Development of Drugs, Devices, and DrugDevice Combinations: FDA for the Next Generation

1
(No Transcript)
2
CIRCULATORY SYSTEMDEVICES PANELWednesday, April
21, 2004

• Cordis Corporation
• Precise 5.5F and 6.0F OTW and RX nitinol Stent
  System
• Angioguard XP OTW and RX Emboli Capture Guidewire
  System
• PMA P030047
• Lead FDA Reviewer
• Lisa Kennell

3
Introduction

• Regulatory history of the Cordis system
• Non-clinical study summary
• Statistical Summary
• Clinical Summary
• Panel Questions

4
FDA Review Team

• Team Leader Lisa Kennell
• Clinical Reviewers Ronald Weintraub, M.D.
• Wolf Sapirstein, M.D.
• Paul Chandeysson, M.D.
• Statistics Heng Li
• Engineering Deanna Busick
• Vivianne Holt
• Terry Woods
• Animal data/remainder Lisa Kennell

5
Device Description/Sizes

• PRECISE 5.5F OTW
• 135 cm long
• 0.018 guidewire
• Sizes 5, 6, 7, and 8 mm x 20, 30, or 40 mm
  straight and tapered 8-6 x 30 mm
• PRECISE 6.0F OTW
• 135 cm long
• 0.018 guidewire
• Sizes 9 and 10mm x 20, 30, and 40 cm straight and
  9-7 and 10-7 x 30 cm tapered

6
Device Description/Sizes continued

• PRECISE RX 5.5 and 6.0F
• 135 cm long
• 0.014 guidewire
• Same sizes as OTW but no tapered configurations
• RX not under consideration today

7
Device Description/Sizes continued

• ANGIOGUARD XP OTW
• 300 or 180 cm long
• 0.014 guidewire
• Filter diameters 4, 5, 6, 7, and 8 mm
• For vessel diameters 3 - lt/ 7.5
• ANGIOGARD XP RX
• 180 cm long
• 0.014 guidewire
• Filter diameters 4, 5, 6, 7, and 8 mm
• RX not under consideration today

8
Recent Developments

• Cordis submitted unsolicited amendment to PMA
  4/5/04
• problem with air entrained in the RX version when
  used off label in carotid and other
  non-approved indications
• Has resulted in adverse events from air embolism
• Event rate estimated at 0.14 for all procedures
  (carotid and others)

9
Recent Developments continued

• Performed simulated bench testing to determine
  root cause, but this testing not optimal
• Corrective actions
• stipulate use of larger guiding
  catheters/introducer sheaths to ? potential for
  air entrapment
• Modify IFU prep procedure
• Did not perform animal testing to verify if
  corrective action corrected problem

10
Precise/AngioGuard Indication

• The proposed indication for use for the system
  is
• The Cordis PRECISE Nitinol Stent System used in
  conjunction with the ANGIOGUARD XP Emboli Capture
  Guidewire is indicated for use in the treatment
  of carotid artery disease in high-risk patients.
  High-risk is defined as patients with
  neurological symptoms (one or more TIA s or one
  or more completed strokes) AND gt/ 50
  atherosclerotic stenosis of the common or
  internal carotid artery by ultrasound or
  angiogram
• OR
• Patients without neurological symptoms AND gt/
  80 atherosclerotic stenosis of the common or
  internal carotid artery by ultrasound or
  angiogram.
• Symptomatic or asymptomatic patients must also
  have one or more condition(s) that place them at
  high-risk for carotid endarterectomy.

11
Regulatory History

• IDE submitted in 1998
• Many design changes to devices
• Most significant were addition of Angioguard,
  lowering profile and rapid exchange configuration
  

12
Regulatory History Continued

• Sponsor terminated randomized study early.
  Sponsor states reasons for early termination
  being
• too many competing studies,
• physicians reluctant to randomize,
• Surgeons unwilling to refer patients

13
Regulatory History Continued

• Competing studies involved Cordiss own devices
• Competing studies were single-investigator
  sponsored studies authorized by Cordis, but not
  followed by Cordis
• Cordis supplied each investigator copy of
  feasibility (non-randomized) protocol, CRFs,
  consent, and letter of authorization to
  facilitate opening their own IDE

14
Regulatory History Continued

• Most single investigator-sponsors followed Cordis
  protocol, with little deviation, however Cordis
  not certain of exact protocol amendments
• Investigator-sponsored studies each approved for
  50-100 subjects

15
Regulatory History Continued

• No contractual relationship between Cordis and
  investigator-sponsors for sharing data
• HOWEVER
• PMA regulation stipulates that sponsor must
  report all data that they are aware of or should
  be aware of, so Cordis made effort to supplement
  PMA with this data

