Wednesday, September 21, 2016

Are FDA, Medicare, and Moonshot Converging on Cancer Testing? ...Medicare?!

Putting together precision medicine policy milestones of the past year, it looks like Medicare is becoming closely aligned with goals of the FDA, the Precision Medicine Initiative, and Cancer Moonshot.

PMI/Moonshot
One of the cardinal goals of the Precision Medicine Initiative and especially the Cancer Moonshot headed by Joe Biden are to break down information silos and creative public databases on cancer genomics.
In June 2016, Foundation Medicine announced it was supporting Cancer Moonshot by "contributed 18,000 genomic profiles from our FoundationCORE knowledgebase to the NCI Genomic Data Commons Portal (GDC)" (here).  
FDA
The FDA is producing groundbreaking guidance documents about the FDA validation of genomic tests based on public curated databases.  A harbinger was the approval of a genomic cystic fibrosis test based on a public variant database.  This approach is full steam forward; see a recent guidance document from the FDA on the calibration and regulatory use of variant databases, here.

Medicare
Medicare has a roughly 25-state unified policy system called MolDX in which local coverage determinations most often originate in the "Palmetto" Medicare contractor jurisdiction and propagate to other multi state jurisdictions.  In September 2016, the Palmetto MAC proposed a new local coverage policy for genomic testing in lung cancer which mandates extensive public data deposition in elaborately defined public access variant databases.    The local decision specifically cites the goals of Cancer Moonshot as being concordant with Medicare's view of eligible open access public genomic databases.  Specifications of an eligible database are listed below the break.  For a deeper dive, see here.

____

Proposed requirements for registry data deposition for genomics labs providing lung cancer gene panel testing (Medicare document DL36143):

Collect CGP test and patient specific information in the MolDX approved registry that meets the following characteristics:

o   National in scope and open to any lab, (commercial or academic), and any provider location (academic, community);

o   Independent of the participating laboratory (commercial or academic);

o   Governed by a well-designed protocol listed on clinicaltrials.gov with national cross-institution leadership, patient consent, Institutional Review Board (IRB) approval, end points, and regular reporting;

o   Requires and verifies that CGP testing is essentially equivalent to MolDX Analytical Performance Specifications for Comprehensive Genomic Profiling AND has demonstrable plans to maintain compliance to both MolDX and other published standards;

o   Will collect detailed genomic information as detailed by the registy– including raw (FASTQ or BAM) data and variant call data in connection with clinical outcomes and report these to the registry in a timely fashion;

o   Will compare CGP identified mutations in the EGFR, ROS1 and ALK regions to companion diagnostic tools (where exist) on a subset of patients to determine concordance;

o   Organization overseeing the registry or essential partners to that organization cannot have a history of data siloing (e.g. not sharing data with competitors) or history of requiring physician-groups purchase or lease any propriety software;

o   Registry organization has to have shown a strong commitment and effort to work with national organizations committed to data sharing (i.e. Genetics Data Commons, NCI, Vice President Biden’s Moonshot initiative, etc.);

o   Registry will allow open, non-commercial research access to the database (with appropriate curation), and will allow equal access to commercial groups;

o   Non-profit registry organization is preferred;

·        Registry to report to Palmetto GBA every six months the following:
o   Number of patients enrolled in registry

o   Biomarker prevalence in registry patients

o   Treatments and time to progression in patients with a given biomarker per line of therapy for at least 2-3 lines of therapy

o   Overall survival of patients by biomarker status and treatment profile

o   Concordance analysis of biomarker testing results between CGP and a FDA approved companion diagnostic test (where one exists for a given biomarker)

·        Registry will assure publication of test results and clinical findings on a regular basis