IVD Roundtable Notes

December 8, 2011

John Kahan of Hogan & Hartson gave an introductory speech.

He pointed out that his firm represents over 700 device companies, work with every different division/office in CDRH, and that in his opinion, CDRH (including OIVD) has been behaving generally different from past years. Policies at CDRH have been changing, and the Center is getting more unpredictable and even arbitrary. There is a lot of new staff, and they are often not managed well. There has been a ramping up of data requirements. The investment community is running scared. CDRH is aware of the problems, CDRH has produced a lot of paper recently, but there has been little effect on time to clearances, or on predictability of data requirements. CDRH-FDA is generally different from past years.

As an example, Mr. Kahan discussed the notice to industry SOP, published this summer.  FDA announced plans to give industry notice of significant changes in data requirements before they submit a filing. However, this has not helped time to market. The percent of NSE 510(k)’s has gone up from 1% in 2000 to 5% in 2011, and this does not count the number of submissions simply withdrawn and re-submitted. Including these the number would be 8 or 9 percent, more if you counted in extra review cycles, many additional requests for information (which may mostly be legitimate).

There has been a lot of change in review staffs, CDRH turnover is double that of CDER.  Kahan said the inexperience of review staff needs to be addressed. Certification and training is going on, but no results on time to market yet. Lots of newbies, lots of new requirements, led to a lot of angst.

There is an increased demand for PMAs for products that might have been 510(k)’s in the past. The PMA is frightening to the investment community, and some projects are not being financed because of increased regulatory requirements.

RUO/IUO guidance articulates what has always been FDAs expectation for compliance with labeling requirements. However, publication of guidance has led some manufacturers to consider stop offering certain RUOs which are used in diagnostic tests, rather than go to the effort of developing a PMA for a fairly small volume product.  Some of these tests have become medically required for some conditions.  The reagents are unavailable except from one manufacturer.

FDAs response to LDTs came in for some criticism. Mr. Kahan describes LDTs as a parallel universes colliding with IVDs. How long can we have two different systems for providing diagnostic information? This cannot last forever. Analytical validation is quite different for each approach. FDA should either down regulate IVDs or up regulate LDTs, so both are treated similarly.

Companion diagnostics are a critical part of pharmaceutical industry strategy. These are still a work in progress. There is at least one clinical trial going on now with an LDT as a companion diagnostic? Can we do this?

Device modifications: This draft guidance has caused a range of reactions with ADVAMED being on the apoplectic end. They claim applying it will increase. 510(k)’s 500%. Mr. Kahan can see both sides of the issue, but knows that FDA could not handle a 5x increase in the number of 510(k)s.

Dr. Gutierrez then discussed OIVD activities.

Dr. G. pointed out that OIVD has a professional roundtable as well as an industry round, and that while Mr. K. put a lot of stuff on the table, it is OIVDs policy to listen to everyone. The implication was that the professional group had different concerns. Dr. G. reviewed the 2012 OIVD priorities.

Most important, they will not scrap the 510(k) program. They have produced a white paper about improvement of 510(k) program which will result in greater transparency, better interaction and better engagement with industry, use of external experts to leverage scientific expertise, and a more patient-centric approach.

OIVD is stressing interactive review, and improving pre-submission meetings. They will leverage outside scientific expertise in their reviews. Now have a science council. Guidance on making benefit/risk determinations. They have put out draft guidance and are planning more. Risk/Benefit guidance is in the works. Industry need to understand how FDA assesses risk/benefit. This guidance will be a draft and we are asking for comments. Also planning IDE guidance, though not very many IVD IDEs. OIVD is very interested in developing an innovation pathway. OIVD has interacted with NCI, other partners, some things beyond, and is committed to fostering innovation in diagnostics.

OIVD’s goal in submission review is predictable and consistent recommendations.

To that end, they are putting a lot of SOPs in place to address “how to” review.

Dr. G. discussed the current draft on modifications, and there was some comment from the floor. Dr. G. said that companies will make modifications that FDA thinks need a 510(k). Some have resulted in major recalls. People are overreacting to the guidance. The principal floor question was, since the new guidance is a draft, should we be following the old changes guidance until this one finalizes. This led to a discussion of guidances.  E. Mansfield offered that there are two kinds, and if FDA wants immediate compliance (some immediate medical need) they will say so, otherwise a draft is a draft, so the old guidance is still the one to look to. Dr. G. Stated that the Agency will be revisiting the draft guidance and looking at comments.

Dr. G. noted that the clinical trials guidance has been fairly well received. He described it as setting new ground, although in my opinion it is simply bringing the US devices up to date with the international requirements.

