-1, 14.2. |

3. NeuroPace® Programmer (model PGM-300), NeuroPace® Remote Monitor (model DTR-300), NeuroPace® Patient Data Management System (PDMS), and Programming Wand (W-02)

The model PGM-300 Programmer, model DTR-300 Remote Monitor, and model 4340 Patient Data Management System (PDMS) underwent software development and verification testing in accordance with the Food and Drug Administration (FDA) guidance, entitled, “Guidance for the Content of Premarket Submissions for Software Contained in Medical Devices” (May 11, 2005) and all requirements were met. Electrical and mechanical verification of the Wand included leakage current and dielectric strength of insulation (per CEI/IEC 60601-1 Second Edition 1998-12), USB 2.0 protocol compliance, telemetry communication protocol compliance, temperature, humidity, liquid ingress and mechanical stresses (per CEI/IEC 60601-1 Second Edition 1998-12), and model W-02 drop test (drop from a height of $\approx 7$ feet) and performed according to their specified requirements.

4. Electromagnetic Compatibility (EMC) and Wireless Technology

EMC and wireless technology testing was performed using appropriate essential performance criteria in accordance with the relevant clauses of the following standards and met specified acceptance criteria except for essential performance criteria associated with radio-frequency identification (RFID) and the electrocorticographic (ECoG) sensing feature of the device as discussed below:

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 9

{9}

○ IEC 60601-1-2: 2007, “Medical electrical equipment - Part 1-2: General requirements for basic safety and essential performance - Collateral standard: Electromagnetic compatibility - Requirements and tests”
○ ISO 14708-3:2008(E): Implants for surgery – Active implantable medical devices – Part 3: Implantable neurostimulators”, Part 27
○ Wireless radio testing per United States FCC CFR Title 47 Part 2 and 15

The System performed as specified except for the electrocorticographic (ECoG) sensing feature of the device when in proximity to radio-frequency identification (RFID) devices. The ECoG recording feature of the System can be affected when exposed to RFID resulting in a sensing artifact that might deliver stimulation therapy to the patient based on the device setting. The stimulation will be in a safe range and will be the same as the output stimulation parameters that are programmed by the physician. Thus, a precaution was placed in the labeling advising the physician and patient of this risk. The conclusion of the testing was that the RNS® System met its essential performance criteria with the necessary contraindications, warnings, precautions, and additional information included in the labeling for the clinician and patient.

5. Sterility

The RNS® System components are terminally sterilized utilizing 100% ethylene oxide (EO) gas with heated aeration to allow for residual sterilant dissipation. The EO sterilization cycle has been validated according to ISO 11135-1: 2007. Sterilization of health care products – Ethylene oxide – Part 1: Requirements for development, validation, and routine control of a sterilization process for medical devices. The results obtained from the sterilization validation studies show that the sterilization process provides a Sterility Assurance Level (SAL) of 10⁻⁶. The sterile device components were tested successfully to meet the most rigorous category (permanent contact) for allowable limits of EO and Ethylene Chlorohydrin (ECH) residuals, as specified in ISO 109937: 2008. Biological Evaluation of Medical Devices – Part 7: Ethylene Oxide Sterilization Residuals. The amount of bacterial endotoxins on fully configured RNS kits and fully configured Leads kits was verified using Limulus Amebocyte Lysate (LAL) testing and found to be within the specification limit of &lt; 0.06 EU/ml (or &lt; 2.15 EU/device), as indicated for devices in contact with cerebrospinal fluid.

6. Packaging and Shelf-life

Packaging and shelf life validation tests for the sterile RNS® System products were successfully completed per AAMI / ANSI/ ISO 11607-1:2006. Packaging for terminally sterilized devices – Part 1: Requirements for materials, sterile barrier systems and packaging systems. Accelerated aging, transportation and handling studies met the requirements for sterile packaging, protection of components, and product functional testing. Shelf-life for the RNS® Neurostimulator has been established as 9 months from the date the battery is attached. All other sterile products including the Connector Cover Kit,

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 10

{10}

Craniectomy Template Kit, Ferrule Kit, Cranial Prosthesis Kit, Lead and Lead Accessory Kits have a 3-year shelf life from the date of sterile packaging.

# 7. Biocompatibility

Biocompatibility testing was performed on the finished, sterilized devices for all patient-contacting components of the RNS® System in accordance with ISO 10993-1 Biological evaluation of medical devices – Part 1: Evaluation and testing within a risk management process. All biocompatibility studies were conducted in compliance with Good Laboratory Practices (GLP), 21 CFR Part 58. The implanted components and accessories are considered permanent (&gt;30 days) implants in contact with tissue/bone and all non-implantable tools to support surgical implantation have limited contact duration (≤24 hours) with tissue/bone. The biocompatibility test data are summarized in Table 5 below. All prespecified acceptance criteria were met and all tests passed.

Table 5: Biocompatibility Test Data on the Implantable Components and Accessories and Non-implanted tools of the RNS® System

| Test Performed (Applicable Standard) | Acceptance Criteria |
| --- | --- |
| Implanted Components and Accessories & Non-implanted Tools: | |
| In Vitro Cytotoxicity (ISO 10993-5) | ISO Elution Method – 1X MEM Extract: Reactivity grade is not greater than mild reactivity (Grade 2). |
| Irritation and Delayed-type Hypersensitivity (ISO 10993-10) | Guinea pig Maximization Sensitization: Grades < “1” in the test group provided grades of < “1” are observed on the control animals. (If grades of ≥ “1” are noted on control animals, then the reactions of the test animals which exceed most severe control reaction are presumed to be due to sensitization.) |
| | Intracutaneous Reactivity: The difference between the test article and the control mean score is ≤ 1.0 (negligible or slight). |
| Acute Systemic Toxicity (ISO 10993-11) | None of the test animals show a significantly greater biological reaction than the controls |
| Implanted Components and Accessories: | |
| Genotoxicity (ISO 10993-3) | Bacterial Reverse Mutation Assay: There is less than 2-fold increase in the number of revertants when compared to the solvent control in strains TA97a, TA100, and TA102 and less than 3-fold increase in the number of revertants when compared to the solvent control in strains TA98 and TA1535. |
| | In vitro Chromosomal Aberration: There is no statistically significant difference in aberrations between the test group and the negative control. |
| | Mouse Bone Marrow Micronucleus: There is no statistically significant increase in the number of micronucleated polychromatic erythrocytes (PCEs) in the test group as compared to the concurrent negative control. |
| Interactions with Blood (ISO 10993-4) | Hemolysis - Indirect Contact (Extract test on depth lead): The hemolytic index of the test article extract is ≤ 2%. |
| Local effects after implantation (ISO 10993-6) | Implantation – Rabbit Intramuscular Implantation Study (13 weeks): An overall interpretation of the degree of biocompatibility exhibited by the test article based on gross and microscopic analysis comparing test to control article(USP high density polyethylene reference standard), as well as clinical observations. |

