VNS THERAPY SYSTEM
Device Facts
| Record ID | P970003 |
|---|---|
| Device Name | VNS THERAPY SYSTEM |
| Applicant | LivaNova USA, Inc. |
| Product Code | LYJ · Neurology |
| Decision Date | Jul 16, 1997 |
| Decision | APPR |
| Device Class | Class 3 |
| Attributes | Therapeutic, Expedited Review |
Indications for Use
The NCP® system is indicated for use as an adjunctive therapy in reducing the frequency of seizures in adults and adolescents over 12 years of age with partial onset seizures, which are refractory to antiepileptic medications.
Device Story
NCP® System is an implantable vagus nerve stimulation (VNS) device; consists of an implantable pulse generator (IPG), lead, and external programming components. IPG is surgically placed in the left chest; lead is tunneled subcutaneously and attached to the left vagus nerve. IPG delivers constant current, capacitively coupled, charge-balanced electrical signals to the nerve. Physician uses an IBM-compatible PC with programming software and a wand to non-invasively activate, program, and interrogate the generator. Patients are provided with a handheld magnet to manually activate stimulation to abort seizures or deactivate the device. Stimulation parameters (current, frequency, pulse width, ON/OFF times) are programmable to individualize therapy. Device provides adjunctive therapy for refractory epilepsy; intended to reduce seizure frequency. Benefits include potential reduction in seizure frequency and improved quality of life. Risks include surgical complications, stimulation-related side effects (e.g., hoarseness, cough, paresthesia), and potential nerve injury.
Clinical Evidence
Five clinical studies (E01-E05) involving 454 patients and >900 device-years. E03 and E05 were randomized, double-blind, parallel, active-controlled trials comparing 'HIGH' (therapeutic) vs. 'LOW' (sub-therapeutic) stimulation. Primary endpoint: mean/median percent change in seizure frequency. E03 HIGH group showed statistically significant mean seizure reduction (23.6%) vs. LOW group (6.4%, p=0.0175). E05 HIGH group showed 27.9% mean reduction vs. 15.2% in LOW group (p=0.0396). Common adverse events included voice alteration (50%), cough (41%), and pharyngitis (27%), typically related to stimulation intensity.
Technological Characteristics
Implantable pulse generator (Model 100) in hermetically sealed titanium case; lithium thionyl chloride battery. Lead (Model 300) features helical platinum ribbon electrodes insulated with silicone rubber. Sensing/actuation: constant current, charge-balanced electrical stimulation. Connectivity: inductive telemetry via external wand and PC. Sterilization: ethylene oxide for sterile components; autoclave for tunneling tool. Software: menu-driven, IBM-compatible PC-based programming system.
Indications for Use
Indicated for adults and adolescents >12 years with partial onset seizures refractory to antiepileptic medications. Contraindicated in patients after bilateral or left cervical vagotomy.
Submission Summary (Full Text)
{0}
Page i
# NeuroCybernetic Prosthesis (NCP®) Vagus Nerve Stimulation System
## SUMMARY of SAFETY and EFFECTIVENESS DATA
### Table of Contents
1. GENERAL INFORMATION ... 1
2. INTENDED USE/INDICATIONS ... 1
3. CONTRAINDICATIONS ... 1
4. WARNINGS and PRECAUTIONS ... 1
5. DEVICE DESCRIPTION ... 2
Figure 1. Placement of the Implanted Components of the NCP® System ... 4
5.1 NCP® Generator (Model 100) ... 4
5.2 NCP® Programming Wand (Model 200) ... 5
5.3 NCP® Programming Software (Model 250) ... 5
5.4 NCP® Vagus Nerve Stimulation Lead (Model 300 Series) ... 5
5.5 NCP® Tunneling Tool (Model 400) ... 6
6. ALTERNATIVE PRACTICES AND PROCEDURES ... 6
7. MARKETING HISTORY ... 6
8. ADVERSE EVENTS ... 6
8.1 Observed Adverse Events ... 6
Table 1. Observed Adverse Events ... 7
8.2 Potential Adverse Events ... 8
9. SUMMARY OF PRE-CLINICAL STUDIES ... 9
9.1 Risk Analysis ... 9
9.2 Components Testing ... 9
9.3 Device Testing ... 10
9.4 Biocompatibility ... 13
9.5 Sterilization and Shelf Life Testing: ... 13
9.6 Animal Testing ... 13
10. SUMMARY OF CLINICAL STUDIES ... 15
10.1 Design and Methods ... 15
Figure 2. E01, E02, E04 Study Designs and Timelines ... 16
Figure 3. E03, E05 Study Designs and Timelines ... 17
Table 2. Protocol Stimulation Parameters ... 17
Table 3. Inclusion Criteria ... 19
Table 4. Exclusion Criteria ... 19
10.2 Description of Patients Studied and analyses for Gender Bias ... 20
Table 5. Description of Study Patients ... 20
Table 6. Description of Clinical Studies ... 21
10.3 Results – Effectiveness and Safety ... 21
Table 7. Seizure Rate Changes ... 22
Table 8. Responder Rates ... 23
Figure 4. Change in Seizure Frequency, Patient Distribution ... 24
Table 9. Observed Adverse Events ... 27
Table 10. Adverse Events Reported During Baseline or Stimulation ... 28
Table 11. Treatment Emergent Adverse Events By Severity ... 29
Table 12. Treatment Emergent Signs and Symptoms (>2%) ... 30
Table 13. Deaths and Other Serious Adverse Events ... 33
Table 14. Device Complications and Observations ... 34
11. CONCLUSIONS DRAWN FROM STUDIES ... 35
12. PANEL RECOMMENDATIONS ... 36
13. FDA DECISION ... 36
14. APPROVAL SPECIFICATIONS ... 36
NCP® SSED, P970003
{1}
Page 1
# SUMMARY of SAFETY and EFFECTIVENESS DATA
## 1. GENERAL INFORMATION
**Device Generic Name:**
Stimulator, Vagus Nerve
**Device Trade Names:**
NeuroCybernetic Prosthesis (NCP®) System
Model 100 NCP® Generator
Model 200 NCP® Programming Wand
Model 250 NCP® Programming Software
Model 300 Series NCP® Vagus Nerve Stimulation Lead
Model 400 Tunneling Tool
NCP® System Accessories
**Applicant's Name and Address:**
Cyberonics, Inc.
17448 Highway 3, Suite 100
Webster, TX 77598-4135 USA
**PMA Number:** P970003
**Date of Panel Recommendation:** June 27, 1997
**Date of Notice of Approval to the Applicant:** July 16, 1997
## 2. INTENDED USE/INDICATIONS
The NCP® system is indicated for use as an adjunctive therapy in reducing the frequency of seizures in adults and adolescents over 12 years of age with partial onset seizures, which are refractory to antiepileptic medications.
## 3. CONTRAINDICATIONS
The NCP® system cannot be used in patients after a bilateral or left cervical vagotomy.
## 4. WARNINGS and PRECAUTIONS
Same as final product labeling.
NCP® SSED, P970003
{2}
Page 2
# 5. DEVICE DESCRIPTION
The NCP® System consists of the following components categorized into those implanted and peripheral components.
## Implanted Components:
NCP® Generator (Model 100)
NCP® Lead (Model 300)
NCP® SSED, P970003
{3}
Page 3
# External Components:
NCP Programming Wand
(Model 200)
IBM Compatible Computer
NCP®
NCP Programming Software
(Model 250)
NCP Tuneling Tool
(Model 400)
Electrical signals are transmitted from the NCP® Generator to the vagus nerve via the Model 300 Series NCP® Lead. Peripheral components are used with an IBM-compatible personal computer to non-invasively activate, program, and retrieve information from the Generator. An NCP® Tunneling Tool (Model 400) is used during implantation to create a subcutaneous path for the Lead and is used in its placement as well.
NCP® SSED, P970003
{4}
Page 4
Figure 1. Placement of the Implanted Components of the NCP® System
The NCP® Generator, a device similar to a cardiac pacemaker, is surgically placed in the left chest. The NCP® Lead is connected to the Generator and attached to the left vagus nerve. Patients are provided with a hand held magnet which, by passing the magnet over the implanted NCP® Generator, can be used to manually activate the Generator to provide additional non-programmed stimulation in an effort to prevent or abort seizures and to test for battery status. Use of the magnet-activated stimulation is part of the NCP® System’s prescribed treatment regimen. The magnet can also be used to deactivate (turn OFF) programmed stimulation by continuous placement of the magnet over the device. Programmed stimulation resumes when the magnet is removed.
## 5.1 NCP® Generator (Model 100)
The Model 100 NCP® Generator is an implantable, multiprogrammable pulse Generator that delivers electrical signals to the vagus nerve. Constant current, capacitively coupled, charge-balanced signals are transmitted from the Generator to the vagus nerve by the NCP® Lead (Model 300 Series).
The NCP® Generator is housed in a hermetically sealed titanium case. Feedthrough capacitors are used to filter electromagnetic interference from the Pulse Generator circuitry. The major components and functions of the Generator are as follows: a microprocessor, a voltage regulator, a 76.8 kHz crystal oscillator, one antenna to transmit information and another antenna to receive information, communication circuitry, DC-DC voltage generation and control circuitry, constant current control circuitry, a dual pole magnetic reed switch for manual activation of the Generator and for inhibition of the output pulses, and a lithium thionyl chloride cell to provide power for stimulation and circuit operation. The lithium thionyl chloride battery chemistry has the low impedance and high energy density characteristics required for the rapid pulsing needed in peripheral nerve stimulation, and similar batteries have been previously used in cardiac pacemakers, implantable spinal cord stimulators, and implantable drug pumps.
### Therapy:
The NCP® Generator has a number of programmable settings which allow a physician to individualize the treatment for a patient. Those settings include pulse width, magnet-activated output current, normal output current, magnet-activated ON time, signal frequency, magnet-activated pulse width, signal ON time, and signal OFF time. The Applicant provides a magnet
NCP® SSED, P970003
{5}
Page 5
that may be used to either manually initiate stimulation or to turn OFF the device. Device labeling instructs physicians to begin stimulation at the lowest output current settings possible (0.25 mA) and to ramp up stimulation to the desired level in 0.25 amp increments, never exceeding a level that is not tolerable to the patient.
**Diagnostic and safety characteristics:**
The NCP® Generator has telemetry capability which supplies information about its operating characteristics, such as parameter settings, lead impedance and history of magnet use. The Generator has a number of characteristics intended to strengthen operational reliability and safety, such as electromagnetic interference (EMI) filter capacitors, a series battery resistor to limit temperature rise in the event of short circuit, defibrillation protection diodes, direct current-blocking capacitors on both Leads that prevent direct current (DC) from being applied to the patient, a software watchdog timer to prevent continuous stimulation, and protection against voltage dips on the battery that could disrupt microprocessor memory.