16
Regulatory History Continued

• Cordis did not fund, sponsor or monitor these
  studies
• 34 sites contributed data in PMA, 2 did not
  participate
• Single investigator sites following to 12 months,
  but only 30 day data in PMA

17
Non-Clinical Study Summary

• Sponsor conducted simulated use, fatigue and
  device specification and integrity tests on the
  bench and in animals for both the stent and the
  embolic protection device, with each iteration of
  the devices.
• RX iteration has only pre-clinical bench and
  animal testing FDA agreed to allow clinical use
  without clinical data since working end not
  changed
• Still working with sponsor on RX validation only
  OTW under consideration today

18
Non-Clinical Study Summary - continued

• Engineering reviews complete and satisfactory
• Biocompatibility review complete and satisfactory
• Sterilization review ongoing but do not
  anticipate issues

19
Other non-Clinical Issues

• FDA issued warning letter on April 1, 2004
• Cited non conformance with the Current Good
  Manufacturing Practice (CGMP) requirements
• FDA is concerned with the breadth and scope of
  the violations
• symptomatic of serious underlying problems in
  Cordis s manufacturing and quality systems
• FDA sought Corporate Corrective and Preventive
  Action plan which ties all facilities

20
Statistical Summary

• Heng Li, FDA Statistician

21
Statistical Issues

• SAPPHIRE Randomized Trial
• SAPPHIRE Stent Registry
• Propensity Score Analysis
• Conclusions

22
Randomized TrialStudy Protocol Adherence

• The randomized clinical study was originally
  designed as a group sequential clinical trial
  using the sequential triangular test.
• Interim analyses were scheduled every 100
  patients.
• The expected sample size was 600 to 900, with a
  maximum sample size of 2400.

23
Randomized TrialStudy Protocol Adherence

• The randomized study was not conducted according
  to the original group sequential protocol
• An alternative protocol seems to have never been
  developed
• FDA was not informed of any change in protocol
  prior to PMA submission

24
Randomized TrialStatistical Inference

• Statistical inferences (e.g., declaring
  non-inferiority) for designed studies should be
  made according to the study design.
• Since the initial protocol was neither followed
  nor replaced by an alternative one, a nominal
  protocol needs to be referred to when statistical
  inference is conducted.

25
Randomized TrialStatistical Inference

• The nominal protocol used by the sponsor for this
  PMA submission is a fixed sample size design with
  the planned sample size equal to the sample size
  at which the trial was discontinued.
• This is probably the most favorable choice for
  declaring non-inferiority.

26
Randomized trialPre-specified analysis
27
Randomized trialAnalysis at 334 patients
28
Stent RegistryPre-specified analysis

• OPC16.94
• Observed 360 day MAE rate 15.76
• 95 CI (12.36, 19.68)
• OPC not met

29
Stent Registrycomparison with randomized CEA

• The sponsor carried out a comparison of the stent
  registry versus the CEA arm of the randomized
  studies
• Since by definition the patient characteristics
  of the two groups are different, a simple
  comparison is not appropriate
• The sponsor used the propensity score method to
  compare the two groups

30
Propensity Score Method

• A class of statistical procedures that can help
  evaluate difference in treatment effect when the
  treatment groups are not necessarily comparable
  (e.g., treatments are not randomly assigned).
• The propensity score methodology reduces bias by
  balancing (on average) a set of chosen covariates

31
Propensity Score Method

• Propensity score method has the advantage of
  typically being able to simultaneously balance a
  large number of covariates.

32
Stent RegistryComparison with randomized CEA

• It is not clear whether the analysis that the
  sponsor performed has taken full advantage of the
  potential to balance all the clinically relevant
  covariates
• Key covariates such as baseline demographics and
  angiographic data were not considered

33
Conclusions

• Randomized trial
• Original group sequential protocol not followed
• A second protocol not developed
• No evidence of crossing non-inferiority boundary
  at termination of study

34
Randomized trialAnalysis at 334 patients
35
Conclusions continued

• Randomized trial
• Original group sequential protocol not followed
• A second protocol not developed
• No evidence of crossing non-inferiority boundary
  at termination of study
• Registry
• Fails to meet original OPC
• Propensity score analysis not adequate

36
Clinical Review

• Dr. Ronald Weintraub, FDA Consultant

37
Overview

• Randomized SAPPHIRE trial
• Stent Registry cohort
• Subgroup results
• Effectiveness results
• Historical surgical trials

38
SAPPHIRE Trial
39
Inclusion Criteria

• Symptomatic patients (ipsilateral TIA or
  completed stroke) with gt/ 50 stenosis
• OR
• Asymptomatic patients with gt/ 80 stenosis
• AND
• A co-morbid condition indicating higher risk for
  CAE