PMA P100026: FDA Summary of Safety and Effectiveness Data

{11}

| Test Performed (Applicable Standard) | Acceptance Criteria |
| --- | --- |
| Systemic Toxicity (ISO 10993-11) | Subchronic Toxicity: The correlation of all data for patterns of toxicity, including death of > 1 animal/group, mean body weight loss for each group, clinical signs of toxicity in > 1 animal/group, hematological and clinical chemistry values, and histopathology of tissues |
| | Material-mediated Pyrogenicity: No rabbit shows an individual rise in temperature of 0.5°C or more above the baseline temperature. |
| Combined implantation and systemic toxicity (ISO 10993-6 and -11) | Combined Neuroimplantation/Chronic Toxicity Study in Rabbits: See section below for details. |
| Carcinogenicity | An adequate carcinogenicity risk assessment was provided. |

Combined Neuroimplantation/Chronic Toxicity Study in Rabbits: The objectives of this study (conducted in accordance with GLPs) were to assess potential neurotoxicity, acute and chronic local tissue responses as well as long-term systemic effects following implantation of the device in rabbits. The study was based on the testing recommendations in the International Organization for Standardization (ISO) 10993: Biological evaluation of medical devices, Part 6: Tests for local effects after implantation, ISO 10993 – Part 11: Tests for systemic toxicity, and ASTM F2901 – 12, Standard Guide for Selecting Tests to Evaluate Potential Neurotoxicity of Medical Devices. The study design is summarized in Table 6 below.

Table 6: Combined Neuroimplantation/Chronic Toxicity Study in Rabbits Study Design Summary

| Termination Interval | Number of Animals | | | | Study Objective | Articles Implanted |
| --- | --- | --- | --- | --- | --- | --- |
| | Control | | Test | | | |
| | Male | Female | Male | Female | | |
| 6 Days | 4 | 4 | 4 | 4 | Brain tissue reaction & Neurotoxicity | Partial Depth Lead (T1), Partial Cortical Strip Lead (T2), Partial Lead body material section (T3), Ferrule (partial coupon) (T4)1 |
| 4 Weeks | 5 | 5 | 5 | 5 | Brain tissue reaction & Neurotoxicity | T1, T2, T3, and T4 |
| | | | | | Systemic toxicity & Local subcutaneous tissue reaction | Neurostimulator (T5)2, Depth Lead (T6), Cortical Strip Lead (T7), Burr Hole Covers (T8) |
| 26 Weeks | 7 | 7 | 7 | 7 | Brain tissue reaction & Neurotoxicity | T1, T2, T3, and T4 |
| | | | | | Systemic toxicity & Local | T5, T6, T7, and T8 |

PMA P100026: FDA Summary of Safety and Effectiveness Data

{12}

| Termination Interval | Number of Animals | | | | Study Objective | Articles Implanted |
| --- | --- | --- | --- | --- | --- | --- |
| | Control | | Test | | | |
| | Male | Female | Male | Female | | |
| | | | | | subcutaneous tissue reaction | |

1 The adapted Ferrule (partial coupon) was a finished device modified for implantation using only manufacturing methods and tooling materials used in finished device manufacturing.
Neurostimulator includes the Upper Strain Relief, Connector Cover, Ferrule, Ferrule Clamp, and Connector Plugs.

For purposes of determining neurotoxicity and brain tissue reaction, the implant location of the control articles replicated the location of the test articles. Table 7 below provides a summary of the examinations performed and the study results.

Table 7: Results of Combined Neuroimplantation/Chronic Toxicity Study in Rabbits

| Examination | Timing | Results |
| --- | --- | --- |
| Clinical Signs of Disease or Abnormality | 2x/day: for general health prior to surgery, weekly At termination: detailed examinations | No clinically significant findings or signs of toxicity were noted |
| Neurological | Days 1-5: daily; Days > 5: weekly; and Prior to termination | No abnormal findings were noted |
| Body Weight | Prior to implantation, weekly for the first 4 weeks, every 4 weeks thereafter, and prior to termination | Clinically acceptable following treatment at 6 days and 4 and 26 weeks No statistically significant differences w/ controls at 4 and 26 weeks and within group |
| Blood Hematology & Clinical Chemistry | 4 and 26-week termination | Hematology: no biologically significant differences between the test and control groups and all mean values were within an acceptable range Clinical Chemistry: no biologically significant differences between the test and control groups for any of the clinical chemistry parameters |
| Necropsy | 4 and 26-week termination | No changes that could be attributed to test articles |
| Organ weights and organ/body weight ratios | 4 and 26-week termination | No biologically significant differences between the test and control groups |
| Macroscopic implant site evaluation | 6-day, 4 and 26-week termination | No test device-related differences in macroscopic observations were detected between those that received the control HDPE implants in brain and subcutaneous locations and those that received various components of the test device in brain |

PMA P100026: FDA Summary of Safety and Effectiveness Data

{13}

| Examination | Timing | Results |
| --- | --- | --- |
| | | and/or subcutaneous locations. |
| Histopathology† | 6-day, 4 and 26-week termination | Microscopically, no evidence for systemic toxicity, neurotoxicity, or local tissue reaction occurred beyond the expected effects due to surgical placement and the physical presence of the various implants in the brain, the calvarium, or the subcutaneous tissue. |

Hematoxylin and Eosin (H &amp; E), Fluoro-Jade B (for evidence of neuronal degeneration), Anti-Glial Fibrillary Acidic Protein (GFAP) antibody (astroglial activation), and Mouse Monoclonal Anti-Rabbit Macrophage (RAM11) antibody (macrophages and activated microglia) were used for histopathological evaluation of the brain/dural implant sites. For histopathological evaluation of the subcutaneous implant sites and the designated tissues, H &amp; E stain was used

## B. Animal Studies

### Lead Implantation Study in Sheep

Testing on sheep was performed to examine the safety of the NeuroPace® Depth Leads and NeuroPace® Cortical Strip Leads in a simulated use condition. Two depth and 2 cortical leads were implanted into each sheep. One of each lead type was then stimulated for a specified amount of time at prescribed intervals over the course of the survival period. Longer continuous stimulation periods were used than would be experienced in human use with the RNS® System. The other of each lead type had lead impedance measurements taken at implantation and explantation.

The chronic implant duration of 11 sheep ranged from 33 days to 200 days, with a mean of 131 days. No histological analysis was performed on 4 sheep which were sacrificed early due to hardware problems relating to externalizing the leads for the scheduled stimulations and two animals that were inadvertently frozen. Five other animals experienced hardware-related complications but were survived until 18-24 weeks (135 - 200 days).

Injury to the neuronal tissue immediately adjacent to the leads was as expected and the reaction did not appear to extend into the surrounding tissue. No significant neuronal disorganization or necrosis was observed. Tissue reactions included chronic inflammation, astrocytic gliosis, some foreign body giant cells, and a fibrous capsule around the implantation track. These are not unexpected. The cortical strip leads appeared to result in no detectable cytoarchitectural changes to underlying tissue. Thirty (30) ECoG recordings were examined and the magnitudes of the signals provided sufficient characterization of brain activity for monitoring purposes. No evidence of electrode-to-tissue sensor block was observed.