## 5.2 NCP® Programming Wand (Model 200)
The NCP® Programming Wand (Model 200) is used with the NCP® Programming Software (Model 250) installed on a dedicated IBM-compatible personal computer to activate, program, reprogram and interrogate the NCP® Generator. The programming software is provided to the user on a 3.5-inch floppy disk. Capabilities include revision of the programmable parameters of the Generator, retrieval of telemetry data, and resetting of the Generator's microprocessor.
## 5.3 NCP® Programming Software (Model 250)
The NCP® Programming Software (Model 250) is a computer program which permits communication with an implanted NCP® Generator. The Programming Software is menu-driven and uses on-screen messages and prompts to assist the operator in using the system. Whenever the Programming Software is initialized, a self test is automatically run on the Software to verify checksum, file lengths, and file names. The programmed parameters and operational status can be interrogated. One or more parameters can be programmed at one time, and the programmed values are verified and displayed. The NCP® Programming System employs a strict communications protocol designed to minimize the possibility of "phantom" programming (i.e., inadvertent programming via environmental sources of electromagnetic interference or partial programming of a parameter). The Programming Software should be used on a computer dedicated only for programming the NCP® system. The Programming Software has been validated on a Compaq Contura Aero 4/25 running MS DOS 6.2 with the disk caching software disabled.
## 5.4 NCP® Vagus Nerve Stimulation Lead (Model 300 Series)
The Lead delivers electrical signals from the Generator to the vagus nerve. The Lead has two helical electrodes with a helical anchor tether on one end and two 5-millimeter (mm) connectors on the other end. The helix of the Lead is available in two sizes of inner diameter (2.0-mm and 3.0-mm) to allow for appropriate fit on different sized nerves. The helical design is soft, pliable, and expands or contracts with changes in nerve diameter, which may occur immediately post implant. These design features allow the 2-mm inside diameter helical electrode to fit most vagus
NCP® SSED, P970003
{6}
Page 6
nerves. The Model 300 Series NCP® Lead is insulated with silicone rubber and is bifurcated at each end. The Lead wire is quadrifilar MP-35N, and the electrode is a platinum ribbon.
## 5.5 NCP® Tunneling Tool (Model 400)
An NCP® Tunneling Tool (Model 400) is used during implantation to create a subcutaneous path for the NCP® Lead and is used in its placement as well. The Tunneling Tool is supplied non-sterile with instruction to autoclave prior to use.
## 6. ALTERNATIVE PRACTICES AND PROCEDURES
### A. Anti-Epileptic Drugs (AEDs)
The NCP® System clinical experience at the time of approval has been as an adjunctive treatment to anti-epileptic drug (AED) therapy. Additional anticonvulsant drugs are an alternative to treatment with vagus nerve stimulation using the NCP® System.
### B. Therapeutic Resective Surgery
Epilepsy surgery may be an option for some patients. Generally less than two percent of the medically refractory patients undergo surgery each year in the U.S. The goal of resective seizure surgery is to excise a small area of brain tissue which contains a discrete seizure focus. Epilepsy surgery does have significant morbidity and mortality, including memory loss and stroke.
## 7. MARKETING HISTORY
On June 1, 1994, the applicant began commercial distribution of the NCP® System in Europe. The device has also been marketed in Australia, Switzerland, South Africa, Israel, China, Myanmar (Burma), Vietnam, Hungary, Hong Kong, and Canada.
The NCP® System has not been withdrawn from any country, for any reasons related to the safety or effectiveness of the device.
## 8. ADVERSE EVENTS
The NCP® System was implanted in 454 patients in five clinical studies involving 611 devices (some patients had Generator replacements). As of August, 1996, total NCP® exposure in these 454 patients was 901 device years. Individual patient exposure averaged 24 months with a range of 8 days to 89 months.
A total of 9 patients died during these five studies. One patient died from each: thrombotic thrombocytopenic purpura, drownings, aspiration pneumonia, pneumonia, and renal failure associated with drug and alcohol ingestion. No cause of death was apparent for the other four and they may be classified as Sudden Unexpected Death in Epilepsy (SUDEP). None of the deaths were attributed by the Investigators to the NCP® System.
## 8.1 Observed Adverse Events
Included among the five clinical trials were two randomized trials (Study E03 & E05) which involved 314 patients and implantation of 413 devices yielding a total NCP® System exposure
18
NCP® SSED, P970003
{7}
Page 7
(inclusive of long term follow-up) of 591 years. These trials form the basis for the rates of observed adverse events.
Table 1 reports the adverse events from these studies during the Randomized Phase (14 week observation period) on a per patient basis. For the Extension Phase, events are reported on a per patient and per patient-year basis. The most common side effect associated with stimulation is hoarseness (voice alteration) during stimulation (hoarseness should only occur during the ON time). Most people tolerate the hoarseness well.
Table 1. Observed Adverse Events
(N=413 devices in 314 patients, 152 patients in HIGH treatment group, 591 device years)
| | Randomized + Extension Phase, N= 314 patients, 591 device years | | | | Randomized Phase, HIGH Only, N = 152 pts | |
| --- | --- | --- | --- | --- | --- | --- |
| Adverse Event | # of patients | % of patients | # of Events | Events per device-year | # of patients | % of patients |
| Serious AEs^{1} | | | | | | |
| Surgically Related | 13 | 4.1 % | 13 | 0.022 | | |
| Stimulation Related | 4 | 1.2 % | 4 | 0.007 | 1 | 0.7 % |
| Non-serious AEs | | | | | | |
| Voice Alteration | 156 | 50% | 720 | 1.228 | 91 | 60% |
| Cough Increased | 129 | 41% | 456 | 0.772 | 57 | 38% |
| Pharyngitis | 84 | 27% | 182 | 0.308 | 36 | 24% |
| Paresthesia | 87 | 28% | 377 | 0.638 | 32 | 21% |
| Dyspnea | 55 | 18% | 55 | 0.093 | 32 | 21% |
| Dyspepsia | 36 | 12% | 98 | 0.166 | 22 | 15% |
| Nausea | 59 | 19% | 154 | 0.261 | 21 | 14% |
| Laryngismus | 10 | 3.2% | 30 | 0.051 | 9 | 5.9% |
1 - Serious AEs reported included infection, nerve paralysis, hypersthesia, facial paresis, left vocal chord paralysis, left facial paralysis, left hemidiaphragm paralysis, left recurrent laryngeal nerve injury, urinary retention, and low grade fever.
Status Epilepticus: Valid estimates of the incidence of treatment emergent status epilepticus among VNS treated patients are difficult to obtain because reporters participating in clinical trials did not all employ identical rules for identifying cases. At a minimum, 2 of 441 adult patients had episodes that could unequivocally be described as status. In addition, a number of reports of variably defined episodes of seizure exacerbation (e.g., seizure clusters, seizure flurries, etc.) were made.
Rebound after stimulation is stopped: Seizure frequency was monitored for one to four weeks after stimulation was stopped in 72 instances (68 patients) in Study E03. Seizure rate increased by greater than 1.5 standard deviations above baseline in 10% of instances (compared to 7% expected). Of these instances, 11 of 72 (15%) had a greater than 25% increase above baseline and 42 of 72 (58%) had a greater than 25% decrease in seizure rate.
NCP® SSED, P970003
{8}
Page 8
## 8.2 Potential Adverse Events
Adverse events which may occur (including those reported in Table 1), reported in alphabetical order:
- Cough increased
- Dyspepsia, duodenal or gastric ulcer
- Dyspnea
- Facial paresis or paralysis
- Hemidiaphragm paralysis
- Hypesthesia
- Infection
- Laryngismus
- Muscle twitching during stimulation
- Nausea and vomiting
- Nerve injury
- Paresthesia
- Pharyngitis
- Voice alteration
Patients who manipulate the NCP® Pulse Generator and the Lead through the skin may damage or disconnect the Lead from the Pulse Generator and/or possibly cause damage to the vagus nerve.
Possible complications of NCP® System treatment include those related to implantation, those related to performance of the implanted Generator, and those related to long-term patient tolerance of the implant.
Except for Lead positioning, implantation of the NCP® Generator is similar to implantation of a cardiac pacemaker. In addition to the normal risks associated with a surgical procedure, complications associated with NCP® System implantation include, but may not be limited to, skin irritation; pain at the incision site; infection; extrusion or migration of the NCP® Generator and/or Lead; dislodgment, disconnection (Lead from Generator), breakage (Lead), or corrosion of the stimulating Lead; hematoma; fluid accumulation; cyst formation; inflammation; and histotoxic reactions. These phenomena may occur either acutely or chronically and may require device replacement to correct the complication.
Complications can include damage to the vagus nerve, either due to surgical trauma, compression by the electrode, or excessive stimulation. Hoarseness not associated with the stimulation suggests possible nerve irritation. Persistent hoarseness may be caused by nerve constriction, nerve fatigue, or Generator malfunction.
Normal stimulation of the vagus nerve and its branches may be associated with some side effects. In two randomized parallel controlled studies, a comparison between baseline rates of occurrence and treatment rates, for either HIGH or LOW, revealed the following adverse events (AEs) which had a statistically significant difference between periods in any study group: cough, dyspepsia, dyspnea, hoarseness/voice alteration, hypesthesia, infection, insomnia, laryngismus, non-specific pain, abdominal pain, paresthesia, pharyngitis, and vomiting.
NCP®SSED, P970003
{9}
Page 9
The investigators reported that hoarseness, coughing, pharyngitis, and dyspnea were usually reported at the time of stimulation and were directly related to the strength of stimulation. Intolerable adverse events can generally be reduced or eliminated by a reduction in the output current, ON time, or increase in OFF time.
Most of the reported Study events were mild and well tolerated; very few clinical Study patients discontinued therapy due to side effects.
## 9. SUMMARY OF PRE-CLINICAL STUDIES
### Pre-clinical Studies on Leads
The helical electrode component of the Leads used in the NCP® System was developed by Huntington Medical Research Institute (HMRI) and the National Institute of Neurological Disorders and Stroke (NINDS) Neural Prosthesis Program. A number of animal studies have been performed at HMRI that showed that damage associated with peripheral nerve stimulation is of two types: mechanical and electrical. Mechanical damage, manifest as connective tissue changes in the perineurium and epineurium, is apparently caused by physical trauma to the nerve, although a chemical interaction has not been ruled out. Mechanical damage can be minimized by pliable helical electrode arrays of the appropriate size which will not compress or otherwise traumatize the nerve, and by routing and stabilizing the Lead to minimize electrode movement on the nerve. Electrical damage results in changes to myelin and the axons it surrounds. Larger, more heavily myelinated axons are most susceptible. This type of damage can be minimized by lower stimulation frequencies, less total stimulation, and lower duty cycles.