There was some discussion of the Notice to industry letters. Dr. G. said examples of these were the notice to manufacturers of glucose meters, and manufacturers of tests for blood borne diseases. A question from the floor was whether the April notice about troponin was considered a notice to industry letter and under the SOP. Dr. G. said no, that was not a formal notice to industry, but simply assistance FDA provided to manufacturers who wanted to get a new claim for their troponin assay. Dr. G. Stated that the myocardial infarctions claim was not currently one that a troponin test has. Dr. G. noted that FDA is willing to look at other ways to resolve issues than the one suggested in a notice.

Discussing down classification, Dr. G. said the project needs a lot of work. At present still employing enforcement discretion for products that still need downclassifiacation. Only a very few left (10-12) . FDA is dealing with some using a draft guidance completed by spring – we will implement a triage program.

The organization of OIVD is changing. They will be getting an Increasing number of managers. At present the ratio of manager to staff is 1/27. They hope to get b back to 1/10. We are adding postmarked authority for radiology, and while OIVD is currently 175 people, 2012 will see it grow to 213. OIVD will be adding a function specifically for management of public meetings. Dr. G. reminded everyone that the docket is still open on the advance microbiology multiple analytes meeting and he encourages persons to contribute as they will be using the information to write guidance.

Adding management, public meetings Ultra hi throughput and advanced microbiology

Dr. G. Touched on LDTS.

He acknowledged that there is uncertainty around completion of LDT guidances, and mentioned the House legislative proposal.  Gutierrez himself wants to proceed, describes the situation as a “Loop hole getting larger” and having an increasing impact on public health, and especially discussed oncology as an area where manufacturers are doing good clinical trials, but LDTs are simply getting on the market without going through review.

In response to a question about down classification petitions, Dr. G. said to go from Class II to I they would either need a reference to a regulation or for the petitioner to identify a protocol for standard. He stated that OIVD was planning to move a lot of devices through down classification. Dr. G. not sure if there is a public record of planned down classification. He commented that mass specs were down classified and became virtually non-regulated. This is concern for high-tech products.

In response to a question about the RUO guidance, Dr. G. stated that these are, are old policies, the guidance is just a reminder. Everything in it is enforceable. But for the most part, wait for the final to comply. There was a question about certification, which was s not in the guidance. Would FDA include it in the final? Gr. G. said they were not going to address specific methods of compliance.

M. Burns of FDA ODE and A Veoukas from Abbott gave a review of the deNovo guidance.

The principal change is that there will be a pre-deNovo submission that will include the manufacturer’s suggestion for guidance. This will be similar to the current pre-IDE, but will be formalized in terms of the de novo process. The response, if OK, will be a suitability letter, not binding, but laying out the requirements. Then there will be simultaneous submission of the 510(k) and the de novo petition. Draft is open thru Jan3 2012.

Industry concern is that use of published literature is not sufficiently addressed. Also, FDA needs more examples of principles for determining suitability relationship to statutory criteria for de Novo. Read on risk/benefit determinations.

They presented a case study of a biomarker.

Question about pilot: would FDA be willing to go through pilot process, Response, from Alberto Gutierrez was yes. He stated that in OIVD-the new procedure is not much different from current pre-IDE process. De novo does step it up for management review. However, presently not unusual to include the upper management in the tough decisions. Dr. G. thinks the transition in OIVD will be relatively seamless.

A Panel session on companion diagnostics and personalized medicine was chaired by Robert Di Tullio.

E. Mansfield-(FDA) stated the companion diagnostic guidance is moving forward. She defines a companion diagnostic as a test that is required by a drug or biologic, in order to tell whether treatment with that drug/biologic will be medically useful.  In other words, the diagnostics is required to make the Rx products safe and effective. Several companies have already had companion diagnostics approved. The test needs to be approved along with the drug. They received a lot of comment on the draft guidance, in particular the labeling, and whether a companion diagnostic should be called out by name in the drug labeling. Mansfield leans toward a generic description of the analyte in question, not use of brand names.

W. Pignato, speaking for AdvaMed, was pleased to see guidance with molecular diagnostics tied into the notion of companion diagnostics. AdvaMed comments are a reflection of ongoing practices.  They would like to see more concentration on the development cycle, and contemporaneous review. Sore discussion of when is IDE necessary. They would like to know how there can be priority review for the drug but not for the companion diagnostic

A. Dayton from CBER indicated that there will be differences for first: of kind, second in turn compared to one ahead. FDA will want to see a comparison with the standard of care.

E. Lawson, speaking for AMDM, mentioned that AMDM did not comment as AMDM. However, review of IVD labeling and claims by CDER is a concern as different. standards are employed by CDER. Offices have different approaches to claim review. Also there are questions about how orphan drugs requiring non-orphan diagnostics will be treated. He would like this discussed in the guidance. Orphan drug treatment for the drug, but how will this relate to device?

Q. A question from the floor about complementary vs. companion diagnostics.

How does guidance define companion? W. Pignato attempts to respond. A complimentary diagnostic is related to a therapeutic area but not a specific drug.   For example, KRAS in colorectal cancer.