## C. Additional Studies

The detection algorithm was not evaluated for its efficacy in accurately identifying specified ECoG activity and seizures. Verification and validation testing provided for the detection algorithm was determined by the FDA to be adequate for proceeding

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 14

{14}

with the IDE feasibility and pivotal studies (i.e., the risks to the subjects for participating in the studies were outweighed by the anticipated benefits to the subjects and the importance of the knowledge to be gained). Testing of the detection algorithm included the following:

- Evaluation of the algorithm in detecting epileptiform activity using artificial ECoG waveforms, using a MATLAB simulator. The detection tool met predefined success criteria.
- Evaluation using archived ECoG data from 11 consecutive patients admitted to the Emory Epilepsy Monitoring Unit (between January 1997 and May 1999). The majority of seizures were of mesial or neocortical temporal origin, but seizures recorded in 3 of the subjects originated extratemporally. Detection tools were programmed individually for patients and training sets were used for tuning the detectors in some cases. Of 125 seizures marked by an epileptologist all but 2 were detected (98.4% sensitivity) and the average false positive rate was 0.013 (range: 0-0.051) per hour.
- As part of software verification testing simulated data were used to test the minimum, maximum, and at least one intermediate value for each detection parameter of each detection tool by performing simulations using real-time ECoG data. The device met specifications.
- A multi-center feasibility IDE clinical investigation (G010288) was performed using an external model (eRNS) of the RNS® System that incorporated the same detection algorithm as the RNS® System. The study enrolled at 8 sites 125 subjects who were candidates for epilepsy surgery and were under-going video-EEG monitoring in an epilepsy monitoring unit. The study demonstrated that the eRNS could safely deliver electrical stimulation in response to detected ECoG activity. However, the study was not designed to determine specificity and sensitivity of the detection algorithm.

X. SUMMARY OF PRIMARY CLINICAL STUDIES

The sponsor performed a Feasibility study, a Pivotal study, and a Long-term Treatment (LTT) study (ongoing) as described below.

Feasibility Study

A Feasibility study was performed to evaluate preliminary safety and effectiveness and the results were used to inform the design of the Pivotal Study and to assess the integrity of the blind. Additionally, the safety data from the Feasibility study were included in the primary safety analysis for this PMA. It was a multi-center clinical investigation of individuals with medically intractable epilepsy. The first subject was enrolled on January 19, 2004 and the last subject transitioned to the Long Term Treatment (LTT) study (see below) on December 17, 2007. Eligible subjects were 18-65 years of age with medically intractable partial onset seizures and a minimum of 4 simple partial seizures (motor or sensory), complex partial seizures, and/or secondarily generalized seizures in each of the

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 15

{15}

previous three months. Subjects were required to be on a stable AED regimen and to have previously undergone diagnostic testing that localized one or two epileptogenic region(s). Subjects with psychogenic or non-epileptic seizures, status epilepticus, active psychosis, severe depression, or suicidal ideation within the preceding year were excluded. Sixty-five subjects were implanted with the RNS® Neurostimulator and Leads in the Feasibility study. The first four subjects implanted with the RNS® Neurostimulator and Leads at a clinical site participated in an open label protocol (all subjects received responsive stimulation), and subsequent subjects at that site participated in a randomized, double-blind, concurrent sham-stimulation control protocol in which the Treatment group received stimulation and Sham group did not. Forty-two (42) subjects were in the open label protocol and 23 were in the blinded protocol. Following completion of the 16 week Evaluation Period, subjects transitioned to an Open Label Period, and all subjects were able to receive responsive stimulation.

## Pivotal Study

A Pivotal study was performed to establish a reasonable assurance of safety and effectiveness of the RNS® System in reducing the frequency of seizures in individuals 18 years of age or older with partial onset seizures who have undergone diagnostic testing that localized no more than 2 epileptogenic foci, are refractory to two or more antiepileptic medications, currently have frequent and disabling seizures (motor partial seizures, complex partial seizures and/or secondarily generalized seizures, and in patients who average 3 or more disabling seizures per month over the three most recent months (with no month with fewer than two seizures) in the US under IDE# G030126. Data from this clinical study combined with safety data from the Feasibility study were the basis for the PMA approval decision. Patients were enrolled in the Pivotal trial beginning on December 29, 2005 and the last subject finished the blinded evaluation period (BEP) and transitioned to the Open label period on October 16, 2009. A summary is presented below.

## Long-term Treatment Study (LTT)

The LTT is an ongoing open label, multi-center, prospective clinical investigation of two-hundred and thirty (230) subjects who consented to enroll once they completed the Feasibility or Pivotal study. Each subject participates for a maximum of 7 years. Adverse event and seizure data are collected at 6-month intervals, and data regarding quality of life are collected at yearly intervals. AED adjustments are permitted as needed. The first subject was enrolled on April 6, 2007 and the study is ongoing.

## A. Study Design

The database for this PMA includes data from all 3 studies (Feasibility, Pivotal, and LTT) and reflects data collected through May 12, 2011 and includes 191 subjects in the Pivotal trial and 65 in the Feasibility study. Note that data for deaths, including Sudden Unexplained Death in Epilepsy (SUDEP), are current through October 24, 2012. There were 12 investigational sites in the Feasibility study, 28 investigational sites in the Pivotal study, and 29 investigational sites in the LTT study.

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 16

{16}

The RNS® System Pivotal study was a randomized, double-blinded, multi-center, sham-controlled clinical study. The investigation had five periods: the Baseline Period (which includes the Pre-Implant Period defined in Section 2 below), Post-Operative Stabilization Period, Stimulation Optimization Period, Blinded Evaluation Period (BEP), and Open Label Period. Enrolled subjects were implanted with the RNS® Neurostimulator and Leads within 28 days following the date of qualification for implantation. Subjects were randomized 1:1 at the end of the Post-Operative Stabilization Period (4 weeks post-implant). To ensure equal representation in the two therapy groups, an adaptive randomization approach (minimization) was used to balance variables that might influence the clinical response to responsive stimulation. These variables (listed in order of priority) were:

1) Investigational site;
2) Seizure onset zone location (partial onset seizures of mesial temporal origin versus partial onset seizures arising from any other region of the cortex);
3) Number of seizure foci (unifocal versus bifocal); and
4) Previous therapeutic epilepsy surgery (resection, subpial transection and/or corpus callosotomy).

Subjects randomized to the Treatment group received responsive stimulation during the Stimulation Optimization and BEP and subjects randomized to the Sham group did not receive responsive stimulation during these periods. Following completion of the BEP (20 weeks post-implant), subjects transitioned to the Open Label Evaluation Period and both Treatment and Sham group subjects received responsive stimulation.

The Pivotal study was designed to have 80% power with an overall 2-sided Type 1 error of 0.05, assuming responder rates (i.e., subjects with a 50% or greater reduction in seizures from baseline) in the Treatment group and Sham groups of 40% and 20%, respectively. To meet these criteria, 180 subjects were required in the BEP. Assuming approximately 10% of subjects would not be compliant (including subjects who did not complete the BEP), approximately 200 subjects were to be randomized, 100 each into the Treatment and Sham groups.