### 9.1 Risk Analysis
The implantable portions of the NCP® System employ known implantable materials generally considered to be biocompatible and safe for permanent implant use. The Company has conducted a risk analysis which included a system hazard analysis, fault tree analysis, component reliability analysis, drift analysis and failure modes and effects analysis. The results of these and other evaluations provide evidence of the safety of the NCP® System.
### 9.2 Components Testing
#### Integrated circuits:
Each component was subjected to qualification tests by the Applicant. These tests include 1000-hour life testing, high temperature storage, temperature shock, exposure to high temperature/high humidity, for integrated circuits. All qualification tests were successfully completed.
#### Battery:
The Model 8602 battery has been subjected to a series of safety and qualification tests, including vibration, temperature cycling, hermeticity, low pressure, low temperature, high temperature, mechanical shock, soldering heat, high and low temperature storage, voltage reversal, incineration, impact shock, puncture at high velocity, slow dent and puncture, crush and recharge testing. No unexpected behavior of the cells was observed. Short-circuit behavior was also characterized; the cells did not explode or vent, but the cell temperature, measured at body temperature, rose from 37 to 60 °C under an extremely low impedance discharge path. The
NCP® SSED, P970003
{10}
Applicant added a 10-ohm current limit resistor to the Generator, which limited heating during a short to a negligible amount in qualification tests.
## Feedthoughts:
The feedthrough utilized in the Model 100 NCP® Generator has undergone qualification testing in accordance with applicable portions of ASTM/ANSI Standard F18, ASTM F134-85, and MIL-STD 883B, which consisted of thermal shock, mechanical shock, vacuum bake, vibration, solvent resistance (ten samples), and terminal strength. All samples passed all test requirements.
## 9.3 Device Testing
Qualification testing was conducted to provide adequate data to support the intended use of the device system.
"Type testing" for CE Mark approval was conducted in accordance with prEN 45502-1:1993, Active Implantable Medical Devices and the requirements of Annex 1 and Annex 3 of the Commission of the European Communities Council Directive 90/385/EEC, amended by 93/42/EEC and 93/68/EEC.
Qualification tests were largely based on commonly recognized test methods and standards, such as military standards; International Standards Organization (ISO), European Standards (EN), European Committee for Standardization (CEN/CENELEC), and International Electrotechnical Commission (IEC) standards; American Society for Testing and Materials (ASTM), National Safe Transit Association (NSTA), Medical Device Standards (MDS), and United States Pharmacopeia (USP) standards; and others. More than 130 qualification tests have been performed on the devices or their components.
## Model 100 Generator
The Model 100 NCP® Generator was evaluated using electrical, mechanical, electromagnetic interference, and firmware testing.
### Electrical:
The electrical qualification tests consisted of
- Electrical characterization, End of Service characterization, and Communication Distance Characterization
- Direct Current leakage / Charge Balance per EN50051:1988,
- Electrosurgery Immunity - CENELEC "Standard for the Safety of Implantable Cardiac Pacemakers" (1986) and prEN45502-1 (May 1992),
- Defibrillation Immunity - prEN45502-1 (May 1992),
- Electromagnetic Interference (EMI) - EN50061:1988/prA1:1991; MDS201-0004 with modifications to testing facility and immunity levels: tested at 20V/m for frequencies from 20 MHz to 1 GHz and high field strengths at communication
- Electromagnetic Compatibility - MDS201-0004, IEC801-2 "Electromagnetic Compatibility for Industrial Process Measurement and Control Equipment, Part 2: Electrostatic Discharge (ESD) Requirements."
NCP® SSED, P970003
{11}
- Cellular phone (digital/GSM and analog) compatibility frequencies (including Cell Phone frequencies) of 450 MHz (386 V/m), 900 MHz (206 V/m), 1.45 GHz (213 V/m), and 1.9 GHz (96 V/m).
- Magnetic Resonance Imaging (MRI) compatibility -should not be done with the MR body coil. MRI should only be done using a head coil.
## Mechanical
The mechanical qualification tests consisted of the following:
- Temperature Cycle - prEN45502-1 (MAY 1992), Sec. 27.1
- Shock - prEN45502-1 (MAY 1992), Sec. 24.1
- Vibration - prEN45502-1 (MAY 1992), Sec. 24.2
- Diagnostic Ultrasound - prEN45502-1 (MAY 1992), Sec. 23.1
- Gross Leak - prEN45502-1 (MAY 1992), Sec. 26.1
- Fine Leak - Helium Leak Test performed 100% in production
- Wet Rub - prEN45502-1 (MAY 1992), Sec. 14.1
- Header Adhesion - (Header sheer test, strength exceeds 50 lb.)
- Header Seal (Isolation) - prEN45502-1 (MAY 1992), Sec. 16.1
- Shipping Shock - (NSTA - 30 inch drop onto each box corner and face)
- Shipping Vibration - (NSTA - 14,200 vibratory impacts)
- Shipping Atmosphere Pressure Changes - prEN45502-1, Sec. 12.1
- Wet Wipe - prEN45502-1 (MAY 1992), Sec. 9
- Shipping Humidity - prEN45502-1 (MAY 1992), Sec. 10.1
The Model 100 Generator performed under all conditions in accordance with its design specifications.
## Model 300 Series Lead
The Model 300 Series Lead was evaluated by electrical and mechanical testing.
### Electrical
- Electrode electrical characterization, $V_1$ and dV/dt pre and post sterilization
- Lead tubing leakage current - (3 month soak at +75C in saline).
### Mechanical
- Lead Dielectric - prEN45502-1, Sec, 16.2,
- Lead Tensile (Mandrel Wrap) - prEN45502-1, Sec, 24.3,
- Flexural Fatigue - prEN45502-1, Sec, 24.4,
- Connector Integrity - prEN45502-1, Sec, 24.5,
- Connector Insertion/Extraction Force, (5 kg force)
- Repeated Insertion, (10 cycles), Lead fatigue with tether, ($10^{+7}$ cycles), Lead fatigue without tether, ($10^{+7}$ cycles),
- Electrode weld pull test - (Shear pull to destruction - 310 to 540 g),
- Lead connector crimp pull test - (Pull to destruction - 0.90 to 1.05 Kg),
- Tubing bond pull test - (Pull to destruction 10 assemblies - 1.0-1.5 Kg),
Page 11
NCP* SSED, P970003
{12}
Page 12
- Other tests: Connector leak test - 10 psi, Lead Connector Crimp Pull Test, Connector Leak Test, Tubing Bond Pull Test, Electrode Weld Pull Test, Electrode Fatigue, Electrode Helix Force, Electrode Anchor Tether Pull-Off (90°), and Electrode Pull Test (axial)
- MRI Compatibility
The Model 300 Series Lead performed under all conditions in accordance with its design specifications. However, exposure to MRI resulted in excess heating of the electrode whenever the exposed Lead length was greater than 10 cm, when it was positioned against the outer bore of the MRI instrument and when worst case scan conditions were used. No heating was observed under worst case conditions of a fully extended Lead, when the Lead was positioned in the center of the bore, or in any position for a Lead shortened to 10 cm. Final product labeling included the following PRECAUTION:
**Magnetic Resonance Imaging (MRI)** should not be done with the MR body coil. The heat induced in the leads by a body MRI scan can cause injury. MRI should only be done using a head coil. Conditions which have been tested include:
- Transmit and Receive coil type: Head Coil Only
- Static Magnetic Field Strength: ≤ 2.0 Tesla
- SAR: < 1.3 W/kg for 70 kg patient
- Time varying intensity < 10 Tesla /sec
## Model 200 Programming Wand
The Model 200 Programming Wand was evaluated by electrical and mechanical testing.
### Electrical:
- Electrical Characterization,
- Maximum Programming Distance,
- Evoked Potential Adapter Qualification,
- Electromagnetic Interference (EMI),
- Magnetic Field Susceptibility - 30 Hz to 100 kHz per requirements of RS101 MIL-STD 461D,
- Magnetic Field Emissions - magnetic field emissions testing per requirements of RE101 MIL-STD 461D.
- Quasi Static Electric Field - sinusoidally varying electric field at 0.5 Hz with field strengths of 500 V/m to 2000 V/m per requirements of FDA Reviewer Guidance document.
- Electric Field Susceptibility - radiated electromagnetic energy field rated at 3 V/m as specified in the requirements of IEC 801-3 and the FDA Reviewer Guidance documents.
- Fast Transient Burst - transients of 0.25, 0.50, and 1 kV coupled by way of a capacitive clamp to the signal Leads per requirements of IEC 801-4 and the Reviewer Guidance documents.
- Conducted Susceptibility - conducted electromagnetic energy as specified in CS114, of MIL-STD-461D and the FDA Reviewer Guidance documents.
- Radiated Emissions - conducted and radiated emissions testing per requirements of CISPR 11.
### Mechanical:
- Temperature Cycle - prEN45502-1, Sec. 27.1, (-20 °C to 55 °C, 2 cycles)
NCP® SSED, P970003
{13}
Page 13
- Free Fall Test - prEN45502-1, Sec. 24.1 (3 ft drop)
- Vibration - prEN45502-1, Sec. 24.2, Wet Wipe - prEN45502-1, Sec. 11
- Shipping Shock - (NSTA - 30 inch drop), Shipping Vibration - (NSTA - 14,200 vibratory impacts)
The Model 200 Programming Wand performed under all conditions in accordance with its design specifications.
## Model 250 Programming Software
The Applicant has carried out extensive development, evaluation and testing of the software in the NCP® System.
A software development procedure which governs the creation, development, review, testing, documentation, manufacturing release, discrepancy reporting and maintenance of the software products in the NCP® System was used. The Model 250 Programming Software and the Model 100 Generator software were verified and validated to perform in accordance with their requirement specifications.
## Model 400 Tunneling Tool
The Model 400 Tunneling Tool was evaluated by mechanical testing.
- non-destructive shear testing - (14 lb.)
- destructive pull testing of the bullet/shaft connection (230 lb.)
The Model 400 Tunneling Tool performed under all conditions in accordance with its design specifications.
## 9.4 Biocompatibility
The following testing has been conducted for all tissue contacting materials used in the Model 100 Generator and Model 300 Lead:
- Cytotoxicity (MEM elution method), Sensitization
- Intracutaneous reactivity, Systemic toxicity, acute (USP method)
- Pyrogenicity (USP method), Genotoxicity (Ames mutagenicity test) subcutaneous implant (USP method, rat model) for fourteen, thirty, and one hundred-five days duration, with investigation of histology and systemic effects
The results of these tests showed that the materials used in the Applicant’s implantable products, as processed by and for them, are biocompatible, non-toxic, and non-pyrogenic.
## 9.5 Sterilization and Shelf Life Testing:
The ethylene oxide sterilization process was validated. Testing was performed to evaluate the effectiveness of the recommended steam autoclave sterilization method for the Model 400 Tunneling Tool. The shelf life of the sterile product was validated for a period of two years.
