New Draft Guidance for Multiplex Tests
Elizabeth Mansfield and
Michele Schoonmaker
Office of In Vitro Diagnostic
Device Evaluation and Safety (OIVD)
[Powerpoint]
Draft Multiplex Guidance
Major Points
- Recommendations are non-binding
- OIVD has little experience reviewing multiplex tests
- OIVD hopes that industry will step forward with meaningful ideas for good
guidance
Major Points
- As with other devices, regulatory path will be determined using a risk-based
approach (not technology-based)
- FDA does not consider multiplex tests as ASRs
- Genomics and genetics have been combined in a single draft
Technical Issues in
Multiplex Test Validation
Intended Use
- FDA requires that a device have an intended use, which usually encompasses
the indications for use.
- Intended use specifies what the test measures, why, and in what population
it should be used.
- FDA generally does not recommend multiple intended uses in a single submission.
Platform Design and Manufacturing
- Should conform with applicable parts of the Quality System Regulation
- Recommend characterization of design, components, instruments/software,
methods/conditions, layout and stability, etc.
- Controls and calibrators: identity and physical location (if applicable),
value assignment, span decision points (if applicable), etc.
Test Design: Pre-analytical
- Describe collection, storage, handling processing of sample (identity, acceptance
criteria, etc.)
- Validate purification and/or amplification methods
- Describe acceptance criteria for material used in assay (e.g., 260/280,
conc., etc.)
Specific performance characteristics
Analytical studies on clinical samples (where appropriate)
- Sensitivity
- Reproducibility/precision
- Cut-off, ref. range, decision point
- Assay range
- Effect of excess/limiting sample
- Specificity and interfering substances
Array and data processing
- Optimization of multiple simultaneous target detection
- Sample carry-over/signal bleeding
- Computational methods
- Limiting factors, e.g., saturation level of hybridization
Instrumentation
- Instrumentation can be general purpose or part of a system
- If general purpose, should be characterized and have specifications
- If part of a system, will be reviewed
- Describe calibration of and uncertainties introduced by instrument
Regulatory Strategies
510(k): Class II devices
Compare to:
- A commercially available predicate device (percent agreement)
- A reference method or "gold standard" (Sensitivity/specificity)
- Standard is substantial equivalence
PMA: Class III devices
- Comparison to clinical diagnosis
- Define "clinical truth" first
- Sample adequate specimens/populations
- Determine ref. ranges if appropriate
- May verify with second detection system (e.g. RT-PCR, other appropriate
system)
- Standard is "safe and effective"
De novo 510(k)
- A clearance process for certain moderate-risk novel devices that have no
predicate
- Compare to reference method or clinical diagnosis
- Standard is "safe and effective"
Pre-IDE (protocol review)
- IDE not required
- Sponsor describes proposed intended use and supporting studies
- Interactive process to provide feedback prior to initiating studies.
- Non-binding on either party
- Recommended for novel devices or uses
Clinical Effectiveness
- New markers, mutations, patterns should meet FDA standard for effectiveness
for clinical use (see 21
CFR 860.7)
- Established markers may refer to clinical literature to support the effectiveness
of the marker for clinical use.
Use of Literature
For some markers, mutations or patterns, a sufficient literature base may
exist to support clinical validity.
- Provide summary of available information pertinent to device.
- Should use same technology as "new" test and similar patient population.
Problem areas
Study Design
- How to establish appropriate sample numbers, identities, etc., esp. when
available samples are rare
- Archived vs. prospective sampling?
- Multiple intended uses for a single test?
Statistical Methods
- Depending on type of test, statistical methods may be straight-forward,
complex and/or novel. FDA is uncertain what types of methods may be presented.
- Describe statistical methods used for calculations, including measures of
precision, confidence intervals
Quality Control
- What will be the measure of quality control?
- Will there be universal standards/controls for arrays?
- Importance of controls in inter- and intra-assay reproducibility at manufacturer
and user level
Regulatory Environment
- Least Burdensome
- TPLC principles
- Transparency (truth in labeling)
Your role
- Review draft guidance carefully
- Make comments to the docket
- On the guidance
- On related issues
- Unvalidated features on clinical diagnostic
- General vs. specific claims
- Combination devices (CDRH and CDER or CBER)
- Understand QSR (formerly GMP)
Future Directions
- Specific guidance is probably needed for QSR/GMP for microarray
and multiplex devices—input from interested parties would
be valuable
|