An independent Data Monitoring Committee (DMC) for the Feasibility, Pivotal, and LTT studies was established. The DMC was responsible for independently monitoring the safety of interventions during the investigation by reviewing data made available by NeuroPace acting in the capacity of the Coordinating Center. The DMC made recommendations to NeuroPace about safeguarding the interests of trial participants and about stopping, modifying or continuing the investigation. Information regarding all deaths that occurred during the investigation, the SUDEP Analysis Committee's classification with respect to SUDEP, as well as the data supporting that classification, was communicated to the Chair of the Data Monitoring Committee (DMC) by NeuroPace. The DMC reviewed composite safety and effectiveness data on a regular basis depending on subject enrollment, at a minimum of every six months.

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 17

{17}

# 1. Clinical Inclusion and Exclusion Criteria

The inclusion and exclusion criteria for the Feasibility and Pivotal studies were similar (the key inclusion and exclusion criteria for the studies are presented in Table 8 and Table 9).

Table 8: Key Inclusion Criteria

| Inclusion Criteria | Feasibility Study | Pivotal Study |
| --- | --- | --- |
| Subject has simple partial motor seizures, complex partial seizures and/or secondarily generalized seizures | Yes1 | Yes |
| Seizure counts per month | 4 or more2 | average of ≥ 33 |
| Age | 18-65 years | 18-70 years |
| Subject has seizures that are severe enough to cause injuries or significantly impair functional ability in domains including employment, psychosocial, education and mobility. | Yes | Yes |
| Subject has seizures that are distinct, stereotypical events that can be reliably counted | Yes | Yes |
| Subject failed treatment with a minimum of two AEDs (used in appropriate doses) with adequate monitoring of compliance and the effects of treatment. | Yes | Yes |
| Subject has remained on the same AED(s) over the preceding three (3) months | Yes | Yes |
| Subject has undergone diagnostic testing that has established the epileptiform activity onset region(s) | Yes | Yes, ≤ 2 epileptogenic regions |

1 The Feasibility study also included simple partial sensory seizures.
2 Subject has a minimum of four (4) or more countable seizures every month over the last three (3) months.
3 Subject has an average of three or more disabling simple partial seizures, complex partial seizures, or secondarily generalized seizures per month (28 days) over the three most recent months, with no month with less than two seizures.

Table 9: Key Exclusion Criteria

| Exclusion Criteria | Feasibility Study | Pivotal Study |
| --- | --- | --- |
| Subject has been diagnosed with psychogenic or non-epileptic seizures in the preceding year. | Yes | Yes |
| Subject has been diagnosed with primarily generalized seizures. | Yes | Yes |
| Subject has experienced unprovoked status epilepticus in the preceding year. | Yes | Yes |
| Subject has a clinically significant or unstable medical condition or a progressive central nervous system disease. | Yes | Yes |
| Subject is taking anticoagulants. | Yes | Yes |
| Subject has been diagnosed with active psychosis, severe depression or suicidal ideation in the preceding year. | Yes | Yes |
| Subject has an implanted Vagus Nerve Stimulator (VNS). | Yes1 | Yes2 |

PMA P100026: FDA Summary of Safety and Effectiveness Data

{18}

| Exclusion Criteria | Feasibility Study | Pivotal Study |
| --- | --- | --- |
| Subject has had therapeutic surgery to treat epilepsy | in the preceding year | in the preceding 6 months |
| Subject is implanted with an electronic medical device that delivers electrical energy to the head. | Yes | Yes |
| Subject requires repeat MRIs | Yes | Yes^{3} |
| Subject’s epileptiogenic region(s) is / are located caudal to the level of the thalamus. | Yes | Yes |
| Subject is pregnant. | Yes | Yes |

1. A subject with an inactive VNS could be enrolled so long as the VNS was explanted prior to or at the same time as the RNS® System implant.
2. A subject could be enrolled if the subject is willing to have the VNS explanted (excluding leads) prior to or at the time of the RNS® System implant. (Subjects with VNS devices must have had VNS therapy discontinued for at least three months prior to enrollment.)
3. In which the head is exposed to the radio frequency field.

Subjects were eligible to enroll into the LTT study if they had completed either the Feasibility or Pivotal study, had the RNS® System implanted, had elected to continue to receive responsive stimulation, and were able to attend scheduled appointments for the study. They were not eligible if they had an active psychiatric or mental illness that made it inadvisable for the subject to continue to receive responsive stimulation or if the subject had been diagnosed with psychogenic or non-epileptic seizures, or primarily generalized seizures during the Feasibility or Pivotal studies.

2. Follow-up Schedule

A schematic of the study timeline is provided in Figure 2. The primary effectiveness analysis compares changes in seizure frequency in the Treatment group and in the Sham group during the 12-week BEP relative to the 12-week Pre-Implant Period. The Pre-Implant Period (not shown in Figure 2) is defined as the 12-weeks in the Baseline Period leading up to and including the date of qualification for implantation. Primary safety analyses include adverse event data over the first 12 weeks post-implantation. Secondary safety and effectiveness analyses include data from all periods of the study.

Information regarding daily seizure counts, adverse events and subject well-being was collected at all visits by a physician investigator who was blinded to the subject’s randomization status and a second non-blinded physician investigator was responsible for Neurostimulator programming.

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 19

{19}

Figure 2: RNS® System Pivotal Clinical Investigation – Trial Flow and Periods

## 3. Clinical Endpoints

### Safety:

The primary safety endpoint variables for the Feasibility and Pivotal studies were the serious adverse event (SAE) rates during the Acute Period (initial implant procedure and the following month) and the Short-Term Chronic Period (initial implant procedure and the following three months) compared to similar procedures reported in literature. Specifically, the study intended to demonstrate that the SAE rate is no worse than historical implantation of intracranial electrodes for localization procedures and epilepsy resective surgery rate and the historical Deep Brain Stimulator (DBS) rate for movement disorders from the published literature. To demonstrate this, the upper limit of the one-sided 95% confidence interval for the RNS® System SAE rate was not to exceed 42%. The SAE rate is defined as the proportion of subjects having a serious adverse event. The SAE rate includes all SAEs whether reported as device-related or not.

The secondary safety endpoints were as follows:

- The rate of occurrence of any adverse event (AE) observed during each of the post-implant time periods: The Post-operative Stabilization Period, the Stimulation Optimization Period, the BEP and the Open Label Evaluation Period. Data were compared for Treatment versus Sham stimulation.
- Affective status (by summary scores from the Beck Depression Inventory (McNair et al., 1971), the Profile of Mood State (Beck et al., 1961) and the Center for Epidemiologic Studies Depression Scale (CES-D) (Radloff, 1977) inventories) were described for each treatment group for the Baseline and for the 84-day BEP of the investigation, as well as at three time points during the subsequent Open Label Evaluation Period (56-week, 80-week, and 104-week visits).