## 9.6 Animal Testing
Animal studies pertaining to vagus nerve stimulation can be divided into those dealing primarily with efficacy and those related to safety.
NCP® RSED, P970003
{14}
Page 14
# 9.6.1 Animal Efficacy Studies
Animal models of epilepsy have been used to evaluate the efficacy of vagus nerve stimulation (VNS). A dog model was used to demonstrate the feasibility of vagus nerve stimulation to inhibit strychnine-induced seizures.
A second feasibility study, used the rhesus monkey to study the effects of vagus nerve stimulation on focal seizures induced by alumina gel. Long-term seizure frequency was reduced when the vagus nerve stimulation was delivered during a seizure in two of the four monkeys with alumina gel induced epilepsy. In the other two, seizures continued but became synchronous in time, i.e., they appeared at regular intervals. The protective effect observed during stimulation was found to be maintained for weeks when the therapy was discontinued, suggesting a carry over effect.
Three groups of investigators used rat models of epilepsy to investigate the effect of vagus nerve stimulation on seizures. One group used pentylenetetrazol (PTZ), 3-mercaptopropionic acid (3-MP), and maximal electroshock (MES) to induce seizures in rats. Another used a penicillin induced seizure model in rats to study the effects of vagus nerve stimulation on interictal spiking. The third group, used the kindling and PTZ models. These rat models are also used to evaluate potential AEDs.
Using the PTZ model, it was reported that vagus nerve stimulation affected two seizure phenomena: mild status epilepticus and discrete intense seizures. Muscle activity associated with status epilepticus was suppressed when enough current was delivered to recruit the vagus nerve C fibers and depress heart rate and respiration. Discrete seizures could be prevented by vagal stimulation, but not stopped once they started. Also using the PTZ model, another group reported that efficacy was proportionally dependent on the cumulative time of stimulation and that the inhibitory effect of stimulation decreases gradually after cessation of stimulation.
Using the MES model, it was reported that vagus nerve stimulation can abolish or reduce the tonic component of the convulsive response. The degree of inhibition was related to the fraction of C fibers stimulated and the frequency of stimulation (as measured by heart rate depression).
The effects of vagus nerve stimulation on seizures in rats produced by intraperitoneal injection of 3-MP appeared to be variable.
With penicillin induced seizures in the rat, VNS significantly reduced interictal spike frequency, and the spiking frequency remained lower for about three minutes after cessation of stimulation at levels which also reduced heart rate and respiration.
In the kindled rat model, VNS reduced seizure duration, and the amount of reduction was dependent on duration of stimulation prior to the seizure.
Overall, studies with a number of animal models for human seizure, which are also used to evaluate efficacy of AEDs, showed that the effect of vagus stimulation can be detected in various brain structures, that vagus nerve stimulation decreases the frequency and/or duration of seizures and that the effect is related to stimulation parameters and timing. These studies, however, did not determine whether these effects are specific to stimulation of the vagus nerve, and whether they are secondary to depressed heart rate and respiration.
26
NCI 88ED, P970003
{15}
Page 15
## 9.6.2 Animal Safety Studies
Two safety studies were performed in monkeys and one was performed in sheep. In the first of these, the effect of a titanium cuff electrode on the vagus nerve of monkeys was examined. The effect of vagus nerve stimulation on cardiac and gastric function was also examined. Of the five animals implanted, two received 72 hours of stimulation without evidence of clinically significant effects on the stomach lining, and no pro-arrhythmic cardiac changes were observed.
In the second monkey safety study, spiral platinum electrodes were used. Six animals were exposed to electrodes for a minimum of 14 weeks and received at least 72 hours of 143 Hz, 50% duty cycle stimulation. Cardiovascular system effects included transient asystole and bradycardia. No stomach ulcerations were found. Histological analysis of the nerves indicated that stimulation with the spiral electrode did not result in electrical damage, and no mechanical damage to the nerve occurred in subjects in which proper electrode placement was achieved through adequate strain relief on the lead. Compression damage to large axons was found in 2 of the 4 experimental animals.
The effects of stimulation of the vagus nerve of sheep was studied in nine animals. Nerves from three of the animals which had been exposed to the system for at least ten months and which had received at least three months of stimulation were harvested. No evidence of nerve fiber degeneration, regeneration, or injury was present in any of the nerves analyzed, whether or not the nerve was stimulated. In all three animals studied for histology, epineural fibrosis and fatty infiltration of the nerve was found.
Overall, studies in animals indicate that stimulation of the vagus nerve may be conducted safely, with no damage to the axons of the vagus nerve. The spiral electrode design used in the NCP® System appears to be appropriate for chronic therapy.
## 10. SUMMARY OF CLINICAL STUDIES
Five acute phase studies were conducted. All patients that exited the acute phase studies were also followed in two long-term follow-up studies. The total stimulation exposure during these studies is greater than 900 patient years.
## 10.1 Design and Methods
### 10.1.1 Objective of studies
The overall objective of the clinical trial program was to collect clinical data regarding the NCP® System as an implantable vagus nerve stimulator indicated for use as adjunctive therapy in patients with medically refractory partial onset seizures.
The primary objective of the studies was to demonstrate a between group difference in mean and/or median percent change in seizure frequency between patients treated with “High” stimulation (optimal stimulation for seizure reduction) and “Low” stimulation (not expected to result in as great a seizure reduction).
Secondary objectives included the effect of VNS on within group (High vs. Low) changes in seizure frequency, response rates, number of seizure-free days, seizure intensity and duration, global evaluations by patients, caregivers and investigators and QOL measures.
NCP®USED, P970003
{16}
Page 16
# 10.1.2 Design of studies
Two of the studies (E01 and E02) were multicenter pilot studies, one (E04) was an open-label controlled longitudinal study, and two of the studies (E03 and E05) were multicenter, prospectively randomized, double blinded, parallel, active control studies. In the studies, treatment was adjunctive to AEDs; in E03 and E05, AEDs were to be held constant.
Figure 2. E01, E02, E04 Study Designs and Timelines
| Qualification/ | | Long-Term | | | | |
| --- | --- | --- | --- | --- | --- | --- |
| Historical Baseline ⇒ | ↔Baseline↔ | ↔Rec↔ | ↔C1↔ | ↔Stim↔
(Acute Phase) | ↔C2↔ | Follow Up Stim↔ |
| | Weeks 0-4 weeks | 2 Weeks | 3 - 4 Weeks | 12-16 Weeks | 0-4 Weeks | Indef.↔ |
KEY: S Surgery
Rec Post-surgical recovery period
C1, C2 Control periods
S1 - S4 4-week treatment periods
Analysis of the E01 and E02 acute phase mean percent seizure frequency reduction was based on the average mean percent change in seizure frequency over the Acute Phase Stimulation (16 weeks) in comparison to the first placebo stimulation control period (C1). Analysis of the E04 acute phase mean percent seizure frequency reduction was based on the average mean percent change in seizure frequency over the Acute Phase Stimulation (12 weeks) in comparison to the pre-implant Baseline Period
Patient's % Change = [seizure frequency/day]treatment - [seizure frequency/day]Baseline x 100
[seizure frequency/day]Baseline
NCI® SSED, P970003
{17}
Page 17
Figure 3. E03, E05 Study Designs and Timelines
E03 and E05 Study Design
A = Admission
S = Surgery
R = Randomize
For E05, V6 was after two weeks of stimulation, followed by V7, V8, and V9 corresponding to V6, V7, and V8 in E03.
Analysis of the E03 and E05 acute phase seizure frequency reduction was based on the difference in the average mean and median percent change in seizure frequency for the HIGH Group vs. the LOW Group over the first three stimulation periods (12 weeks) in comparison with the 12 week baseline period.
Analysis of the Long Term Follow-up Study (for Studies E01 through E04) and XE5 Study seizure frequency was based on the difference in the average mean and median percent change in seizure frequency over the three stimulation periods ending at 6, 9 (XE5 only), 12, 18, 24, and 36 months in comparison with the control period used in the respective acute phase studies.
The following stimulation parameters were used in the Studies.
Table 2. Protocol Stimulation Parameters
| Parameter | STUDY | | | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- |
| | E01 | E02 | E04 | E03 HIGH (LOW) | E05 HIGH (LOW) | XE1-4 | XE5 3 mo. (Longer) |
| CURR (mA) | 1 - 6 | 0.5 to 5 | ≤ 1 | 0.5 - 3.0 | 3.5 | 0-12 | 0-3.5 |
| FREQ (Hz) | ≤ 50 | 20 to 50 | 30 | 20 - 50 (1 to 2) | 30 (1) | 1-143 | 20 (1-30) |
| PW (μsec) | 250 | 250 to 500 | 500 | 500 (130) | 500 (130) | 130-1000 | 750 (130-1000) |
| ON (sec) | 60 | 30 to 60 | 30 | 30 - 90 (30) | 30 (30) | 7-270 | 30 (7-60) |
| OFF (min) | 60 | 5 to 20 | 10 | 5 (90) | 5 (180) | 0.2-180 | 1.8 (1.1-180) |
NCP® SSRD, P970003
{18}
Page 18
The parameters used in Studies E03, E04, and E05 were similar, whereas Studies E01 and E02 used higher frequencies, longer OFF times and shorter pulse widths.
## 10.1.3 Statistical methods
Standard statistical tests were used in comparing changes in seizure frequency between stimulation and control or baseline periods for Studies E01, E02, and E04 between and within HIGH and LOW stimulation groups for E03 and E05. The statistical test employed most extensively in E03 was the Student's t-test with the Wilcoxon rank-sum test used as a corroborative non-parametric test. Fisher's exact test was used to compare equality of proportions between HIGH and LOW stimulation groups. All tests were two-tailed and normality assumptions were checked as required.
E05 was analyzed in essentially the same manner as E03, with the exception that different statistical tests were applied. The null hypothesis that the two treatment percent changes are equal was tested using an extended Mantel-Haenszel test with modified ridit scoring of this ordinal outcome. The ranking for the modified-ridits is done for all patients from all sites after adjustment for investigative site through the subtraction of the means for the sites.
As a supplementary parametric analysis, analysis of variance (ANOVA) techniques were carried out to analyze between-treatment group comparisons of percent change in mean seizure frequency. One ANOVA model had treatment, investigative site, and treatment-by investigative site as independent variables. A second model additionally included baseline seizure frequency as a covariate and a baseline seizure frequency-by-treatment interaction term. Treatment-by investigative site interaction was addressed in the ANOVA models even though the primary method for comparing treatments is a non-parametric extended Mantel-Haenszel test.
All hypothesis tests were two-sided with an alpha of 0.05. Statistical significance was determined by p-values less than 0.05. A p-value between 0.05 and 0.10 was considered marginally statistically significant.
