Post on 03-Jan-2021
FDA Overview of Molecular Diagnostics and the Critical Path Initiative
Steven Gutman, MD, MBADirector, Office of In Vitro
Diagnostics
Medical Device Amendments of 1976
General controlsRegistration and listingGood manufacturing practicesPost market reporting
Premarket Review
510(k)sPMAAdministrative differencesCommon scientific base
Analytical Performance
AccuracyPrecisionAnalytical specificityAnalytical sensitivity
Clinical Performance
Signals can be turned in to clinical actionDiagnostic sensitivityDiagnostic specificityPredictive value or positive or negative results
Labeling -- 809.10(b)
Intended usePerformance characteristicsLimitations
Life Cycle
Analytical performanceFeasibility *Clinical performance FDA approval – market access---------------------Real world use
The Real World of Regulation
FDA CMS (CLIA)CMS and others (third party pay)
Critical Path Initiative
Biomarkers appear in two contextsDiagnosisDrug discoveryConcept of personalized medicine grounded in genomic map but not bounded by this
Critical Path Initiative
InfrastructureOpportunities listPilot programs
Bad News
Cutting edge new technology --multiplex, bioinformatics, nanotechnologyPaucity of material or method standardsBiological and clinical nuancesFinancial uncertainties
Bad News
Scientific limitations are clear Spectrum biasVerification biasImpact of missing data pointsDiscrepancy
Good News
Regulatory trail is well litLiteratureStandardsGuidances
Principle Road Maps
STARD InitiativeReMARK Initiative
Growing Literature on Co-Development
Simon and Wang, 2006Pennello and Vishnuvajjala, 2005Sargent et al, 2005Pustzai and Hess, 2004
Growing FDA Guidance
Voluntary Genomic Data Submissions guidanceConcept paper on co-developmentStatistical guidance on IVD labelingGuidance on pharmacogenetic and heritable genetic testsBayesian statistics
Good News
Broad menu of regulatory toolsMandate to be least burdensomeNew scientific resources -- MDUFMANew regulatory programs -- FDA data template, critical path
Flexible Regulatory Tools
Pre-IDEExpedited reviewsReal time reviewsDe novo classification
Review Successes
Cystic Fibrosis test – 109 daysAvian flu – 14 daysUGT1A1 – 9 days
Co-Development
If predictive diagnostic determines drug choiceSafety and effectiveness of drug becomes hostage to diagnosticNeed to understand system as a whole
Predictive Marker
Identify patients by biomarker status; randomize therapy across all patientsIdentify patients by biomarker status; randomize therapy in subsetsRandomize by treatment; look back at biomarkers
Predictive Marker
H (non-response)
G (response)
F (non-response)
E(response)
Placebo
D (non-response)
C(response)
B (non-response)
A (response)
Therapy
Neg testNeg testPositive test
Positive test
Predictive Marker
F (non-response)
E(response)
Placebo
B (non-response)
A (response)
Therapy
Positive test
Positive test
FDA Mission
Promote public healthProtect public healthTension in objectives
Good Science