40
Inclusion Criteria ContinuedCo-Morbid Risks

• Significant Cardiac disease
• Severe pulmonary disease
• Contralateral carotid occlusion
• Contralateral laryngeal palsy
• Post radiation treatment
• Previous CEA/stent
• Other anatomic risk factors

41
Exclusion Criteria

• Stroke-in-progress, or stroke within 48 hours
• Intracranial mass
• stent in target vessel
• Intraluminal thrombus visible
• Total occlusion of target vessel site
• Known PVD, supra-aortic or ICA tortuosity
  precluding interventional approach

42
Exclusion Criteria Continued

• Intracranial aneurysm gt9mm
• Lesion requires gt2 stents
• Stent in contralateral vessel lt30 days
• Subclavian ostial lesion (added later)
• Percutaneous interventions planned lt30 days after
  index procedure (initially one year)
• Staged procedure for bilateral disease lt30 past
  index procedure

43
Protocol overview

• Sponsor contracted out some aspects, or had
  independent oversight
• Clinical Events Committed to adjudicate adverse
  events
• Core lab for angiographic analyses
• Core lab for ultrasound analyses
• Core lab for analysis of filter baskets
• Data analysis contracted to Harvard Clinical
  Research Institute

44
Study Endpoints

• Primary Endpoints
• Composite major adverse events (death, any
  stroke, and/or MI at 30-days)
• Composite MAE as above, plus death and/or
  ipsilateral stroke 31 days to 12 months

45
Study Endpoints Continued

• Secondary Endpoints
• Successful stent deployment
• Successful filter deployment and retrieval
• lt30 residual stenosis by angiography
  post-dilatation
• Access site complications
• Surgical site complications
• Patency (gt50 restenosis by US at 48 hours, 6,
  12, 24, and 36 months

46
Study Endpoints Continued

• Independent neurological assessments at 24 hours,
  30 days, 6, 12, 24, 36 months
• 30-day and 6 month evaluation for stroke
• MAE composite at 6, 12, 24, and 36 months (death
  and ipsilateral stroke)
• Safety assessment of Angioguard XP
• Presence of trapped material in filter
• Laboratory analysis of trapped material

47
Enrollment Distribution

• Most patients were enrolled at 5 sites
• Cleveland Clinic 93 subjects (27.8)
• Prairie Cardiovascular 43 subjects (12.9)
• St. Lukes Medical Center (WI) 30 subjects (9.0)
• St. Lukes Medical Towers (AZ) 29 subjects (8.7)
• Midwest Card. Res. Found. 19 subjects (5.7)
• Total enrolled in these sites 214/334 (64)

48
MAE Distribution by Site
49
Randomized Pivotal Trial 30 Day Major Adverse
Event Rates
50
Randomized Pivotal Trial 360 Day Major Adverse
Event Rates
51
Randomized Pivotal Trial 720 Day Major Adverse
Event Rates
52
30-day MAE Rates with and without MI included
53
MI Details
54
Stent Registry Cohort
55
Stent Registry Cohort

• Patients entered into registry cohort prior to
  randomization
• Total of 406 entered
• Reasons for entry were given for 196/406 (48.3).
  These included
• Prior CEA 62/406 (15.3)
• Previous radiation 27/406 (6.7)
• High Cervical Lesion 20/406 (4.9)
• CAD 20/406 (4.9)
• Age gt80 15/406 (3.7)
• COPD 11/406 (2.7)
• Multiple co-morbidities 9/406 (2.2)
• Unknown 210/406 (51.8)
• Asymptomatic 281/406 (69.2)

56
Stent Registry CohortMajor Adverse Event Rates
57
Statistical Inference (Stent Registry)

• OPC12.94
• d4
• Observed 360 day MAE rate 15.76
• 95 Confidence Interval for 360 day MAE
• (12.36, 19.68)
• p-value for H0 360 day MAEOPC d
• p0.29
• Pre-specified criterion of non-inferiority not
  met.
• It is not clear that the propensity score method
  has been thoroughly explored questions remain
  about adequacy of analysis.