## 10.1.4 Patient Population and Collection Criteria
The final Inclusion and exclusion criteria were similar for each study.
NCP® SSED, P970003
{19}
Page 19
Table 3. Inclusion Criteria
| INCLUSION CRITERIA | E01 | E02 | E03 | E04 | E05 |
| --- | --- | --- | --- | --- | --- |
| Age | 18-60 | 18-60 | 12-60 | 2*-60 | 12-65 |
| Seizure type | SPS, CPS | SPS, CPS, 2ND | SPS, CPS, 2ND | ALL TYPES | SPS, CPS, 2^{ND} |
| # AED's | 1-2 | 1-2 | 0-3 | Not Specified | 1-3 |
| Good physical and health | yes | yes | Not Specified | Not Specified | Yes |
| Mental ability | IQ>80 | IQ>80 | Ability to understand IC, patient or guardian | Ability to understand IC, patient or guardian | Ability to understand IC, patient or guardian |
| Steady state serum levels | Yes | Yes, 1 month prior to study | Yes, +/- 20% over 3 months | Not Required | Yes, +/- 20% over 3 months |
| Seizure rate | 6/mo. | 6/mo | 6/mo | 1/mo | 6/mo with AOC |
| Max seizure free interval | 14 days | 14 days | 14 days | None Specified | 21 days |
| Women with acceptable birth control | Not Specified | Not Specified | Yes | Not Specified | Yes |
| Ability of patient or caregiver to accurately record seizures | Implied | Implied | Implied | Implied | Yes |
Table 4. Exclusion Criteria
| Exclusion Criteria | E01 | E02 | E03 | E04 | E05 |
| --- | --- | --- | --- | --- | --- |
| Living alone | Yes | Yes | Not Specified | Not Specified | Not Specified |
| Ulcers | Yes | Yes | Not Specified | Not Specified | Yes, peptic |
| Diabetic | Yes | Yes | Not Specified | Not Specified | Not Specified |
| Prior cervical vagotomy | Yes | Yes | Yes | Yes | Yes |
| Significant heart, lung or chronic bowel | Yes (heart and lung) | Yes | Yes (heart and lung) | Not Specified | Yes (heart and lung) |
| Medical condition likely to deteriorate | Yes | Yes | Yes | Yes | Yes |
| Investigational drugs | Not Specified | Not Specified | within 2 wk. + 5X half life | Not Specified | Within 2 wk. + 5X half life |
| History of pseudo seizures | Implied | Implied | Implied | Not Specified | Yes |
E05 patients were excluded for prior VNS treatment or any other stimulation treatment for epilepsy, as well as prior resective surgery for epilepsy. E05 patients were also excluded if they had experienced more than two episodes of status epilepticus within the last year.
NCP® SSED, P970003
{20}
Page 20
* E04: Patients under 12 years old had to meet the following additional criteria:
1. During the last five years (or over lifetime if younger) at least three anticonvulsants must have been tried (alone or in combination) to control seizures.
2. The Investigator and family must be of the opinion (and Investigator must so document in the chart) that seizures and/or drug side effects are detrimental to the patient.
3. Admission must first be discussed with the Applicant’s staff and approval given on a case by case basis. The purpose of this condition is to provide Investigators and family with all current information, so that the best risk/benefit decision can be made for the patient.
The important differences in the inclusion/exclusion criteria are:
- E05 excluded patients with resective surgery and required patients to have at least 6 seizures/month with AOC.
- E04 included patients with all types of seizures and allowed children down to 2 years of age.
## 10.2 Description of Patients Studied and analyses for Gender Bias
### 10.2.1 Patients Studied
Overall the populations across Studies was comparable as displayed in Table 4.
Table 5. Description of Study Patients
| Study No. | Group | Number of Patients | Gender | Average Age, Range | Average Years with Epilepsy, Range | Seizures/ Day | Ave. No. AEDs |
| --- | --- | --- | --- | --- | --- | --- | --- |
| E01 | --- | 11 | 7 M, 4 F | 31.6 (20-58) | 22.2 (13-32) | 2.95 mean 0.70 median | 1.0 |
| E02 | --- | 5 | 3 M, 2 F | 33.0 (18-42) | 20.3 (5-36) | 0.44 mean 0.44 median | 1.0 |
| E04 | --- | 124 | 67 M, 57 F | 23.9 (3.6-63) | 17.4 (0.8-48) | 25.2 mean 0.65 median | 2.2 |
| E03 | HIGH | 57 | 35 M, 22 F | 32.9 (20-57) | 22.4 (6-47) | 1.45 mean, 0.7 median | 2.1 |
| | LOW | 57 | 36 M, 21 F | 33.7 (13-50) | 20.5 (4-44) | 1.76 mean, 0.85 median | 2.1 |
| E05 | HIGH | 95 | 49 M, 46 F | 32.1 (13 - 54) | 22.1 (2 - 52) | 1.59 mean, 0.58 median | 2.2 |
| | LOW | 103 | 44 M, 59 F | 34.2 (15 - 60) | 23.7 (2 - 48) | 0.97 mean, 0.51 median | 2.1 |
32
NCI-SSED, P970003
{21}
Page 21
# Table 6. Description of Clinical Studies
All patients enrolled in all clinical studies, N=537
| Study | E01 | E02 | E04 | E03 | E05 | Total |
| --- | --- | --- | --- | --- | --- | --- |
| Type of Study | Pilot Longitudinal | Pilot Longitudinal | Open Longitudinal | Randomized Parallel High/Low | Randomized Parallel High/Low | - |
| Patients Enrolled | 11 | 5 | 133 | 126 | 262 | 537 |
| # Centers^{a} | 3 | 2 | 24 | 17 | 20 | 45 |
| Reference (baseline) period | Weeks 2 through 4 | Weeks 3 through 6 | Weeks -4 through 0 | Weeks -12 through 0 | Weeks -12 through 0 | - |
| Seizure Type | Partial | Partial | All types | Partial | Partial | - |
| Num. AEDs | 1 to 2 | 1 to 2 | not specified | 0 to 3 | 1 to 3 | - |
a Total includes OUS centers (Canada, Holland, Germany 2, and Sweden), several US centers participated in more than one study
## 10.2.2 Gender Bias
Inclusion and exclusion criteria were designed and carried out to avoid gender bias in patient enrollment. Of all patients enrolled 254 of 454 (54%) were male. This proportion (254/210 =1.21) of males is consistent with the male to female incidence (1.1 to 1.7) for nonfebrile epilepsy.¹
Separate analyses of safety and effectiveness data for males and females indicated no differences between the genders; hence, the results presented in the following analyses are representative for both men and women.
## 10.3 Results – Effectiveness and Safety
### 10.3.1 Effectiveness
The following table provides efficacy data for all studies. The mean percent reduction in seizure frequency is consistent across all studies except E04. E03, E04 and E05 studies established efficacy with statistical significance, however, two interim analyses were performed on the E03 data.
The median reduction in seizure frequency in the trials consistently improved over time. However, this data is uncontrolled and may be due to changes in drug therapy, rather than VNS.
Study E03 enrolled 115 patients, one patient dropped out prior to randomization/stimulation of the device and is not included in the efficacy tables below leaving an N=114. Study E05 enrolled 254 patients and had 55 baseline failures. In addition, 1 patient was explanted prior to randomization/stimulation of the device, 1 patient was excluded due to unreliable seizure diary counts, and 1 patient dropped from the study prior to collection of any effectiveness data leaving an N=196. All randomized/stimulated patients are included in the safety tables which follow.
¹ page 16 in: Hauser WA, Hesdorffer DC: Epilepsy: Frequency, Causes and Consequences. Demos Publications, New York, 1990, 378 pages.
NCI® SSED, P970003
{22}
There was no statistically significant change in seizure intensity or seizure duration in either the E03 or E05 Studies.
Table 7. Seizure Rate Changes
| ACUTE CHANGE IN SEIZURE FREQUENCY | | | | | EXTENSION PHASE FOLLOW-UP N, MEDIAN PERCENT SEIZURE REDUCTION | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Study | N | Mean % Seizure Reduction | Median % Seizure Reduction | Between Group Means p Value | 6 MO | 12 MO | 18 MO | 24 MO | 36 MO |
| E01 | 10 | 24.3% | 31.6% | N/A | 10
41.90 | 10
43.43 | 10
46.19 | 8
58.69 | 4
64.84 |
| E02 | 4 | 39.9%, | 48.1% | N/A | 4
3.01 | 4
30.04 | 3
67.64 | 2
60.32 | 2
31.88 |
| E04 | 116 | 7% | 21.8% | N/A | 107
32.46 | 86
26.81 | 72
41.41 | 34
32.18 | Insufficient data |
| E03 | 57 | 23.6% | 22.6% | 0.0175 | 108 | 102 | 71 | 51 | 49 |
| HIGH | | | | | 26.84 | 32.51 | 45.79 | 40.74 | 40.42 |
| E03 LOW | 57 | 6.4% | 6.3% | | | | | | |
| E05 | 94 | 27.9% | 23.4% | 0.0396 | 89 | 40 | N/A | 0 | 0 |
| HIGH | | | | | 32.5 | 34.2 | | | |
| E05 LOW | 102 | 15.2% | 20.5% | | | (9 mo.) | N/A | | |
| E01 | 244 | 11.8% | 17.1% | N/A | 229 | 202 | 156 | 95 | 55 |
| E04 Pooled | | | | | 29.75% | 31.31% | 44.25% | 40.74% | 40.42% |
| Total N for effic calculation | 440 | N/A | N/A | N/A | 318 | 242 | 156 | 95 | 55 |
NS = Not Significant
N/A = Not Applicable
NCP® SSED, P970003
{23}
Page 23
# Table 8. Responder Rates
| ACUTE
Responder Rate (% pts >50% Reduction) | | | | LONG TERM FOLLOWUP
Responder Rate (% pts >50% Reduction) | | | | |
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Study | N | Responder Rate | Between Group p Value | 6 MO | 12 MO | 18 MO | 24 MO | 36 MO |
| E01 | 10 | 30% | | 20%
(2/10) | 30%
(3/10) | 50%
(5/10) | 63% (5/8) | 50%
(2/4) |
| E02 | 4 | 50% | NA | 25%
(1/4) | 25%
(1/4) | 67%
(2/3) | 100%
(2/2) | 0%
(0/2) |
| E04 | 116 | 29.3% | | 37%
(40/107) | 31%
(27/86) | 39%
(28/72) | 35%
(12/34) | na |
| E03HIGH | 57 | 29.8% | 0.042 | 26%
(28/108) | 31%
(32/102) | 45%
(32/71) | 39%
(20/51) | 39%
(19/49) |
| E03LOW | 57 | 14.0% | | NA | NA | NA | NA | NA |
| E05HIGH | 94 | 23.4% | 0.171 | 39.3% | 40% | NA | NA | NA |
| E05LOW | 102 | 15.6% | | | (9 mo.) | | | |
| E01E04
Pooled | 244 | 25% | NA | 31.4% | 31.2% | 42.9% | 41.1% | 38.2% |
N/A = Not Applicable
35
NCP® SSED, P970003
{24}
Page 24
Figure 4. Change in Seizure Frequency, Patient Distribution
All Patients Completing Effectiveness Evaluation, N=196
| | Percent Change (seizures/day) from Baseline | | |
| --- | --- | --- | --- |
| | HIGH | LOW | Difference |
| N | 94 | 102 | 196 |
| Median | -23% | -21% | n/a |
| 25%, 75% Quartiles | -8.9%, -49% | 4.0%, -43% | n/a |
| 95% Confidence Intervals | -35%, -21% | -23%, -7.7% | -23%, -2.3% |
| Range (min, max) | -100%, 52% | -89%, 171% | -23%, -2.3% |
| Mean ± SD | -28% ± 34% | -15% ± 39% | -13%* ± 37% |
* Difference statistically significant (p<0.05) by Analysis of Variance (p=0.032) and by Cochran-Mantel-Haenszel Aligned Ranks (p=0.040)
36
NCP® SSED, P970003
{25}
Page 25
Table 9. Patient Summary Chart: Patients Continuing Treatment as of August 22, 1996
| STUDY PERIOD | E01 | E02 | E03 | E04 | Total |
| --- | --- | --- | --- | --- | --- |
| Patients Randomized/Stimulated (N) | 10 | 5 | 114 | 123 | 253 |
| Patients (N) Entering Extension Phase | 10 | 5 | 113 | 123 | 251 |
| Total continuing patients being treated for up to one year (N) | 10/10 | 5/5 | 111/115 | 112/121^{(4)} | 238/251 |
| Continuing patients being treated for up to two years (N) | 9/10 | 4/5 | 71/87^{(1)} | 58^{(2)}/70 | 142/173 |
| Continuing patients being treated for up to three years (N) | 7/10 | 3/5 | 57/87 | 21^{(3)}/24 | 88/127 |
(1) Twenty eight (N=28) commercial European patients were excluded from follow-up after one year of treatment due to the commercial release of the NCP® system in those countries.