PMA P100026: FDA Summary of Safety and Effectiveness Data

{20}

- Neuropsychological functioning as assessed by neuropsychological testing with validated, standardized inventories obtained pre-implant (within 28 days of the implant) and then at 20 weeks, 56 weeks and 104 weeks after implantation. The neuropsychological testing assessed visual and verbal memory, verbal fluency and naming, and cognitive flexibility.

- At the time of IDE approval, the sponsor prespecified in the clinical protocol that the SUDEP occurrence rate for the RNS® System would be no worse than 6.3/1000 patient-years. This rate was based on the reported incidence of SUDEP which ranges from 3.5 deaths per 1000 person years in a population-based cohort with epilepsy (Lhatoo et.al, 1991); 3.5/1000 patient-years in a well-defined cohort of 4,700 patients (5,747 patient-years of exposure) included in the worldwide clinical development database of the AED lamotrigine (Leetsma et.al., 1997); 4.5/1000 patient-years for the Cyberonics Vagus Nerve Stimulator; 6/1000 patient-years in patients with medically refractory epilepsy followed in an epilepsy clinic (Sperling et.al., 1999); to 6.3 deaths/1000 person-years in a population based Swedish cohort with refractory epilepsy who were candidates but did not choose to undergo epilepsy surgery (Nilsson et.al., 2003). The protocol stated that data from the RNS® System Feasibility Clinical Investigation, as well as the 5 year Long-term Treatment Investigation, would be used to collect approximately 1500 patient-years of data in order to confidently calculate the rate of SUDEP.

The number of patient-years necessary to determine that the 95% confidence interval for SUDEP does not exceed 6.3/1000 patient-years can be calculated based on the number of patient deaths attributed to SUDEP that occur during the RNS® System Clinical Investigations and the number of patient-years of data. As seen in Table 10 below, if there are 4 patient deaths identified as possibly or probably related to SUDEP, then 1446 years of patient follow-up will permit NeuroPace to be 95% confident that the true rate of SUDEP does not exceed 6.3/1000 patient-years. After three SUDEP events occurred the sponsor, with FDA concurrence, increased the acceptable SUDEP rate to 9.3/1000 patient-years.

Table 10: Follow-Up Required to Establish SUDEP Rate
| K^{1} | 1 | 2 | 3 | 4 | 5 | 6 |
| --- | --- | --- | --- | --- | --- | --- |
| M^{2} | 823 | 1016 | 1231 | 1446 | 1657 | 1864 |
1 K = number of deaths identified as possibly or probably related to SUDEP.
2 M = number of patient-years necessary to establish that the 95% confidence interval is &lt; 6.3/1000 patient-years.

## Effectiveness:

The primary effectiveness objective for the Pivotal study was to demonstrate a significantly greater reduction in the frequency of total disabling seizures in the Treatment group compared to the Sham group during the BEP relative to the Pre-Implant Period. Seizure frequency was modeled using the generalized estimating

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 21

{21}

equations (GEE) method, which accounts for within-subject correlations and variability across subject populations. The pre-specified primary efficacy endpoint variable was the group-by-time interaction term in the generalized estimating equation (GEE) model, where “group” refers to the therapy allocation (Treatment group or Sham group) and “time” refers to the trial period (Pre-Implant Period or BEP). The dependent variable was each subject’s daily seizure frequency during the Pre-Implant and BEPs.

The pre-specified GEE analysis assumed that daily seizure count data would follow a Poisson distribution. However, the observed distribution did not follow a Poisson distribution as a result of a large variability in day to day seizure counts in most subjects, as well as a large variability between subjects. The increased variability in the seizure frequency was not anticipated. In order to better fit the observed data, the following post hoc modifications were made to the pre-specified GEE analysis (referred to as the post hoc GEE model):

1) Using monthly rather than daily seizure count data
2) Modeling data with a negative binomial distribution rather than a Poisson distribution
3) Including the following clinical covariates that were used in the adaptive randomization:

a. Seizure onset zone location (subjects with seizure onsets exclusively in the mesial temporal lobe versus any other region(s) of the cortex)
b. Number of seizure foci (unifocal versus bifocal)
c. Prior therapeutic epilepsy surgery (resection, subpial transection and/or corpus callosotomy, versus no such surgery)

The GEE analysis, with these post hoc modifications, was used to demonstrate efficacy.

The secondary efficacy endpoints were as follows:

- Comparison of the Treatment group responder rate to the Sham group rate over the 84-day BEP of the investigation. (Responder rate is defined as the proportion of subjects who experience a 50% or greater reduction in mean disabling seizure frequency compared to the Pre-Implant Period.)
- Change in average frequency of disabling seizures during the BEP versus the Pre-Implant Period for the Treatment group compared to the Sham group.
- Proportion of seizure-free days during the BEP versus the Pre-Implant Period for the Treatment group compared to the Sham-stimulation group.
- Change in seizure severity, as determined by the Liverpool Seizure Severity Scale during the BEP versus the Pre-Implant Period for the Treatment group compared to the Sham group.

Additionally, subjects continued to be followed to gather long-term experience. The following endpoints were assessed:

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 22

{22}

- Change in average frequency of disabling seizures in the group originally randomized to the Sham-stimulation group (Therapy OFF) once therapy has been enabled in that group (Open Label Evaluation Period). Average seizure frequency during 84 days of the Open Label Evaluation Period was compared to the average seizure frequency during the 84-day BEP.
- Each subject’s responder status over the Open Label Evaluation Period.
- Daily seizure frequency counts compared to baseline during the Open Label Evaluation Period for both Sham and Treatment groups.
- Quality of life in individual subjects as measured with the QOLIE-89 assessment inventory to provide a descriptive analysis for each treatment group for the Baseline, Blinded Evaluation, and Open Label Evaluation Periods.

## B. Accountability of PMA Cohort

Subject participation in the Feasibility, Pivotal and LTT studies is presented in Figure 3 as of May 12, 2011. The safety and effectiveness analysis populations for the Pivotal study included all 191 subjects implanted and randomized; this is the intent-to-treat population. The combined safety analysis population includes the intent-to-treat safety population from the RNS® System Feasibility, Pivotal and Long-term Treatment (LTT) Clinical Investigations combined. This includes all 256 subjects implanted with the RNS® Neurostimulator and Leads.

## Withdrawals and Discontinuations

Forty-three (43) of the 256 enrolled subjects, discontinued treatment from all three studies. An additional 6 subjects did not transition. In the Feasibility study, 6 of 65 subjects discontinued. In the Pivotal study, 16 of 191 subjects discontinued. Two hundred thirty subjects transitioned to the LTT including 2 subjects who discontinued the Pivotal study early and later enrolled in the LTT study. As of May 12, 2011, 21 of 230 subjects had discontinued the LTT study. An additional 2 subjects have died since May 12, 2011, bringing the total number of death to 11 and the total number of subjects who had discontinued treatment to 45.