Q. About use for a specific class of drugs.

Class effect is defined in drugs.  If a diagnostic works for one member of a drug class, can it be used for other members of the same class? The question here is if there could be a claim that a companion diagnostic offers same level of safety and efficacy, for a second drug of the same class, without specific trial with the second drug. FDA does not attempt to answer in the guidance, and will continue on case by case.

Q. How will FDA assure that only IVD cleared biomarkers are used as companion diagnostics?

Simply put, they won’t.  Mansfield will not get into relationship. of LDTs for companion diagnostics. LDTs continue under enforcement discretion. She mentioned the option of brand naming the test in the drug labeling, but this has its own problems.

Q. What kind of process to facilitate interaction between centers?

Operating through clinical data review, drugs request a review. It is informal, but we are trying to build in something more formal. We want to build in mechanisms to alert other division to companion diagnostics. We do have consult SOPs already, but not specific for companion diagnostics. All IT systems are different. Informal but we are on top of it. When a drug has a companion diagnostic, she recommend ask for consult with CDER OIVD.

Q. Are you using c modular PMAs for most companion diagnostics?

Yes. Can OIVD approve the diagnostic out of Phase 2 data?  Yes, if final product is tested. Need good coordination of trial with diagnostic supplier. Other considerations in clinical trials are that. the manufacturer may need a clinical trial version of the test. Then you need residual sample to get final version validated.  Also, need to craft informed consent to allow use of residual sample.  China, Japan won’t let samples out of country, may have to perform analysis in the country.

FDA recommends that the drug maker send data using the proposed companion IVD in the IND. There will be a consult.  Some companies do this. Other times people have requested separate IDEs, where confidentiality is an issue. OIVD OK either way. You will get a review of the diagnostic portion of the protocol by OIVD. FDA will probably clarify acceptable routes to obtain comment on protocol in final guidance.

In any case, it is a good idea to have both OIVD and CDER involved in review, and is a good idea to have personnel from both drug and diagnostic companies at all meetings.

Mansfield pointed out how pharmacy and diagnostic companies interact is up to industry, not FDA. AMDM is trying to develop some best practices. FDA is still working on the 2005 co-development guidance. There is a large variety in the way these are dealt with to date.

Jim Kelly from ADVAMED and Kathy Serrano from OIVD talked about the RUO draft guidance.

A lot of what is in the guidance is OK with Industry according to Kelly.

RUO products should not be promoted for clinical use. These are not new ideas, but a long-established rule, made clear in the 92 and 98 compliance policies there was some discussion of the usefulness of a certification program.

From Industry standpoint, the possibility of patient harm seems remote. The more basic issue is LDT.  Proper use of RUO necessary to discovery. The extent of manufacturers’ responsibility to monitor customers’ use of product is not clear.

RUO products are defined as product “in the research phase of development”. However, RUO products are used in basic research not always being researched themselves.

FDA Needs to explicitly describe a certification program. Cert program should be re-stated in current guidance as it was in 98 guidance. Also, research products often need detailed instructions for use.  Often, these products support highly sophisticated methods

There are problems with having user defining intended use of the product. Any customer support could be interpreted as acquiescence in use of product.

FDA standpoint

This represents current FDA thought. RUO products are products prior to design control to do feasibility for an IVD, or develop fundamental knowledge related to disease. No premarket submission is required, but there are labeling requirements for both RUO. And Investigational products – products used to get performance data prior to completion of design for finished product.

Products are shipped for clinical use but labeled RUO/IUO. FDA concern is these products are not made under QSR; performance claims have no independent review, etc. If you are promoting a product for clinical use, you need to label it that way. Not new, been around. 1972 labeling rule.

1997 ASR regulation was written to provide components for LDTs. However, entire kits were sold as ASR, with instructions for use. Final ASR guidance in 2007 made it clear that provision of kits was not the purpose of the ASR Regulation. Manufacturers shifted back to labeling the product RUO/IUO. This current guidance is meant to be published as reminder.

It is not OK to insinuate clinical use. (naming product clinical use)

Unacceptable for the RUO provider to give technical help to a clinical lab, to help with validation, etc.

Guidance has product codes for RUO products.

FDA is not planning to regulate labs, but does regulate manufacturers. If the main market is for clinical use, should be filing. Cant comment on timeline

There have been different types of comments depending on the source. Manufacturers have different issues from lab professionals. FDAs major concern is where major market of RUO product is really diagnostic. There needs to be a balance between not interrupting critical testing and selling mislabeled product.

Sayah Nedjar discussed CBER IVD priorities and updates, mainly by referring to the CBER website.

There was a discussion of regulatory considerations for HLA test kits.

The meeting adjourned at 3:30 pm.