(2) As of 8/22/96. Only 70 patients had been implanted long enough to reach the 2 year treatment period; only 58 of the 70 were continuing.
(3) As of 8/22/96, Only 24 patients had been implanted long enough to reach the 3 year treatment period; only 21 of the 24 were continuing.
(4) Two of the original 123 patients entering the extension phase had been implanted for less than one year as of 8/22/96.
A total of 238 out of a possible 251 completed one year of therapy (95%); 142 out of 172 completed two years of therapy (83%); and 88 out of 126 completed three years of therapy (70%). Although these numbers of patients remained in the study, only those patients who had completed the annual follow up visits are represented in the above efficacy calculations. NCP® System treatment did not cause changes in AED plasma concentrations as measured during studies E03 and E05.
Figure 5. Median Percent Change in Seizure Frequency
(1) Note: The acute phase results include seizure frequencies of the E03 Study LOW stimulation group, which includes one-half of the E03 patients, N=57 Patients were permitted to change their AEDs during these studies and these changes may have contributed to the change in seizure frequency.
NCP® SSED, P970003
{26}
Page 26
# Pediatric Patients
Study E05 allowed patients as young as 12 years to enter the study. There were 20 patients in the E05 Study between 12 and 18. Their response during the Study was a 22.5% median reduction and a 26.2% mean reduction in seizures compared to baseline.
# Global Rating
Global ratings by Investigators and patients support the conclusion that HIGH stimulation is associated with an improved quality of life, although improvements were seen in both groups. Patients, companions, and Investigators who were blinded to the treatment were asked to give a global rating of the patient's overall condition at each visit. A visual analog scale consisting of a 100 mm line was use. All Global scores for the Investigator, Patient and Companion are significantly improved over baseline. In the E05 study, the Investigators and Patients rated the HIGH group significantly improved over the LOW group.
# Quality of Life and Cognitive Functioning
The E03 and E05 Studies used several instruments to measure QOL and cognitive functioning, with those used in the E05 Study being more comprehensive. Observed changes in the health related quality of life (HQL) and psychosocial scale scores were generally more favorable in the HIGH VNS group. There were few significant correlations between changes in cognitive tests and changes in global ratings of patient’s overall well-being and seizure activity.
38
NCP® SSED, P970003
{27}
Page 27
# Other secondary endpoints
There was a statistically significant difference in mean % change in seizure frequency for within group stimulation compared to baseline. There was no statistically significant difference between groups for the number of responders (≥50% seizure reduction), number of seizure free days, seizure intensity or duration.
## 10.3.2 Safety
In each of the studies, the system was safe and well tolerated. The following treatment emergent signs and symptoms were observed in the Studies. Almost all were mild or moderate and anecdotally reported by Investigators to occur when the stimulator was ON and the intensity of the AE’s were related to the strength (output current) of the stimulation. Symptoms that decreased or disappeared are not included in this analysis. In the E05 Study a symptom check list was used to prompt patients for adverse event reporting to ensure complete reporting of all events. Use of the checklist is likely the cause of slightly higher adverse event reporting rates in E05 Study as compared to E03.
(n=413 devices in 314 patients, 152 in HIGH (optimal stimulation) treatment group, 591 device years)
Table 9. Observed Adverse Events
| | Randomized + Extension Phase, N= 314 patients, 591 device years | | | | Randomized Phase, HIGH Only, N = 152 pts | |
| --- | --- | --- | --- | --- | --- | --- |
| Adverse Event | # of patients | % of patients | # of Events | Events per device-year | # of patients | % of patients |
| Serious AEs1 | | | | | | |
| Surgically Related | 13 | 4.1 % | 13 | 0.022 | | |
| Stimulation Related | 4 | 1.2 % | 4 | 0.007 | 1 | 0.7 % |
| Non-serious AEs | | | | | | |
| Voice Alteration | 156 | 50% | 720 | 1.228 | 91 | 60% |
| Cough Increased | 129 | 41% | 456 | 0.772 | 57 | 38% |
| Pharyngitis | 84 | 27% | 182 | 0.308 | 36 | 24% |
| Paresthesia | 87 | 28% | 377 | 0.638 | 32 | 21% |
| Dyspnea | 55 | 18% | 55 | 0.093 | 32 | 21% |
| Dyspepsia | 36 | 12% | 98 | 0.166 | 22 | 15% |
| Nausea | 59 | 19% | 154 | 0.261 | 21 | 14% |
| Laryngismus | 10 | 3.2% | 30 | 0.051 | 9 | 5.9% |
1 - Serious AEs reported included infection, nerve paralysis, hyperesthesia, facial paresis, left vocal chord paralysis, left facial paralysis, left recurrent laryngeal nerve injury, urinary retention, and low grade fever
NCP® SSED, P970003
{28}
Page 28
# Table 10. Adverse Events Reported During Baseline or Stimulation
Test of Difference between Baseline (12 Weeks) and Stimulation (14 Weeks)
| Adverse Event | E05 | | | | | |
| --- | --- | --- | --- | --- | --- | --- |
| | HIGH | | | LOW | | |
| | Baseline | Treatment | p-value^{a} | Baseline | Treatment | p-value^{a} |
| Cough | 28.4 | 52.6 | <0.0001 | 23.3 | 51.5 | <0.0001 |
| Dyspepsia | 5.3 | 21.1 | 0.0011 | 13.6 | 15.5 | 0.6171 |
| Dyspnea | 4.2 | 27.4 | <0.0001 | 6.8 | 14.6 | 0.0114 |
| Infection | 4.2 | 14.7 | 0.0184 | 8.7 | 15.5 | .1615 |
| Pain, nonspecific | 20.0 | 33.7 | 0.0124 | 19.4 | 37.9 | 0.0030 |
| Paresthesia | 1.1 | 24.2 | <0.0001 | 4.9 | 33.0 | <0.000 |
| Throat pain^{b} | 15.8 | 42.1 | <0.0001 | 16.5 | 29.1 | 0.0236 |
| Voice alteration/hoarseness | 6.3 | 72.6 | <0.0001 | 8.7 | 32.0 | <0.0001 |
| Vomiting | 8.4 | 17.9 | 0.0389 | 6.8 | 14.6 | .0325 |
| Adverse Event | E03 | | | | | |
| --- | --- | --- | --- | --- | --- | --- |
| | HIGH | | | LOW | | |
| | Baseline | Treatment | p-value^{a} | Baseline | Treatment | p-value^{a} |
| Cough | 0.0 | 12.3 | 0.008 | 3.5 | 10.5 | .157 |
| Dyspepsia | na | na | na | na | na | na |
| Dyspnea | 1.8 | 10.5 | 0.059 | 0.0 | 0.0 | na |
| Infection | 0.0 | 3.5 | 0.157 | 1.8 | 3.5 | .564 |
| Pain, nonspecific | na | na | na | na | na | na |
| Paresthesia | 0.0 | 15.8 | 0.003 | 0.0 | 7.0 | 0.046 |
| Throat pain^{b} | 0.0 | 7.0 | 0.046 | 0.0 | 5.3 | 0.083 |
| Voice alteration/hoarseness | 0.0 | 38.6 | <0.0001 | 0.0 | 14.0 | 0.005 |
| Vomiting | 1.8 | 1.8 | ns | 1.8 | 1.8 | ns |
A As reported or observed
a As elicited using Symptoms Checklist, reported or observed
b Within group analysis
c Throat pain is specifically reported as an AE in E03 but not E05. For comparative purposes a separate analysis of E05 data, aggregating neck pain, Pharyngitis and Laryngismus, was done and is presented here
Statistically significant < 0.05
Marginally significant 0.05 < p < 0.10
NCP® SSED, P970003
{29}
Page 29
# Table 11. Treatment Emergent Adverse Events By Severity
| | AE's Reported During Baseline and Stimulation | | | Treatment Emergent AE's by Severity (% of reports) | | |
| --- | --- | --- | --- | --- | --- | --- |
| Adverse Event | E05 HIGH Group | | | Severity Ratings | | |
| | Baseline^{A} | Treatment^{A} | pvalue^{c} | Mild | Moderate | Severe |
| Cough | 28.4% | 52.6% | .0001 | 74% | 23% | 3% |
| Dyspepsia | 5.3% | 21.1% | .0011 | 76% | 24% | 0% |
| Dyspnea | 4.2% | 27.4% | <.0001 | 46% | 54% | 0% |
| Hypesthesia | 0.0% | 5.3% | .0253 | 100% | 0% | 0% |
| Infection | 4.2% | 14.7% | .0184 | 70% | 30% | 7% |
| Insomnia | 1.1% | 5.3% | .0455 | 75% | 25% | 0% |
| Pain, NonSpecific | 20.0% | 33.7% | .0124 | 63% | 33% | 4% |
| Paresthesia | 1.1% | 24.2% | <.0001 | 82% | 18% | 0% |
| Throat Pain & Laryngismus/Pharyngitis | 15.8% | 42.1% | <.0001 | 55% | 43% | 2% |
| Voice Alteration/Hoarseness | 6.3% | 72.6% | <.0001 | 73% | 27% | 0% |
| Vomiting | 8.4% | 17.9% | .0389 | 76% | 24% | 0% |
*AE's that are > 10% in the E05 HIGH stimulation group and were statistically significantly different from baseline are included.*
Since the E03 and E05 Studies were active controlled studies as opposed to the traditional placebo controlled studies, a comparison of adverse events to baseline is required in order to obtain a more complete understanding of the differential effects of VNS therapy. A within group analysis of AEs for both HIGH and LOW groups in the E03 and E05 Studies in comparison to baseline AE rates is shown in the table above. Ninety nine percent of the side effects were rated as mild or moderate.