For the subjects who discontinued treatment, 7 subjects were explanted because of infection, 1 subject was explanted because of hemorrhage, 3 subjects were lost to follow-up, 9 subjects died and 23 subjects withdrew electively. The reasons given for elective withdrawal included: to pursue other treatments (13), because the reduction in seizures was not sufficient (4), and because the subject did not want to have the Neurostimulator replaced when the battery reached expected end of service (3). Another subject had a seizure-related fall that caused a scalp laceration that exposed the Neurostimulator: this subject chose not to have the laceration sutured and withdrew from the trial. Another subject was withdrawn because the physician felt that the subject was no longer a suitable candidate to participate because of psychiatric issues not related to treatment with the RNS® System. The reason for withdrawal for one subject was not specified.

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 23

{23}

Figure 3: Patient Flow Diagram
* Two subjects withdrew early (discontinued) from the Pivotal study to undergo resective epilepsy surgery. Waivers were granted to allow enrollment into the LTT study so that the subjects could continue to receive responsive stimulation to treat seizures arising from the non-resected seizure focus.

# C. Study Population Demographics and Baseline Parameters

Demographic information for subjects implanted in the Feasibility and Pivotal studies is presented in Table 11. All subjects participating in the LTT study originally enrolled in the Feasibility or Pivotal study. Table 12 provides information on epilepsy clinical characteristics and prior treatment for epilepsy by randomization group. Table 13 provides information on the number and types of leads implanted.

Table 11: Demographics

| Characteristic | All (N = 256) | By Study | |
| --- | --- | --- | --- |
| | | Feasibility (N = 65) | Pivotal (N = 191) |
| Gender (percent female) | 49% (125/256) | 52% (34/65) | 48% (91/191) |
| Age in years† (average, SD, range) | 34.0 ± 11.4 (18 - 66) | 30.9 ± 10.3 (18 - 56) | 34.9 ± 11.6 (18 - 66) |
| Years with epilepsy (average, SD, range) | 19.6 ± 11.4 (2 - 57) | 17.0 ± 10.1 (2 - 42) | 20.5 ± 11.6 (2 - 57) |
| Number of AEDs (average, SD, range) | 2.9 ± 1.1 (0 - 8) | 2.9 ± 1.0 (1 - 6) | 2.8 ± 1.2 (0 - 8) |
| Seizures per month (average, SD, range, median) | 50.7 ± 177.4 (0 - 2320) median = 10.2 | 99.2 ± 332.8 (0 - 2320) median = 11.3 | 34.2 ± 61.9 (3 - 338) median = 9.7 |

PMA P100026: FDA Summary of Safety and Effectiveness Data

{24}

Due to hospital confidentiality requirements some institutions did not provide date of birth for subjects

Table 12: Subset Populations of Interest (Implanted Subjects)

| Characteristic | All Implanted (N = 191) | By Randomization Group | | |
| --- | --- | --- | --- | --- |
| | | Treatment (N = 97) | Sham (N = 94) | p-value1 |
| Seizure onset location - Mesial Temporal Lobe Only (v. other)2 | 50% (95/191) | 49% (48/97) | 50% (47/94) | 0.943 |
| Number of seizure foci - Bifocal (v. unifocal)2 | 55% (106/191) | 49% (48/97) | 62% (58/94) | 0.089 |
| Prior therapeutic surgery for epilepsy2 | 32% (62/191) | 35% (34/97) | 30% (28/94) | 0.437 |
| Prior EEG monitoring with intracranial electrodes | 59% (113/191) | 65% (63/97) | 53% (50/94) | 0.098 |
| Prior VNS | 34% (64/191) | 31% (30/97) | 36% (34/94) | 0.443 |
| Anatomical brain abnormality (by neuroimaging) | 67% (128/191) | 68% (66/97) | 66% (62/94) | 0.759 |
| Benzodiazepine use (acute)3 | 36% (69/191) | 31% (30/97) | 41% (39/94) | 0.129 |

1 p-value per chi-square
2 Characteristics used as strata in adaptive randomization algorithm
3 Subjects who used acute benzodiazepines as rescue medications for seizures at any time during the Pre-Implant Period up until the implantation procedure. Does not include daily use of benzodiazepines

Table 13: Leads Implanted at Time of Initial Neurostimulator Implantation (Treatment and Sham)

| | Subject Population | | |
| --- | --- | --- | --- |
| | Implanted (N = 191) | Treatment (N = 97) | Sham (N = 94) |
| Number of Leads: | % (n/N) of subjects | | |
| 1 | 0% (0/191) | 0% (0/97) | 0% (0/94) |
| 2 | 58% (110/191) | 57% (55/97) | 59% (55/94) |
| 3 | 14% (26/191) | 14% (14/97) | 13% (12/94) |
| 4 | 29% (55/191) | 29% (28/97) | 29% (27/94) |
| Types of Leads: | % (n/N) of subjects | | |
| Cortical Strip Leads Only | 31% (59/191) | 31% (30/97) | 31% (29/94) |
| Depth Leads Only | 39% (74/191) | 37% (36/97) | 40% (38/94) |
| Cortical Strip and Depth Leads | 30% (58/191) | 32% (31/97) | 29% (27/94) |

PMA P100026: FDA Summary of Safety and Effectiveness Data

{25}

D. Safety and Effectiveness Results

1. Safety Results

The analysis of the primary safety endpoint was based on the cohort of 191 subjects available for the 12-week evaluation. Additional safety analyses were provided for the 256 implanted subjects in the Feasibility, Pivotal, and LTT as of May 12, 2011. The key safety outcomes for this study are presented below in Table 14 through Table 27. All adverse effects through 2 years are reported in Table 28.

Adverse effects that occurred in the PMA clinical study:

The RNS® System Feasibility, Pivotal and LTT studies evaluated the safety of the RNS® System for epilepsy in 256 implanted subjects over 903 patient-years of implant experience and 819 patient-years of responsive stimulation. All adverse event data are current as of May 12, 2011 with the exception of deaths and SUDEP analysis which are current through October 24, 2012. The investigator classified each adverse event as serious or non-serious and as device-related (which includes device-related and device-relation uncertain) or not device-related. Adverse events were considered serious if the event resulted in significant risks or consequences to the subject's acute or long-term health, serious injury or death, hospital admission, or if invasive medical intervention was required to alleviate the adverse event. Adverse events are presented using MedDRA Coding according to the Preferred Term (PT).

During all study periods, 165/256 (64.5%) subjects experienced a serious adverse event and 254/256 (99.2%) subjects experienced a non-serious adverse event, including common and expected illnesses. The RNS® System Feasibility and Pivotal studies met the safety endpoints pre-specified in the investigational plans (see Table 14 below). Adverse events associated with neuropsychological function are shown in Table 22. Serious adverse events (SAEs) and Adverse Events occurring in ≥ 2.5% of subjects are reported by study period in Table 15 through Table 21 below. Adverse events of special interest are also discussed. A full listing of adverse events (AEs) in subjects by study period is presented in Table 28.

There were no unanticipated device-related serious adverse events during the RNS® System studies. The primary safety endpoint was to compare similar procedures to the significant adverse events for the surgical procedure and following 28 days (acute) and to compare similar procedures to the surgical procedure and the following 84 days (short-term chronic). The primary safety endpoint was met. There was no difference between the Treatment and Sham groups in the overall percentage of subjects experiencing an adverse event, or any specific type of adverse event during the evaluation periods of the studies. The overall frequency of adverse events or of specific adverse events did not increase over time, whether or not the investigator considered the event as device-related or not device-related.

