Medical Device Clinical Studies and Protocol Design

1
Medical Device Clinical Studies andProtocol
Design IVT Medical Device Conference San
Francisco August 17, 2006
Michael A. Swit, Esq. Vice President, Life
Sciences
2
Presentation Overview

• Standards of Approval What the Protocol Targets
• Key Considerations in Designing Clinical Studies
• Practical Lessons in Clinical Trial Design
  Execution

3
Approval Standard PMA

• PMA Approval Standard reasonable assurance
  that the device is safe and effective under the
  conditions of use prescribed, recommended, or
  suggested in the labeling
• Valid Scientific Evidence FDA relies only
  on to determine reasonable assurance 21 CFR
  860.7(c)(1)

4
Approval Standard PMA

• Valid scientific evidence evidence from
  well-controlled investigations, partially
  controlled studies, studies and objective trials
  without matched controls, well-documented case
  histories conducted by qualified experts, and
  reports of significant human experience with a
  marketed device, from which it can be fairly and
  responsibly concluded by qualified experts that
  there is a reasonable assurance of the safety and
  effectiveness of a device under its conditions of
  use.
• source 21 CFR 860.7(c)(2).

5
Clearance Standard 510(k)

• "substantially equivalent" or "substantial
  equivalence" means, with respect to a device
  being compared to a predicate device, that the
  device has the same intended use as the predicate
  device and that the Secretary by order has found
  that the device
• (i) has the same technological characteristics as
  the predicate device, or
• (ii)(I) has different technological
  characteristics and the information submitted
  that the device is substantially equivalent to
  the predicate device contains information,
  including appropriate clinical or scientific data
  if deemed necessary by the Secretary or a person
  accredited under section 523, that demonstrates
  that the device is as safe and effective as a
  legally marketed device, and (II) does not raise
  different questions of safety and effectiveness
  than the predicate device.

6
Conformity Standard for CE Marking in the EU

• Remember unlike with pharmaceuticals, there is
  no pre-market role for devices either at an EU
  centralized authority (indeed, unlike drugs,
  there is no central authority for devices) or in
  member states
• CE Marking handled via reviews by notified
  bodies
• Must meet applicable Device Directive
• Implantable
• Non-Diagnostic
• Diagnostic

7
Background 510(k) Studies

• Substantial Equivalence clinical studies are only
  required in 10 of all Class II 510(k)
  submissions
• When no reliable method is available to validate
  substantial equivalence to a predicate device
• Product-related issues
• Novel design
• New technology
• New indications for use
• Upon request by FDA

8
Background 510(k) Studies

• Superiority and economic data not required for
  FDA clearance of a 510(k) submission. . .
• . . .but these data are required to support
  reimbursement applications with CMS or private
  payers
• Data to support FDA clearance may not be the data
  needed for reimbursement marketing goals

9
Background 510(k) Studies

• What is an equivalence trial?
• a clinical trial designed to evaluate whether
  an experimental treatment E is similar to a
  control treatment S, by an appropriate definition
  of similarity

Reference W.C. Blackwelder, 2004, J. Dent. Res.
83
10
Background 510(k) Studies

• Equivalence
• Two-sided or bi-directional
• e.g. pharmaceutical bio-equivalence
• Non-inferiority
• One-sided or uni-directional
• e.g. most 510(k) clinical equivalence studies
• NOTE Equivalence and non-inferiority are similar
  but not the same thing.most 510(k)
  substantial equivalence trials are
  technically non-inferiority trials

11
Regulatory Considerations 510(k) Studies

• Remember prime focus of 510(k) is substantial
  equivalence
• So, how do you know you need a clinical study?
• Comparison to predicate technology
• Comparison to predicate intended use
• Comparison to requirements for similar devices
• Any factor raising new questions of safety or
  effectiveness that cannot be alleviated through
  bench testing points towards a clinical trial

12
Regulatory Considerations 510(k) Studies

• Communicating with FDA
• When if the path is not evident
• Recommend an informal guidance meeting
• Requires meeting request letter, preparation of
  pre-meeting package with pointed questions
  regarding strategy
• Maximizing your chances of a positive outcome
  organized, complete, concise package and well
  constructed strategy and questions for discussion