NCP® SSED, P970003
{30}
Page 30
Table 12. Treatment Emergent Signs and Symptoms (>2%)
| Study Number | E01 | E02 | E04 | E03 | | E05 | | XE01 04 | | XE5 |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Group | | | | HIGH^{1} | LOW^{2} | HIGH | LOW | 1 YR | 2 YR | 9 MO |
| N | 11 | 5 | 123 | 57 | 57 | 95 | 103 | 244 | 166 | 41 |
| BODY AS A WHOLE | | | | | | | | | | |
| Asthenia | | | | | | 7.4 | 1.9 | | | 12.2 |
| Chills | | | | | | | | | | 2.4 |
| Cyst | | | | | | 1.1 | 1.0 | | | 4.9 |
| Edema Face | | | | | | 0.0 | 1.0 | | | 2.4 |
| Fever | | | 7.3 | 3.5 | 1.8 | 11.6 | 18.4 | 0.8 | 1.2 | 2.4 |
| Flu Syndrome | | | | | | 1.1 | 3.9 | | | 2.4 |
| Headache | 60 | 20 | 8.9 | 5.3 | 7.0 | 24.2 | 23.3 | 4.5 | 1.8 | 17.1 |
| Infection | | | | 3.5 | 3.5 | 11.6 | 11.7 | 2.1 | 1.2 | 4.9 |
| Injury Accident | | | 5.7 | | | 12.6 | 12.6 | 2.1 | 1.2 | 14.6 |
| Malaise | | | | | | 1.0 | 0.0 | | | 4.9 |
| Overdose | | | | | | 0 | 2.9 | | | |
| Pain | 30 | | | 8.8 | 1.8 | 28.4 | 30.1 | 0.4 | 1.2 | 31.7 |
| Pain Abdominal | 20 | | 4.1 | 3.5 | 3.5 | 6.3 | 6.8 | 2.1 | 0 | 9.8 |
| Pain Back | | | | 0 | 1.8 | 0 | 3.9 | 0 | 0 | 9.8 |
| Pain Chest | 20 | | | 1.8 | 0 | 9.5 | 12.6 | 0.8 | 0 | 4.9 |
| Pain Neck | 30 | | | | | 3.2 | 1.0 | | | 4.9 |
| CARDIOVASCULAR SYSTEM | | | | | | | | | | |
| Arrhythmia | | | | | | | | 0 | 0 | 2.4 |
| Arrhythmia Vent | | | | | | | | | | 2.4 |
| Fibrillate Atria | | | | | | | | | | 2.4 |
| Hemorrhage | | | | | | 3.2 | 3.9 | | | |
| Hypertension | | | | | | 4.2 | 1.0 | 0 | 0 | 7.3 |
| Phlebitis | | | | | | 2.1 | 0 | | | |
| Tachycardia | | | | | | 2.1 | 2.9 | 0.4 | 0 | |
| DIGESTIVE SYSTEM | | | | | | | | | | |
| Appetite Increase | 10 | | | | | | | | | |
| Anorexia | 10 | | | | | 1.1 | 1.9 | 1.2 | 0 | 4.9 |
| Constipation | | 20 | 2.4 | 0 | 0 | 2.1 | 7.8 | 0.4 | 0 | 2.5 |
| Diarrhea | 20 | | 4.1 | 3.5 | 0 | 6.3 | 24.3 | 1.2 | 1.2 | 12.2 |
| Dyspepsia | | | | | | 17.9 | 12.6 | | | 9.8 |
| Dysphagia | 20 | 60 | 4.9 | 3.5 | 0 | 2.1 | 6.8 | 2.1 | 1.2 | 2.4 |
| Gingivitis | | | | | | 2.1 | 1.9 | | | |
| Nausea | 10 | | 2.4 | 7.0 | 1.8 | 14.7 | 20.4 | 0.8 | 1.2 | 2.4 |
| Vomit | 10 | | 4.9 | 0 | 1.8 | 17.9 | 13.6 | 1.6 | 1.8 | 4.9 |
| HEMIC AND LYMPHATIC SYSTEM | | | | | | | | | | |
| Anemia | | | | | | 0.0 | 1.0 | | | 2.4 |
| Ecchymosis | | | | | | 3.2 | 3.9 | | | 7.3 |
| METABOLIC AND NUTRITIONAL DISORDERS | | | | | | | | | | |
| Cyanosis | | | | | | 0 | 2.9 | | | |
NCP® SSED, P970003
{31}
Page 31
Table 12. Treatment Emergent Signs and Symptoms (>2%)
| Study Number | E01 | E02 | E04 | E03 | | E05 | | XE01 04 | | XE5 |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Group | | | | HIGH^{1} | LOW^{2} | HIGH | LOW | 1 YR | 2 YR | 9 MO |
| N | 11 | 5 | 123 | 57 | 57 | 95 | 103 | 244 | 166 | 41 |
| Edema | | | | | | 3.2 | 6.8 | | | 2.5 |
| Edema Periph | | | | | | 0.0 | 1.0 | | | 2.4 |
| Weight Decrease | | | | | | 1.1 | 2.9 | | | 2.4 |
| MUSCULO SKELETAL SYSTEM | | | | | | | | | | |
| Arthralgia | | | | | | 2.1 | 1.0 | 0.4 | 0 | |
| Bone Fract Spontan | | | | | | 5.3 | 2.9 | | | 4.9 |
| Myalgia | 30 | | 6.5 | | | 2.1 | 1.0 | 1.2 | 0 | 4.9 |
| Myasthenia | | | | | | | | | | 2.4 |
| NERVOUS SYSTEM | | | | | | | | | | |
| Agitation | | | | 0 | 1.8 | 1.1 | 2.9 | 1.6 | 0 | |
| Anxiety | 10 | | | | | 1.1 | 1.0 | | | 2.4 |
| Ataxia | 10 | | | 0 | 0 | 4.2 | 3.9 | 0.4 | 0.6 | 9.8 |
| Coordination
Abnormality | | 20 | | | | | | | | |
| Convulsions | | | | | | 2.1 | 3.9 | | | |
| Depression | 10 | | 0.8 | 1.8 | 1.8 | 3.2 | 0 | 0 | 0.6 | 12.2 |
| Diplopia | | | | | | 5.3 | 2.9 | 0.4 | 0.6 | |
| Dizziness | | | | 0 | 3.5 | 5.3 | 3.9 | 3.3 | 0.6 | 4.9 |
| Dream Abnormality | 10 | | | | | 1.1 | 0.0 | | | |
| Emotional liability | | | 2.4 | | | 1.1 | 1.9 | 1.2 | 0 | |
| Euphoria | 20 | | | | | | | | | |
| Gait Abnorm | | | | | | | | | | 2.4 |
| Hallucin | | | | | | | | 0.4 | 0 | 2.4 |
| Hostility | | | | | | 1.1 | 1.0 | | | 2.4 |
| Hypesthesia | | | | | | 2.1 | 0 | | | 2.4 |
| Insomnia | 10 | 20 | 2.4 | 5.3 | 1.8 | 4.2 | 4.9 | 0 | 0.6 | 2.4 |
| Movement Dis | | | | | | 1.1 | 0.0 | | | 2.4 |
| Neuralgia | | | | | | | | | | 2.4 |
| Nervousness | | | 3.3 | | | 1.1 | 1.9 | 1.6 | 0 | 2.4 |
| Neurosis | | | | | | | | | | 2.4 |
| Paresthesia | | | 5.7 | 5.3 | 3.5 | 17.9 | 25.2 | 0.8 | 2.4 | 41.5 |
| Person Disorder | | | | | | 1.1 | 0.0 | | | 2.4 |
| Somnolence | | | | 5.3 | 3.5 | 2.1 | 1.9 | 1.2 | 0 | 2.4 |
| Tremor | | | 2.4 | 1.8 | 1.8 | 3.2 | 2.9 | 0 | 1.2 | 2.4 |
| Twitch | 50 | 20 | 3.3 | 1.8 | 0 | 3.2 | 1.0 | 0.8 | 0 | 9.8 |
| Vertigo | | | | | | | | | | 2.4 |
| RESPIRATORY SYSTEM | | | | | | | | | | |
| Asthma | | | | | | 2.1 | 3.9 | | | 2.4 |
| Cough Increased | | | 17.1 | 8.8 | 8.8 | 45.3 | 42.7 | 2.9 | 3.0 | 17.1 |
| Dyspnea | 30 | 20 | 3.2 | 7.0 | 0 | 25.3 | 10.7 | 2.5 | 0 | 9.8 |
NCP® SSED, P970003
{32}
Page 32
# Table 12. Treatment Emergent Signs and Symptoms (>2%)
| Study Number | E01 | E02 | E04 | E03 | | E05 | | XE01 04 | | XE5 |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| Group | | | | HIGH^{1} | LOW^{2} | HIGH | LOW | 1 YR | 2 YR | 9 MO |
| N | 11 | 5 | 123 | 57 | 57 | 95 | 103 | 244 | 166 | 41 |
| Hiccup | 10 | | | | | 3.2 | 0 | | | |
| Hyperventilate | | | | | | 2.1 | 1.0 | | | |
| Laryngismus | | | | | | 7.4 | 1.0 | | | 2.4 |
| Lung disorder | | | | | | 1.1 | 3.9 | | | 4.9 |
| Pharyngitis | 10 | 40 | 4.9 | 12.3 | 10.5 | 34.7 | 25.2 | 4.1 | 1.2 | 19.5 |
| Respiratory Disorder | | | | | | 2.1 | 0 | | | |
| Rhinitis | | | 11.4 | 0 | 3.5 | 7.4 | 14.6 | 1.6 | 1.2 | 4.9 |
| Sinusitis | | | | | | 3.2 | 2.9 | | | 4.9 |
| Voice Alteration | 100 | 80 | 13 | 36.8 | 10.5 | 66.3 | 30.1 | 7.8 | 5.4 | 75.6 |
| SKIN AND APPENDAGES | | | | | | | | | | |
| Application Site Reaction | | | | | | 1.1 | 2.9 | | | 2.4 |
| Alopecia | | | | | | 0.0 | 1.0 | | | 2.4 |
| Pruritis | | | | | | | | | | 4.9 |
| Rash | 10 | | | | | 5.3 | 1.9 | 0 | 0 | 4.9 |
| Rash Vesis Bull | | | | | | 2.1 | 1.0 | | | |
| SPECIAL SENSES | | | | | | | | | | |
| Amblyopia | | | | | | 2.1 | 1.9 | 0.4 | 0.6 | |
| Conjunctivitis | | | | | | 1.1 | 1.0 | | | 2.4 |
| Glaucoma | | | | | | | | | | 2.4 |
| Pain Ear | | 20 | | 0 | 1.8 | 4.2 | 3.9 | 1.2 | 0.6 | |
| Parosmia | | | | | | | | | | 2.4 |
| Taste Pervers | | | | | | | | 0 | 0 | 2.4 |
| Tinnitus | | | | 3.5 | 1.8 | 4.2 | 5.8 | 0 | 0 | |
| UROGENITAL SYSTEM | | | | | | | | | | |
| Dysmenorrhea | | | | | | 2.1 | 3.9 | | | |
| Dysuria | | | | | | 0 | 2.9 | 0.4 | 0 | |
| Hematuria | | | | | | 0.0 | 1.0 | | | 2.4 |
| Incontin Urine | | | | | | 1.1 | 0.0 | | | 2.4 |
| Infection, Urin Tract | | | | | | 2.1 | 6.8 | | | |
| Menstrual Disorder | | | | 3.5 | 0 | 1.1 | 0 | 0 | 0 | |
| Merorrhagia | | | | | | 2.1 | 1.9 | | | |
| Urination Frequency | | | | | | 0.0 | 1.0 | | | 4.9 |
1 = HIGH Stimulation parameters generally believed to provide maximum efficacy
2 = LOW Stimulation parameters generally believed to provide sensation of stimulation but less effective.