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 26

{26}

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 27

# Pivotal Study: Primary Safety Endpoint

The primary safety endpoint was met. The rate of serious adverse events after implantation of the Neurostimulator and Leads was similar over the first 4 weeks (Acute Period) and in the first 12 weeks (Short-Term Chronic Period) compared to similar procedures, i.e., the combined risks of implantation of intracranial electrodes for purposes of an epilepsy surgery evaluation and epilepsy surgery, and the risks of deep brain stimulation for treatment of movement disorders. The results, presented in Table 14, demonstrate that the SAE rate over the first month and the first 3 months after implantation is comparable to the literature based historical controls.

Table 14: Pivotal Study – Primary Safety Endpoint

| Period | SAE Rate¹ | | Met primary safety endpoint?³ |
| --- | --- | --- | --- |
| | RNS® System | Comparator² | |
| | % (n/N) subjects [upper 95% CI] | | |
| Acute (Surgery – Week 4) | 12.0% (23 /191) [16.5%] | 15% [20%] | Yes (16.5% < 20%) |
| Short-Term Chronic (Surgery – Week 12) | 18.3% (35 /191) [23.4%] | 36% [42%] | Yes (23.4% < 42%) |

¹ Upper limit of the one-sided 95% confidence interval, estimated using the Score Interval (also known as the Wilson Interval). Upper limits for literature comparators were pre-specified in the protocol, estimated using the Score Interval based on a sample size of 180.
² Protocol-specified Acute endpoint, based on literature: SAE rate associated with implantation of intracranial electrodes and epilepsy surgery (Tanriverdi et al., 2009; Wong et al., 2009; Fountas and Smith, 2007; Hamer et al., 2002; Behrens et al., 1997). Protocol-specified Short-Term Chronic endpoint, based on literature: SAE rate associated with deep brain stimulation for movement disorders (Oh et al., 2002; Summary of Safety and Effectiveness, Activa Tremor Control System P960009; Beric et al., 2001; Behrens et al., 1997; Hariz, 2002; Joint et al., 2002; Koller et al., 2001).
³ Upper limit for the RNS® System is less than that of the comparator.

# Pivotal Study: Post-operative Stabilization Period AEs

Table 15 presents the SAEs that occurred in 23 subjects during the Post-operative Stabilization Period. Five (5) subjects had implant site infections, one of which required explant of the leads and stimulator. Two additional subjects had serious adverse events reported as effusion or discharge at the implant site. One subject was reported to have bacterial meningitis, which was diagnosed before the RNS® Neurostimulator and Leads were implanted. The infection was a chronic infection from a prior evaluation with intracranial electrodes. The investigator elected to implant the RNS® Neurostimulator and NeuroPace Leads despite the observed infection and treated the patient with antibiotics. Three subjects had intracranial hemorrhages. Additional information regarding all hemorrhages is provided in the section below entitled, “Combined Studies: Adverse Events of Particular Relevance”. Adverse events that occurred in ≥ 2.5% of subjects are presented in Table 16. A full listing of adverse events during the Post-operative Stabilization Period is presented in Table 28.

{27}

Table 15: Pivotal Study – Serious Adverse Events during the Post-operative Stabilization Period¹

| MedDRA Preferred Term | % (#) Subjects with events
N=191 |
| --- | --- |
| Implant site infection | 2.6% (5) |
| Extradural hematoma | 1.0% (2) |
| Hydrocephalus | 1.0% (2) |
| Procedural headache | 1.0% (2) |
| Apraxia | 0.5% (1) |
| Biopsy brain | 0.5% (1) |
| Cerebral hemorrhage | 0.5% (1) |
| Complex partial seizures exacerbated | 0.5% (1) |
| Depression suicidal | 0.5% (1) |
| Device lead revision | 0.5% (1) |
| Drug hypersensitivity | 0.5% (1) |
| Dysphemia | 0.5% (1) |
| Implant site discharge | 0.5% (1) |
| Implant site effusion | 0.5% (1) |
| Meningitis bacterial² | 0.5% (1) |
| Pneumothorax | 0.5% (1) |
| Postictal state | 0.5% (1) |
| Procedural vomiting | 0.5% (1) |
| Subdural hematoma | 0.5% (1) |
| Therapeutic agent toxicity | 0.5% (1) |
| Summary of All SAEs in this Period | 12.0% (23) |

¹ All SAEs resolved.
² Subject was diagnosed with bacterial meningitis before implant.

Table 16: Pivotal Study – Adverse Events occurring in ≥ 2.5% of subjects during the Post-operative Stabilization Period

| MedDRA Preferred Term | % (#) Subjects with events
N=191 |
| --- | --- |
| Implant site pain | 28.3% (54) |
| Procedural headache | 27.2% (52) |
| Procedural nausea | 4.7% (9) |
| Implant site swelling | 4.2% (8) |
| Dizziness | 3.7% (7) |
| Postoperative constipation | 3.7% (7) |
| Swelling face | 3.7% (7) |
| Postoperative fever | 3.1% (6) |
| Therapeutic agent toxicity | 3.1% (6) |
| Adverse drug reaction | 2.6% (5) |
| Implant site infection | 2.6% (5) |

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 28

{28}

Pivotal Study: Stimulation Optimization Period AEs

Table 17 presents the SAEs that occurred in 12 subjects (6 in the Treatment group and 6 in the Sham group) during the Stimulation Optimization Period. Serious adverse events of particular interest include implant site infection and subdural hematoma (due to a seizure). Adverse events that occurred in $\geq 2.5\%$ of subjects are presented in Table 18.

Table 17: Pivotal Study – Serious Adverse Events during the Stimulation Optimization Period (Treatment and Sham)

| MedDRA Preferred Term | Treatment
N=97 | Sham
N=94 |
| --- | --- | --- |
| | % (#) Subjects with events | |
| Summary of All SAEs in this Period | 6.2% (6) | 6.4% (6) |
| Device lead revision | 1.0% (1) | 1.1% (1) |
| Medical device removal (VNS) | 1.0% (1) | 1.1% (1) |
| Adverse drug reaction | 1.0% (1) | -- |
| Arthritis | -- | 1.1% (1) |
| Central venous catheterisation | 1.0% (1) | -- |
| Death | -- | 1.1% (1) |
| EEG monitoring | -- | 1.1% (1) |
| Implant site infection (dts^{1}) | 1.0% (1) | -- |
| Meningioma benign | 1.0% (1) | -- |
| Non-cardiac chest pain | -- | 1.1% (1) |
| Psychotic disorder | -- | 1.1% (1) |
| Skin laceration (dts) | 1.0% (1) | -- |
| Subdural hematoma (dts) | 1.0% (1) | -- |
| Syncope | 1.0% (1) | -- |

1 dts = due to seizure

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 29

{29}

Table 18: Pivotal Safety – Adverse Events in ≥ 2.5% of Subjects during the Stimulation Optimization Period (Treatment and Sham)

| MedDRA Preferred Term | Treatment (N=97) | Sham (N=94) | p-value^{1} |
| --- | --- | --- | --- |
| | % (#) Subjects with events | | |
| Headache | 4.1% (4) | 8.5% (8) | 0.245 |
| Nasopharyngitis | 5.2% (5) | 3.2% (3) | 0.721 |
| Depression | 3.1% (3) | 3.2% (3) | 1.000 |
| Implant site pain | 2.1% (2) | 3.2% (3) | 0.679 |

¹ Comparison of percentage of subjects with events in Treatment vs. Sham groups per Fisher’s exact test.