13
Regulatory Considerations 510(k) Studies

• Case Studies
• Clinical data probably required
• Non-invasive blood glucose meter
• New or significant change of clinical instrument
  software algorithms
• Thermal regulation catheter system
• Clinical data probably not required
• New LED vendor for existing pulse oximeter sensor
• Data interface for fingerstick blood glucose meter

14
Different Clinical Study Hurdles For Different
Audiences
Effectiveness
Pricing
Reimbursement Code
Safety
Efficacy
Patient
Prescriber
End-user
Payer
Regulator
Adapted from Perfetto 2001
15
Regulatory Considerations Reimbursement
Studies

• CMS -- primary focus is effectiveness
• Requires systematic evaluation of the performance
  and properties of the technology
• All available clinical and outcomes data
• Comprehensive review of relevant literature
  (published and unpublished)
• If marketed, opinions/data from leaders in the
  field on real-world use
• Analysis of competitive advantage
• Overall economic impact including costs offsets

16
Regulatory Considerations Reimbursement
Studies

• Data needs for reimbursement purposes often go
  beyond what is required to fulfill regulatory
  requirements
• To the extent possible, requisite data should be
  determined as part of the medical devices
  product development plan
• Different types of studies could be required at
  different times throughout the process
• A strategy for data collection needs to be
  started early

17
Regulatory Considerations Reimbursement
Studies

• Avoid misconceptions such as
• It is soft science, so it doesnt need much
  attention
• Nothing can be done until you are close to
  marketing
• It can be done quickly
• Piggybacking is all that is needed
• This is a drug thing biotech, genomic, and
  device products do not need to worry about this

18
Trial Design Key Issues

• Trial Goal or Objective -- clinical endpoints
• Pilot or Feasibility Study
• Evidence to support trial
• Identification Selection of Variables
• Confounding variables
• Study Population
• Control Population
• Methods of Assigning Interventions

19
Trial Design Key Issues

• Masking
• Trial Site and Investigator
• Bias
• Sample Size and Statistical considerations
• Design challenges
• Key elements of trial design

20
Trial Goal i.e., the Research Question

• Clinical Goal measured via endpoints
• Primary endpoint should be clinically relevant
  objective measurable with known precision
• Secondary endpoints
• Support marketing/reimbursement claims,
• confirm cost-effectiveness, etc.
• Comparative claims
• Safety
• Caveat -- dont lose sight of the primary focus,
  which is data to support a regulatory filing!!

21
Evidence Qualitative Hierarchy

• Evidence to support clinical trials (in order of
  quality)
• Systematic review of randomized, controlled
  trials
• Randomized, controlled trials
• Prospective studies
• Retrospective studies
• Cross-sectional surveys
• Case series
• Case reports

22
Pilot (a.k.a. Feasibility) Studies

• Done when a sponsor can not answer key questions
  that would allow them to focus a clinical trial
• Used to
• Identify possible medical claims
• Monitor potential study variables
• Test study procedures (e.g., logistics)
• Refine the device prototype itself
• Determine precision of potential response
  variables
• Refine the protocol for a future pivotal study
• FDA will expect you to do under an IDE

23
Observation Variables Identification Selection

• Two types outcome and confounding
• Outcome those that define and answer the
  research question a.k.a. endpoints should be
• Directly observable
• Objectively measured
• Relate to the biological effects of the clinical
  condition (which may need to be validated)
• Example if a device reduces a particular blood
  value, must validate that the blood value is
  clinically meaningful relative to the condition
  that will be in the devices labeling

24
Variables -- Confounding

• Confounding variables
• A factor associated with both the outcome measure
  and the variable of interest (e.g., patient)
• Example a study of blood pressure treatment
  might be confounded if there were more young
  people in one arm of the study as younger people
  simply tend to have lower blood pressure
• No study is entirely free of confounders
• Methods to reduce confounding
• Inclusion/exclusion criteria
• Randomization
• Statistical measurement methods quantitative
  vs. qualitative