Cardiac monitoring: For Study E05 all patients had 24 hour Holter monitoring twice during baseline and five times during stimulation. Data from the Holter monitoring showed no difference between either of the Treatment groups, and/or Baseline periods on cardiac endpoints. Some patients in all acute and long term studies were monitored at baseline and stimulation with EKG
NCP® SSED, P970003
{33}
Page 33
or Holter recording. None of the monitoring showed any statistically significant differences between baseline and stimulation.
**Pulmonary monitoring:** Patients in the E05 Study were monitored once during baseline and twice during stimulation for FVC and FEV1. There was no statistical difference between the treatment groups, and/or baseline periods pulmonary endpoints. Although dyspnea was statistically significantly increased over baseline, no effect was detected with pulmonary function monitoring.
**Gastric monitoring:** All patients in E05 were monitored for serum gastrin levels and some patients in the other acute and long term studies were monitored for gastrin or gastric acid output. No clinically meaningful changes were observed.
## DEATHS AND OTHER SERIOUS ADVERSE EVENTS
A total of 17 deaths have been reported in 950 clinical and commercial patients followed for 1650 patient-years. Nine of the deaths were judged to be possible/probable/definite Sudden Unexpected Death in Epilepsy (SUDEP). Additionally, one patient died of SUDEP during the preimplant baseline period, and two patients have died after discontinuation of treatment.
Based on the 17 deaths, there was a mortality of 10.3/1000 patient-years and a SUDEP incidence of 3.0/1000 patient-years for definite/probable and 5 for definite/probable/possible SUDEPs.
Table 13. Deaths and Other Serious Adverse Events
| Study | N | No. of Deaths^{1} | Other Serious AEs |
| --- | --- | --- | --- |
| E01 | 11 | 0 | 1 vocal cord paralysis due to surgery, resolved. |
| E02 | 5 | 0 | 0 |
| E03 | 114 | 1 preimplant | 1 vocal cord paralysis, permanent, device failure
1 infection, hospitalization complete recovery
1 MI |
| E04 | 123 | 0 | 2 Infections; 1 hospitalization for transient increase in seizures; 1 poor appetite; 1 fever; 1 coughing. |
| E05 | 198 | 0 | Surgery & device related 3 infections which ultimately were explanted; 1 vocal cord paralysis; 1 left facial paralysis; 1 post operative fever; 1 localization of fluid around generator; and 1 left recurrent laryngeal nerve injury.
Therapy related: 1 near syncope; 1 Cheyenne Stokes respiration, and 1 bleeding duodenal ulcer. |
| XE14 | 229 | 7 | Definitely related: 1 respiratory distress; 1 obstructive bronchitis; 3 voice alteration; 1 choking
Probably related: 1 Change of seizure type
Possibly related: 1 Fall; 1 pulmonary congestion; 1 obstructive bronchitis. |
| XE5 (6 mo.) | 41 | 2 | 1 probable left hemidiaphragm paralysis, 1 hospitalization for seizures, 1 institutionalized for selfabusive behavior |
1 = Table does not include patients who died using a commercially available product, N = 8
NCP® SSED, P970003
{34}
Page 34
# 10.3.3 DEVICE PERFORMANCE
## DEVICE COMPLICATIONS AND OBSERVATIONS
For the purposes of this report, a device complication is an event which cannot be treated or resolved by reprogramming the device and requires intervention other than reprogramming. A device observation is an adverse event which can be corrected by reprogramming and which does not require surgical intervention
Table 14. Device Complications and Observations
| Study | N | NCP® System Complications | Description |
| --- | --- | --- | --- |
| E01 | 11 | 2 Lead | 1 high impedance, 1 broken wire |
| E02 | 5 | None | None |
| E03 | 114 | 3 Lead
2 Generator | 1 cut by surgeon, 1 Lead not fully inserted in Generator 1 Electrode not on nerve, but reduced seizures
1 Internal short, DC on electrode 1 Premature battery depletion, excessive drain from circuit |
| E04 | 123 | 0 | None |
| E05 | 198 | 0 | None |
| XE14 | | 54 Lead
5 Generator | 24 Lead breaks, 24 High lead impedance, 1 insulation, 1 infection, 1 patient discomfort, 1twiddler, 4Infection (explanted), 1 user error |
| XE5 (6 mo.) | 41 | 1 Lead | 1 Lead high impedance, Lead was twisted, no strain relief or tie down. Replaced Lead |
| Study | N | Observations | NCP® System Observations |
| E01 | 11 | 0 | 0 NCP® System |
| E02 | 5 | 0 | 0 NCP® System |
| E03 | 114 | 3 Generator
4 System
1 Programming Software | 2 reset, 1 communication
4 infections treated with antibiotics
1 exited to DOS, program was modified |
| E04 | 123 | 3 Lead
3 Generator
Several Programming
4 Programming Software | 1 high impedance, unverified, 2 helix stuck together
1 pt discomfort, 2 erratic stim, unverified
Programming difficulties Resolved by repositioning or turning off other equipment in the room
2 Operator error, 1 corrupted history file, 1 background program interfered with printing |
| E05 | 198 | 3 Generator
1 Lead
1 System
12 Communication | 2 Pain during diagnostic tests, User error, explained in Manual 1 Programmer operator error, explained in Manual
1 "Malformed" Lead at implant, used backup Lead returned Lead within specs.
1 infection
Resolved by repositioning Wand or turning off electrocautery equipment. Discovered disk caching routines on certain computers may interfere, instructed users to disable this function. |
NCP® SSED, P970003
{35}
Page 35
Table 14. Device Complications and Observations
| Study | N | NCP® System Complications | Description |
| --- | --- | --- | --- |
| XE14
8 yr. total | 229 | 23 Lead | 11 high impedance, 4 break, 3 user error, 2 connector, 2 helix stuck together, 1 dislodged |
| | | 36 Generator | 10 erratic stim (not duplicated), 8 reset, 8 user errors, 5 communication, 1 each: infection, magnet effectiveness, package anomaly, set screw, other |
| | | 8 Wand | 6 communication, 2 user error |
| | | 14 Programming Software | 7 communication, 3 software difficulties, 4 user error |
| XE5 6 mo. | 89 | 2 Generator | 2 reports of erratic stimulation; not verified |
| | | 2 Communication | 2 communication difficulties resolved by repositioning |
Two Generator complications have occurred related directly to device performance. One Generator malfunction caused by an internal short circuit and DC current on the ground electrode resulted in patient injury, and although the patient's nerve was believed to be permanently damaged the patient's voice fully recovered. The second was premature battery depletion due to a transient high current drain.
Over fifty Leads have been verified to have broken or developed high impedance. Design changes have been made as a result of these failures. Over 500 of the redesigned leads have reportedly been implanted, with no reported complications, other than one case where a suture was tied directly around the lead.
There have been a number of reported programming communication problems. These problems are related, in part, to a strict communication protocol which ensures reliable data transfer and programming, and in part to the use of electrocautery and other electrically noisy equipment in the vicinity of the programmer. The noise immunity of the receiver circuitry in the Programming Wand was improved, and disk caching routines in some computers was identified as a source and corrected.
## 11. CONCLUSIONS DRAWN FROM STUDIES
Valid scientific evidence is necessary to establish that there is reasonable assurance the NeuroCybernetic Prosthesis (NCP®) System is safe and effective for its intended use. The in vivo and in vitro nonclinical laboratory studies together with the clinical investigations provide reasonable assurance that the NeuroCybernetic Prosthesis (NCP®) System is safe when used as indicated in the labeling.
The results of clinical studies of the NCP® system for seizure reduction in patients with partial onset seizures support the effectiveness of the device. The mean seizure reduction experienced by the HIGH treatment groups (the group on whom labeling is based) was statistically significant as compared to the active LOW treatment control group. The HIGH group in both studies had a statistically significant mean seizure reduction as compared to baseline.
The safety of the NCP® System has been demonstrated in clinical trials with reported side effects being negligible in relationship to the severity of the disorder being treated. The most common
NCP® SSED, P970003
{36}
side effects reported are generally a function of the device’s programmed stimulation setting which can be altered or adjusted to lessen the effect or discontinue it altogether.
## 12. PANEL RECOMMENDATIONS
On June 27, 1997, the Neurological Devices Panel recommended that the Premarket Approval Application for the NeuroCybernetic Prosthesis (NCP®) System be approved with some specific labeling recommendations. The approved labeling followed those recommendations. The panel also recommended that post approval studies will be conducted to examine the morbidity and mortality associated with the device a…