## Pivotal Study: Blinded Evaluation Period (BEP) AEs

The total number of subjects that experienced any serious or non-serious adverse event during the BEP was 90/97 (92.8%) for the Treatment group and 88/94 (93.6%) for the Sham group. There was no statistical difference in the frequency of serious adverse events between the Treatment and Sham stimulation groups. Only one type of adverse event was statistically significantly different between the Treatment and Sham stimulation groups. Therapeutic agent toxicity, which refers to side effects of AEDs, was more common in the Sham group (5 subjects, all non-serious events) than the Treatment group (none).

Table 19 presents the SAEs that occurred in 9 subjects (4 in the Treatment group and 5 in the Sham group) during the BEP. Table 20 presents AEs reported in 2.5% or more of the subjects in either the Treatment or the Sham groups who entered the 12-week BEP of the Pivotal Study. This includes all adverse events whether device-related or not.

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 30

{30}

Table 19: Pivotal Safety – Serious Adverse Events during the Blinded Evaluation Period (Treatment and Sham)

| MedDRA Preferred Term | Treatment N=96 | Sham N=93 |
| --- | --- | --- |
| | % (#) Subjects with events | |
| Complex partial seizures increased | 1.0% (1) | 1.1% (1) |
| Alcohol poisoning | 1.0% (1) | -- |
| Hernia | -- | 1.1% (1) |
| Implant site infection (dts¹) | -- | 1.1% (1) |
| Jaw fracture (dts) | -- | 1.1% (1) |
| Myocardial infarction | 1.0% (1) | -- |
| Nephrolithiasis | -- | 1.1% (1) |
| Pneumonia | 1.0% (1) | -- |
| Simple partial seizures (sensory) | -- | 1.1% (1) |
| Simple partial seizures increased (sensory) | -- | 1.1% (1) |
| Summary of All SAEs in Period² | 4.2% (4) | 5.4% (5) |

¹ dts = due to seizure
² Differences between the percentage of subjects in the Treatment group reporting the events and that in the Sham group were not significant (all p-values &gt; 0.05 by Fisher's exact test).

Table 20: Pivotal Study – Adverse Events in ≥ 2.5% of Subjects in Either Group During the Blinded Evaluation Period (Treatment vs. Sham)

| MedDRA Preferred Term | Treatment N=96 | Sham N=93 | p-value¹ |
| --- | --- | --- | --- |
| | % (#) subjects with events | | |
| Nasopharyngitis | 6.3% (6) | 8.6% (8) | 0.588 |
| Headache | 5.2% (5) | 7.5% (7) | 0.563 |
| Contusion (dts²) | 7.3% (7) | 2.2% (2) | 0.170 |
| Skin laceration (dts) | 6.3% (6) | 3.2% (3) | 0.498 |
| Complex partial seizures increased | 4.2% (4) | 3.2% (3) | 1.000 |
| Depression | 5.2% (5) | 2.2% (2) | 0.445 |
| Dysesthesia | 2.1% (2) | 5.4% (5) | 0.273 |
| Influenza | 4.2% (4) | 3.2% (3) | 1.000 |
| Vomiting | 3.1% (3) | 3.2% (3) | 1.000 |
| Adverse drug reaction | 3.1% (3) | 2.2% (2) | 0.445 |
| Therapeutic agent toxicity | -- | 5.4% (5) | 0.027 |
| Upper respiratory tract infection | 1.0% (1) | 4.3% (4) | 0.206 |
| Pain of skin | 4.2% (4) | -- | 0.121 |
| Pharyngitis | 1.0% (1) | 3.2% (3) | 0.363 |
| Abdominal pain | 3.1% (3) | -- | 0.246 |
| Balance disorder | -- | 3.2% (3) | 0.117 |

PMA P100026: FDA Summary of Safety and Effectiveness Data
Page 31

{31}

| MedDRA Preferred Term | Treatment N=96 | Sham N=93 | p-value1 |
| --- | --- | --- | --- |
| | % (#) subjects with events | | |
| Head injury | -- | 3.2% (3) | 0.117 |

Fisher's exact test
2 dts = due to seizure

Pivotal Study: AEs during All Study Periods through Two Years Post-Implant All device-related adverse events (serious and non-serious) occurring during the Pivotal study through 2 years post-implant in 2.5% or more of the subjects are presented by study period in Table 21. A full listing of adverse events during the Pivotal Study is presented in Table 28.

The most frequent serious adverse event during the 28 days after implant was implant site infection, occurring in 2.6% of subjects. There were 5 implant site infections; one of these subjects had the Neurostimulator and Leads explanted. The most common non-serious adverse events were implant site pain, procedural headache and implant site swelling and infection. The most common serious device-related adverse events through two years post-implant were implant site infection (3.7%), device Lead damage (2.6%), and device Lead revision (2.1%). Device-related serious adverse events affecting 1% (2 subjects) were extradural hematoma, hydrocephalus, and premature battery depletion. Device-related serious adverse events affecting 0.5% (1 subject) at any time over the entire Pivotal Study were cerebral hemorrhage, implant site discharge, implant site erosion, implant site pain, intracranial hypotension, medical device removal, procedural headache, subdural hematoma, and suture related complication.

Table 21: Pivotal Study – Device-Related Adverse Events in ≥ 2.5% of Subjects by Study Period through 2 Years

| | Post Op (Implant - Week 4) | Stim Opt (Weeks 4 - 8) | BEP (Weeks 8 - 12) | Open Label Period | | All Study Periods1 |
| --- | --- | --- | --- | --- | --- | --- |
| | | | | (Weeks 20 - 52) | (Weeks 52 - Completion) | |
| Preferred Term | % subjects (# subjects)2 | | | | | |
| Number of Subjects Entering | 191 | 191 | 189 | 187 | 182 | 191 |
| Total Implant-years within Interval | 14.7 | 14.6 | 43.2 | 113.4 | 193.4 | 379.2 |
| Implant site pain | 9.9% (19) | 2.1% (4) | 0.5% (1) | 3.7% (7) | 4.4% (8) | 18.3% (35) |
| Procedural headache | 11.5% (22) | -- | -- | 0.5% (1) | 0.5% (1) | 12.6% (24) |
| Device interaction | 1.6% (3) | 0.5% (1) | 1.1% (2) | 1.6% (3) | 0.5% (1) | 5.2% (10) |
| Implant site infection | 2.6% (5) | -…