25
Study Population

• Key consideration balancing homogeneity of
  population vs. heterogeneity
• Homogeneous
• Advantage more precision in study measurs
• Disadvantage may limit your intended use to a
  smaller population
• Heterogeneous
• Advantage broader label claim
• Disadvantage harder to prove

26
Control Population

• Control is either real or implicit must match
  the study group
• Types
• Concurrent assigned an alternative intervention
  (e.g., placebo or standard of care)
• Cross-over self control
• Historical less reliable requires extensive
  validation may be appropriate when a concurrent
  control might be unethical
• Passive concurrent are not under the direct
  care of the principal investigator

27
Ways to Assign Interventions

• How done is key to minimizing selection bias
• Randomization key way to decrease patients
  assigned to treatment or control in a way that
  they have an equal chance of ending up in either
• But, make sure is truly random e.g., if you
  chose every third subject to come in to clinic,
  might be impacted by external variables that are
  varying the way folks are coming into the clinic
  (e.g., seasonality)

28
Blinding or Masking

• Goal to reduce investigator bias, evaluator
  bias and placebo effect
• QOL measures are particularly subject to
  evaluator bias
• Single patient blind
• Double patient investigator blind
• Third party evaluator blind (e.g., x-ray
  reader)
• Code not broken until analysis is done
• Challenge difficult to blind when a device is
  used (as opposed to drug trials)

29
Trial Site and Investigator

• Selection is key to success because pooling of
  data is usually required due to lack of patients
• Pre-qualify that there really are patients there
• Devices Center often will regard disqualified
  subjects as failures under an intent to treat
  approach
• Thus, investigator compliance with protocol is key

30
Bias

• Bias when a characteristic of the study
  interferes with the ability to measure a variable
  accurately
• is a source of systematic error in a study it
  does not occur by chance
• may occur at each stage design conduct
  analysis
• is a common problem in reports of clinical
  experience with devices
• Types of Bias --
• Observer bias
• Selection bias
• Recall bias
• Reporting bias
• Placebo response
• Learning

31
Sample Size Statistical Considerations

• Sample Size Calculation
• The larger the effect size, the smaller the trial
• Longer trials require more subjects
• Device trials are usually not worse than
  studies (vs. equivalence superiority)
• Typical standards Power 80 p0.05
• Statistical plan
• In writing and in advance!
• Clear statistical tests that are consistent with
  the scientific questions
• Provision for post hoc analysis
• Bayesian Statistics use may allow you to cut
  of patients or use historical controls (yours or
  literature)
• Guidance of Use of Bayesian Statistics in Medical
  Device Clinical Trials, May 23, 2006 --
  www.fda.gov/cdrh/osb/guidance/1601.pdf

32
Practical Considerations

• Cost of trial vs. market potential
• Blinding, randomization often difficult or
  impossible
• Investigator/user skill variability
• Double blinding is often not possible
• Large trials often not feasible
• Pre-clinical data may not predict human
  experience or failure modes

33
Practical Considerations -- Execution

• Regulatory compliance
• NSR vs. SR study (SR study requires IDE)
• Good Clinical Practices
• e.g., Informed Consent, IRB approval, clinical
  protocol, clinical trial and data monitoring
• If not done right, can invalidate data at site
• Monitoring (BIMO) issues
• Data analysis according to plan

34
Practical Considerations -- Data Analysis

• Key steps
• Complete enrollment
• Audit data
• Database lock
• Primary statistical analysis per plan
• Post hoc analysis
• Prepare formal report (either internal or for
  FDA)
• Prepare manuscript for publication

35
Practical Considerations -- Data Analysis

• Methods to Reduce Bias
• Weight of the evidence
• Consistency
• Plausibility
• Temporality
• Mechanism of Action
• Magnitude of effect
• Methods to Reduce Confounding
• Data stratification separate data by
  confounders relies on clinical judgment,
  information and suspicion
• Statistical modeling use correlation and
  regression methods, often complex

36
Practical Considerations What Goes in the
Actual Protocol

• Background of trial previous studies on device
• Clear statement of trial goals i.e., endpoints
• Complete description of trial design
• Design type
• Data collection methods
• Control type
• Blinding parameters
• Sample size justification
• How treatment group assigned