58
Subgroup Results
59
Symptomatic Subgroup Analysis 30-Day Adverse
Event Rates
60
Symptomatic Subgroup Analysis360 Day Adverse
Event Rates
61
Asymptomatic Subgroup Analysis 30-Day Adverse
Event Rates
62
Asymptomatic Subgroup Analysis360 Day Adverse
Event Rates
63
MAE in Significant SubgroupsOther than
Symptomatic/Asymptomatic
64
MAE in Significant SubgroupsOther than
Symptomatic/Asymptomatic Cont.
65
Effectiveness Results
66
Effectiveness ResultsSecondary Endpoints
67
Effectiveness ResultsSecondary Endpoints
Continued
68
Historical Surgical Studies
69
Historical Study Background

• VA Cooperative (asymptomatic) Study
• ACAS (asymptomatic) study
• ECST (European) symptomatic study
• NASCET (North American) symptomatic study

70
Enrollment CriteriaHistorical Comparison

• NASCET/ACAS Exclusions
• gt79 yrs
• Cardiac source emboli
• Mental incapacity
• MI lt6mo.
• Kidney/liver failure
• Lung failure
• Ipsilateral CEA
• Contralateral CEA lt4mo.
• Unstable angina
• Intracranial pathology
• Life expectancy lt5 yr
• Ipsilateral CVA
• Intolerance to anticoagulants

• SAPPHIRE Inclusions
• gt79 yrs
• Cardiac source emboli not addressed
• Mental capacity not addressed
• MI gt24 hrs.
• Kidney/liver failure excluded
• Lung failure included
• Ipsilateral stent excluded
• Contralateral stent lt30 day
• Unstable angina
• Intracranial disease excluded
• Life expectancy lt1yr
• Ipsilateral CVA excluded
• Intolerance to antiplatelet excluded

71
VA Cooperative Study

• Asymptomatic, with gt/50 Stenosis
• MAE (30 days) 4.7
• Risk of Ipsilateral stroke, TIA, Monocular
  blindness markedly reduced (8 CEA v. 20 Med)
• BUT
• Long-term MAE (mean 4 years range to 8 years)
  were equally high (41 CEA v. 44 Med)
• High attrition from Stroke and Cardiac diseases
• Cardiac Mortality 20 in both CEA and Med RX
  Groups

72
ACAS

• Asymptomatic, with gt/60 stenosis
• MAE (30 days) 1.5, plus 1.2 TIA or stroke from
  Angiography
• Over 5 years, 50 reduction in risk of stroke
  (5 CEA 11 Med Rx)
• In asymptomatic patients with Moderate/Severe
  stenosis, CEA is indicated if it can be performed
  with a perioperative Stroke/Death rate lt 3

73
ECST (Europe)

• Symptomatic, with pts stratified by degree of
  stenosis
• MAE (30 days) 4.8, but long-term MAE the same
  (37) for both CEA and Med
• In pts with gt/60 stenosis, risk of MAE at 3 yrs
  was 15 CEA and 26 Med
• 11 absolute (58 relative) benefit
• Pts with lt 60 stenosis do not benefit
• Operative risk does not increase with degree of
  stenosis

74
NASCET (N American)Entire Cohort

• Symptomatic, with pts stratified by degree of
  stenosis
• MAE (30 days), entire cohort 6.7
• MAE (30 days), 70-99 stenosis 5.8
• At 2 years follow-up, study discontinued for pts
  with 70-99 stenosis

75
BENEFITS AND RISKS OF CEA (RCTs)
Modified from Chassin MR, NEJM 1998 3391468
76
Historical RCT Conclusions (1)Symptomatic
Patients

• 70-99 stenosis CEA very effective
• (gt50 reduction in risk of stroke and
• any death at 2 years)
• Risk reduction varies with stenosis
• 50-69 stenosis success less certain
• 23 operations to prevent each
• severe ipsilateral stroke at 5 years
• Each 2 increase in 30 day MAE
• reduces 5-year benefit by 20

77
Historic RCT Conclusions (2)Asymptomatic Patients

• gt 60 stenosis CEA Very Effective
• (50 reduction in risk of ipsilateral stroke
  or peri-op stroke or death)
• IF
• Procedure can be performed with 30-day MAE lt 3

78
CAUTIONARY NOTES FOR ASYMPTOMATIC PATIENTS(ALL
RCTs and AHA)

• Risk of ipsilateral stroke is low (1-3/year) in
  patients treated medically
• As many as 45 of strokes in such patients
  (NASCET) were found to be of lacunar or
  cardioembolic etiology
• Older patients and those with comorbidities
  should be carefully evaluated before CEA
• Asymptomatic patients with an expected lifespan lt
  5 years are not candidates for CEA

79
30-Day Adverse Event Rates-Symptomatic
80
360-Day Major Adverse Event Rates Symptomatic
81
30-Day Major Adverse Event Rates-Asymptomatic
82
360-Day Major Adverse Event Rates Asymptomatic
83
Study Limitations

• The pre-specified enrollment plan and study
  analysis was not carried to completion in the
  SAPPHIRE randomized study
• Resulted in a smaller size study with small
  sample sizes in important subsets of carotid
  populations

84
Conclusions

• Randomized study suggests non-inferiority of
  stent to CEA
• Registry cohort failed to meet the OPC
• Comparability of the registry to the control CEA
  patients has not been optimally defined/conducted