Download pdf

On November 6, 2006, Genetic Alliance hosted a webinar to discuss the FDA draft guidances on genetic testing. Representatives of government, research, clinical care, industry, and laboratories, including Mara Aspinall, Kathy Hudson, David Mongillo, Mike Watson, and Janet Woodcock, participated. To review the major points each panelist presented, click here. To read the transcript of questions and answers presented during the webinar, click here.

While many questions were addressed during the webinar, time did not permit panelists to address all questions. Questions generated during the webinar are presented below with answers supplied by panelists after the event.

Questions asked during Janet Woodcock’s presentation

Upon review of these questions, the FDA has determined it cannot answer these questions since the Guidances are still pending. They recommend you submit these types of comments to the docket [docket #2006D-0347 IVDMIA, #2006D-0336 ASR]. Specific regulatory questions may be addressed directly to the Agency and they will respond on a case-by-case basis.

General

• With the anticipated increase in submissions, does the FDA have sufficient personnel and resources to review the applications in a timely manner?
• Would the agency be willing to extend the comment period to 120 days, so that stakeholders—including Congress—can have adequate time to respond?
• Will the 30-day extension also apply to the draft ASR guidance document, or just the IVDMIA draft guidance?
• Will FDA allow a grace period or allow some tests to be grandfathered before requirements are enforced? If so, what is the anticipated timeline?
• Has the FDA posted a list of the handful of genetic lab tests that are already FDA-approved? By the same token, is there a public list of lab-developed tests other than those on the GeneTest website, where it is difficult to find a list?
• How can healthcare providers know which genetic tests are available and which of them are FDA-approved vs. lab-developed?
• Will public discussions on the draft guidances take place outside of Washington, DC?
• Could you give several real examples of ASRs and IVDMIAs?

IVDMIAs

• Could you please give an example of a device that uses an algorithm but is not an IVDMIA?
• If a reference laboratory in one state develops a home-brew test that meets the three inclusion criteria of the IVDMIA but the test is intended solely for patients in that state, does the FDA have jurisdiction?
• For therapy-related testing, i.e., the drug safety or efficacy predictor test, if the test falls into the IVDMIA category, what is the regulatory pathway in terms of review process? Please give two scenarios to illustrate.
• Laboratory pathologists frequently provide clinical consultations using their professional knowledge and training to evaluate multiple test results and patient clinical information to provide a clinical interpretation. Are these clinical consultative activities considered an IVDMIA? If not, why?
• Are software packages that use published equations based on empirical data to combine multiple test results into a single patient risk considered multivariate tests? For example, the optimal dosing of warfarin using genotyping.
• This IVDMIA guidance and how it would be implemented seems very undefined. The patient community feels that it is not good for a regulatory agency to move forward in any way without clear guidelines. Won’t this guidance delay innovation and tests being delivered to the clinic?
• Are there medical groups calling for the FDA to regulate IVDMIAs because of specific concerns or real reported harm to patients? If so, please give examples?
• How will the FDA resolve the overlaps and conflicts with enforcing the IVDMIA guidance with current CLIA requirements? Won’t labs be subject to both regulations? If not, how will the FDA clarify what they will require and what process the will use?
• Does the FDA want to be the filter between the practice of medicine and these new IVDMIA tests? Is this not a new role for the agency?
• Is it possible to supply software to a laboratory that can be used to compute an index if the software must be configured by the laboratory to create that index? For example, if the index can be a weighted sum, the software only provides the means for the user to specify the analytes used and the weight for each analyte, and the software then computes and reports the weighted sum. What if the software that can be configured to compute an index only works for an instrument from one manufacturer? Can this still be supplied to the laboratory since both the instrument and software can be configured to serve many assays that are completely different in their intended use?
• Does the need for the test-developer to interpret the results impede access to the results for immediate use and medical decision-making?
• Is it the intent of the FDA to declare that all laboratory-developed tests are subject to FDA oversight and regulatory discretion, and if so, are all non-cleared laboratory-developed tests illegal?

