We have previously posted about the heartache and anxiety that is the EUA process (here).  Companies are waiting months for feedback from FDA and are frequently given comically short timelines for response.

Another layer in this onion of poor management is the growing gap between the requirements and expectations that FDA has released publicly to gain authorization and the actual study requirements to support an EUA.

CDRH, helpfully, has created a series of detailed EUA templates for the interactive review process.  Throughout 2020, FDA periodically issued new or revised EUA templates as the pandemic progressed.

Has this been a perfect system? No.  There are still many cases where changes to FDA requirements were applied before new templates were issued and companies with EUAs under review took these changes in the teeth, delaying their authorization while new data was collected.

Sadly, the pace of versioning existing templates has either slowed or stopped for IVD products.  What we are left with now are hard and fast requirements that are barriers to an EUA authorization, that are not disclosed for public consumption.  This gap can leave companies blindsided when their reviewer emails them requesting new data.  In these communications FDA will claim that these requirements have been in place for months, but they are nowhere to be found in any publicly available EUA template or announcement.

Of particular importance:  there are several surprise requirements are all linked to a single issue for PCR and Antigen tests.  This requirement is to ensure that the clinical agreement study is populated with a sufficient number ‘low positive’ samples.

This single concern touches upon many design considerations for a clinical agreement study.

Here are key points to consider:

1. What does it mean for a sample to be ‘low positive’ in the clinical agreement study?

FDA does not have a definitions section in the templates, but the agency does provide some insight as to what ‘low positive’ means in the Molecular Diagnostic Template for Commercial Manufacturers (updated July 28, 2020) when describing the details of a clinical agreement study to support Point-of-Care (Page 34).

You should conduct testing with samples prepared in clinical matrix with SARS-CoV-2 viral load near the LoD of your assay. The testing should be performed by inexperienced users at the clinical sites. The test samples should consist of 10 low positives (<2x LoD) and 10 negatives (matrix) per site.

In reading the above, a reasonable person might assume that when FDA is discussing LoD in this context they are referring to the LoD of the candidate device, as directly stated in the first sentence meaning that the description of ‘low positive’ is <2x LoD of the candidate device.  We have found, however, that a study designed in this fashion would not be considered adequate to support an EUA.  If you are attempting to validate a PCR device for CLIA Labs or Point-of-Care, the above statement is all FDA provides in your template.

To try and get more context for ‘low positive’ one would have to search the templates that were drafted for other indications or device technologies.  For instance, FDA has a slightly different description of low positive in the Template for Manufacturers of Molecular and Antigen Diagnostic COVID-19 Tests for Non-Laboratory Use (July 29, 2020). Low positive is only referred to in this template when describing the Flex Studies, but it largely consistent with the one found in the manufacturers template:  “Flex studies should be conducted by testing a negative sample and a low positive sample (at 1.5x – 2x LoD) for each condition being evaluated.”

FDA, again, has a somewhat different description of low positive in the Molecular Diagnostic Template for Laboratories (updated July 28, 2020) when describing the clinical data required for respiratory panels that implies the description of ‘low positive’ is based on the LoD of the candidate device:

The pre-selection of archived positive samples should represent a range of viral load or Ct values including low positive samples near the assay cut-off.

Finally, FDA has a fourth description of low positive that can be found in the Antigen Template for Test Developers (October 26, 2020)

[L]ow positives (i.e., RT-PCR Ct counts >30)

While it may not seem appropriate to scour an EUA template for an antigen test to gain insight into FDA’s thinking for a PCR test, you will see later in this post that this type of exercise can be informative.

If you are attempting to validate a PCR assay you may be confused as to which of these descriptions of low positive is most appropriate for your test.  The answer, none of them.

Rather, FDA is providing the true or current requirements for low positives during conversations with individual sponsors.  Recently, in the context of PCR assays, FDA has been defining ‘low positives’ as specimens having Ct results obtained with the comparator that are ≤3Ct of the mean Ct for the LoD of the comparator. This description is different than the all of the versions found in the EUA templates.  This is an important difference, because it could mean that your study does not have the correct distribution of sample titers if you did not know about this informally imposed requirement.

2.  Ct values for all positive specimens in the Clinical Agreement study are required

In the Molecular Diagnostic Template for Commercial Manufacturers (updated July 28, 2020) FDA outline the expectations for the positive samples on page 15:

The use of samples previously tested positive by another EUA RT-PCR assay may be acceptable without additional comparator testing. You should indicate the source of the samples, provide results for each tested sample, indicate specimen type, and initial test date.

If you follow the recommendations provided in the template you will likely, at minimum, get follow-up questions, but very possibly a request for new data.  In order to fulfill FDA’s request regarding low positives, you are required to report the Ct values for all positive samples in your study.

While this seems minor, many companies did not ensure that the Ct value was recorded for the positive sample they were purchasing. Most companies purchasing samples confirmed the PCR assay used was EUA authorized. This remediation may not be an issue for a large company as they would likely has direct access to a lab with a comparator device, but smaller companies were relying on the ability use these samples without the need to test them on any other device beyond their own.  In many cases this has forced smaller companies or universities to purchase additional samples and re-run the experiment.