37
Practical Considerations What Goes in the
Actual Protocol

• Complete description of study population
• Study sites
• Selection methods inclusion/exclusion criteria
• Type of patient (inpatient v. outpatient)
• Clinical and demographic characteristics of
  subjects
• Complete description of intervention
• Frequency duration of application
• Compliance measures investigator patient

38
Practical Considerations What Goes in the
Actual Protocol

• Complete description of follow-up visits
• All measures made and info to be collected
• How patient withdrawal to be handled
• How sponsor will follow up on patients health if
  they drop out
• Details on data gathering and analysis
• Data collection and validation methods
• Data Monitoring
• Statistical analysis methods
• Specific rules on how/why study can be ended
  early use of DMCs

39
Practical Considerations What Goes in the
Actual Protocol

• Full info on investigators
• CVs
• Monitoring methods
• Administration of trial, including how to adjust
  protocol
• Glossary of relevant terms
• Informed Consent

40
References

• Statistical Guidance for Clinical Trials of
  Non-Diagnostic Medical Devices, FDA, Center for
  Devices Radiological Health. www.fda.gov/cdrh/od
  e/ot476.html
• Guidelines on Medical Devices Evaluation of
  Clinical Data A Guide for Manufacturers and
  Notified Bodies, European Commission, Enterprise
  Directorate General. ec.europa.eu/enterprise/medic
  al_devices/meddev/2_7.pdf

41
Questions?

• Call, e-mail, fax or write
• Michael A. Swit, Esq.
• Vice President, Life Sciences
• THE WEINBERG GROUP INC.
• 336 North Coast Hwy. 101
• Suite C
• Encinitas, CA 92024
• Phone [phone]
• Fax [phone] or 760.454.2979 (preferred)
• Cell 760.815.4762
• D.C. Office 202.730.4123
• michael.swit_at_weinberggroup.com
• www.weinberggroup.com

42
About the speaker

• Michael A. Swit, Esq., who is Vice President,
  Life Sciences at THE WEINBERG GROUP INC., has
  extensive experience in all aspects of FDA
  regulation with a particular emphasis on drugs
  and medical device regulation. In addition to
  his private legal and consulting experience, Mr.
  Swit also served for three and a half years as
  vice president and general counsel of
  Pharmaceutical Resources, Inc. (PRI) a prominent
  generic drug company and, thus, brings an
  industry and commercial perspective to his
  representation of FDA-regulated companies. While
  at PRI from 1990 to late 1993, Mr. Swit
  spearheaded the companys defense of multiple
  grand jury investigations, other federal and
  state proceedings, and securities litigation
  stemming from the acts of prior management. Mr.
  Swit then served from 1994 to 1998 as CEO of
  Washington Business Information, Inc. (WBII) a
  premier publisher of FDA regulatory newsletters
  and other specialty information products for the
  FDA publishing company. Before joining THE
  WEINBERG GROUP, he served in the FDA Regulatory
  Law Practices at both Heller Ehrman and McKenna
  Cuneo, first in that firms D.C. office and then
  in its San Diego office. He first practiced FDA
  regulatory law with the D.C. office of Burditt
  Radzius from 1984 to 1988. Mr. Swit has taught
  and written on a wide variety of subjects
  relating to FDA law including, since 1989,
  co-directing a three-day intensive course on the
  generic drug approval process, serving on the
  Editorial Board of the Food Drug Law Journal,
  and editing a guide to the generic drug approval
  process, Getting Your Generic Drug Approved,
  published by WBII. Mr. Swit holds an A.B., magna
  cum laude, with high honors in history, in 1979,
  from Bowdoin College, and earned his law degree
  from Emory University in 1982. He is a member of
  the California, Virginia and District of Columbia
  bars.

43

• For more than twenty years, leading companies
  have depended on THE WEINBERG GROUP when their
  products are at risk. Our technical, scientific
  and regulatory experts deliver the crucial
  results that get products to market and keep them
  there.
• Washington, D.C. ? San Francisco ? Brussels