ASRs

• Could you clarify what you mean by single moiety in the ASR definition above and beyond the guidance?
• When multiple ASRs, intended to be used in diagnosing only a specific single disease or condition, appear to be in compliance with ASR rule, is there anything in the rule requiring that this diagnosis must be accomplished by a single ASR?
• If biomarkers have biological association with a particular disease, can they be used in a combined format to diagnose disease? If so, what regulation applies?

Questions during Mara Aspinall’s Presentation

• Should the same regulatory policies apply in the “era of diagnostics” with treating these tests to a lengthy pre-market approval when the market forces and parameters are very different? It seems as if the impact on test producers in terms of cost and time have not been fully reviewed. If someone is developing a test that only has a market of 5,000 people or so, but it is critical to those 5,000 people, should different processes be developed, in your opinion?

Questions during Mike Watson’s Presentation

• Is a normal/common DNA test that uses two primers to produce an amplicon with either wild type or mutation sequence and separate probes to differentiate between the wild type and mutant type sequences, by definition, a medical device?
  

  Yes, the FDA specifically named these in the ASR rules as being “critical components” of a genetic test that is developed by a laboratory. They are also an important part of any class 2 or 3 device for a genetic test that requires primers, though currently there are relatively few to have gone for this higher level of FDA approval.

• Are tests that use empirical data to relate sequence alterations and disease outcome considered medical devices and not ASRs, under the new guidance?
  
  

  This is probably a question for the FDA, but my sense is that they consider any laboratory-developed test to be under their regulatory purview, even though they haven’t exercised that power for most tests.

  
  
• Is there a role for the federal government in examining the clinical validity of genetic tests? If so, which agency? If not, how can patients and healthcare providers be reassured that the tests give clinically-meaningful results?
  
  

  The federal government already fills several roles. Any time a genetic test is developed by a manufacturer, the FDA evaluates the product to ensure that it does what it is supposed to do. The clinical validity has to be shown, but that doesn’t necessarily mean that the test will be accepted by payers as valid from their perspective. All laboratories are licensed under the CLIA rules as well. These are heavy on analytical issues but less robust for clinical issues. However, CLIA could be improved to address this area of laboratory practice.

  Our focus is on tests used in healthcare. However, a number of tests related to “lifestyle” that are marketed as being useful for selecting an appropriate diet of skin cream. It is these tests that have been the source of much concern about genetic tests in the media recently. Alternative federal and state agencies are in place to address some of these problems. When the seller overstates or misrepresents the validity of these or other tests, the Federal Trade Commission can step in to protect consumers. States Attorney’s General are also charged with consumer protection and can step in when the public is being sold products with unsubstantiated claims.

  Another issue arises in this context. It is that all tests have to go through a period of clinical investigation. Once a relationship between a gene and a disease is established, there is period during which testing is done so that sufficient data is acquired to allow clinical and analytical validity to be addressed. For genetic tests, this phase typically addresses the diagnostic uses in patients and their families. Given that it is relatively straightforward to demonstrate a diagnostic utility, the real issues arise when things go off label. As you may know, we are currently in the neighborhood of about 40+% of all prescriptions for FDA-approved drugs being for uses other than that for which it was approved.

  
  
• Are researchers developing paired treatments with genetic tests?
  

  Yes. These can arise through several routes. For instance, in newborn screening, the inclusion of a screening test is dependant on the availability of a treatment. When the treatment is a new pharmacologic agent (there are a number of these already approved or in the clinical trials pipeline for lysosomal and other disorders), it is likely that the confirmatory test that establishes the diagnosis will lead to that specific drug being used to treat the disease. The FDA is also relabeling some of its drug labels to indicate that a genetic test could be used to predict whether a particular drug is best for that patient and may also provide information about the preferred starting dosage. I would expect this to evolve to a requirement that the genetic testing be done for drugs that may have serious side-effects to those with genetic variations that affect the metabolism of the drug. CYP 2C9 and VKORC1 testing for Warfarin may be among the first of these since it already is useful in determining the best starting dose that could minimize the risks of over- or under-coagulation. Since this is a $ half-billion marketplace, the final decisions will await the results of ongoing randomized control trials.