FDA officials have indicated in conversations that reporting of Ct values for clinical agreement is a long-standing request for EUAs, but we have not been able to independently confirm this assertion.

When purchasing samples for a retrospective clinical agreement study you should be able to report the following for samples tests with a high-sensitivity PCR assay: source of the samples, results for each tested sample with Ct values, specimen type, and initial test date

3.  Make sure that your PCR comparator is a high-sensitivity test

The three main diagnostic technologies for the pandemic are:  PCR, Antigen, and Antibody tests.  In each of the templates for these devices you will find a version of the following statement:

Positive percent agreement should be calculated in comparison to an EUA RT-PCR test. We recommend using only a high sensitivity EUA RT-PCR assay which uses a chemical lysis step followed by solid phase extraction of nucleic acid (e.g., silica bead extraction). If available, FDA recommends selecting a comparator assay that has established high sensitivity with an internationally recognized standard or FDA SARS-CoV-2 Reference Panel.

For nearly a year now, we at HPM have been trying to find out what ‘high sensitivity’ meant in this context.  In conversations with FDA, we have been told to propose an assay as the comparator and FDA would then comment.

We recently were finally provided with an answer to this ‘high sensitivity’ question.  To be considered a ‘high sensitivity’ test, we now understand that the assay must have an LoD that is at or lower than 18,000 NDU/mL according to FDA’s Reference Panel.  Setting the threshold to 18,000 NDU/mL effective eliminates at least 17 PCR assays from being used as a comparator for other tests.  We expect the true number to be higher as the list containing reference panel results has not been updated since October 2020.  We were not provided with an explanation as to why 18,000 NDU/mL was an appropriate cutoff, but we will take whatever clarity we can get from the review staff.

This unpublicized requirement disproportionally impacts smaller companies. A larger company has access to more resources and can more easily shoulder the burden of retesting samples or purchasing new ones that were tested with an appropriate comparator, but it can be fatal for smaller companies without these resources.

If you did not pick a test that is high-sensitivity as FDA defines it, it is not your fault, as FDA has not previously provided guidance as to which of the authorized tests can be used as a comparator. The adverse consequences of selecting the wrong assay are real.  Companies run the risk of their data being considered unsuitable to support an EUA.

4.  What does it mean to have a sufficient number of low positive samples

In communications with FDA, the agency has insisted that the clinical agreement study consist of not only a range of viral loads and Ct values as stated in the template but, at minimum, 20% low positive specimens.

The requirement for at minimum 20% low positive samples is not found in the following templates:

• Molecular Diagnostic Template for Commercial Manufacturers(updated July 28, 2020)
• Molecular Diagnostic Template for Laboratories(updated July 28, 2020)
• Template for Manufacturers of Molecular and Antigen Diagnostic COVID-19 Tests for Non-Laboratory Use(July 29, 2020)

In FDA’s templates for PCR assays one can find only the following recommendation when FDA describes the Clinical Evaluation study:

Specimens representing a wide range of viral load including low positive samples should be tested.

We refer to this request for 20% ‘low positives’ as a requirement since FDA has flagged studies as insufficient and requested additional data if a retrospective clinical study does not have at least 20% low positives samples.

One can find only a single reference to a specific amount of low positives in the Antigen Template for Test Developers (October 26, 2020):

Retrospective specimens should be reflective of the natural distribution of SARS-CoV-2 viral loads, and approximately 10-20% of the clinical specimens should be low positives (i.e., RT-PCR Ct counts >30), as has been observed in other sequentially enrolled clinical studies.

As you can see from the excerpt above, the only specific guidance can is found in a template for a different assay technology where the requested amount of low positives is as little as 10%.  If you do follow this guidance and present a study with 10-20% low positives, your study will likely be deemed inadequate.

5.  Risks of picking a too‑sensitive comparator assay

The gut reaction in response to the above may to be over-correct and pick the absolute highest sensitivity device possible.  As stated in section 1 above, low positive is defined by the comparator assay. If you pick a comparator that is more sensitive than your test, you will not only increase the likelihood of discordance, but you will also narrow the window of what it means for a sample to be ‘low positive’. You then run the risk of not meeting the minimum requirement to have 20% of the samples in that range.  It is recommended that you choose a comparator with a similar LoD to your assay that is also still considered ‘high-sensitivity’ per FDA’s Reference Panel.

Based on the foregoing, we have the following recommendations to FDA to improve the EUA process:

1. Issue a statement defining ‘high-sensitivity’ and ‘low positives’ as this lapse is causing unnecessary delay in bringing new products to market and is creating needless inefficiency.
2. Issue updated templates with study design requirements that are consistent with FDA’s expectations for EUA authorization. Some of the templates have not been versioned in 9-11 months.
3. Commit to updating the EUA templates on a regular basis for the duration of the emergency to prevent this problem from recurring.
4. Commit to allowing pending EUAs to resolve issues via new testing as a post‑authorization requirement rather than applying unpublicized requirements retroactively.
5. Update the SARS-CoV-2 Reference Panel Comparative Data Table, as it is 6 months out of date and data from new assays or updated data from existing assays are almost certainly sitting in the shadows.

We strongly urge all developers of PCR assays that are about to submit an EUA or those that have an EUA pending to review the details of your clinical agreement study to make sure that your data will not get flagged by the reviewer.  We are finding that gap analyses are more important than ever to reduce review delays for EUA products.

On April 2, 2021, the Ninth Circuit issued a decision in a False Claims Act (FCA) case against Medtronic.  The Dan Abrams Co. LLC v. Medtronic Inc., No. 19-56377, 2021 U.S. App. LEXIS 9637  (9th Cir. Apr. 2, 2021).  Relator appealed the U.S. District Court for the Central District of California’s dismissal of its FCA lawsuit alleging submission of false claims to Medicare due to: (1) alleged unlawful promotion of spinal surgery implants for off-label and contraindicated uses, and (2) alleged fraud-on-the-FDA in obtaining 510(k) clearance for the subject devices.

Regarding off-label and contraindicated use, Relator alleged that Medtronic marketed the devices without FDA clearance or approval for use in the cervical spine.  Such use was both off-label and specifically contraindicated in the device labeling.  The Ninth Circuit affirmed the district court’s dismissal of these claims, finding that the off-label and contraindicated use was not material to the government’s decision to provide Medicare reimbursement for Medtronic’s device.  Specifically, the Court noted that “the federal government acknowledges that doctors may use medical devices for off-label and even contraindicated uses if they believe that such use is medically necessary and reasonable.”  This appellate-level decision provides welcome precedent to support a lack of materiality argument in other FCA cases involving devices reimbursed for off-label uses.

Relator’s fraud-on-the-FDA claim was based on an allegation that Medtronic defrauded FDA into granting the subject devices 510(k) clearance by “falsely represent[ing] in its clearance application that they were intended for use in the thoracolumbar spine (the part of the spine below the neck) when in fact they could not be used there and could only be used in the cervical (neck-area) of the spine.”

The Court acknowledged that the Supreme Court held in Buckman Co. v. Plaintiffs’ Legal Comm. that the Federal Food, Drug, and Cosmetic Act bars a private party from asserting state law claims that a device manufacturer defrauded FDA during the 510(k) clearance process concerning a device’s intended use.  531 U.S. 341, 348 (2001).  The Court further acknowledged that the First Circuit has extended the Supreme Court’s holding in Buckman to the FCA context (see our blog post on this First Circuit decision here).  However, unlike the First Circuit, because the Ninth Circuit previously allowed fraud-on-the-FDA claims in the FCA context, the Court here reversed the district court’s dismissal of Relator’s claim, allowing the fraud-on-the-FDA theory to go forward.

On March 31st, 2021 BARDA launched a competition to design tomorrow’s mask. The Mask Innovation Challenge, at its core, is about finding new mask designs that can be mass produced at a low cost and provide general protections against respiratory pathogens. One of the lessons learned from this pandemic is that barriers exist within the population when it comes to widespread adoption and consistent use of face masks.  This challenge identifies the specific problems that the contestants should consider in their device designs:

• Currently available retail masks are often untested, with unknown protective capability
• Physical discomfort with prolonged use, particularly in hot and humid environments
• Perceived or actual breathing difficulties
• Irritant contact dermatitis with extended wear
• Inability to effectively communicate with others using facial expressions
• Speech intelligibility and difficulty with glasses wearing,
• Lack of understanding of the features of a mask

Potential contestants should read the rules of the challenge carefully as there are delineations between the two potential tracks for applications.

1. Barrier Face Coverings: Designs in this track focus on improving upon already existing designs for cloth facemasks.
2. Other Designs/Technologies: Designs in this track incorporate new technologies that have not yet been included in current mask designs.

This competition has two phases. Phase 1 could have up to ten (10) winners walking away with $10,000 each and Phase 2 will see five (5) winners splitting a prize pool of $400,000.  If you do not get selected for Phase 1 do not fret as you may still be available for Phase 2. There is no expectation of a prototype or finished product for Phase 1. These submissions are not very long with an approximate 1500 word count limit across the nine discreet sections of the proposal.

If I can make one humble request, please make your masks available in an array of colors.  The only attribute that I share with fictional Telenovela star, Rogelio de la Vega, is that I don’t pop in peach.

The deadline to submit is 5pm (Eastern Time) on April 21st, 2021.

There’s no question that drug pricing is one of the most important issues in health care right now, and while such pricing considerations are outside FDA’s statutory mandate, it has not stopped FDA from trying to address pricing issues through enhancing drug competition (see, for example, our discussions of the Drug Competition Action Plan and the Biosimilar Competition Action Plan).  To that end, FDA has attempted to limit exclusivity awards in an effort to facilitate generic access, by, for example, requiring proof of clinical superiority for Orphan Drug Exclusivity (Depomed) or by limiting the definition of active ingredient to active moiety (Vascepa).  But courts have not necessarily agreed with FDA’s interpretation of its governing statute (the FDCA).  In some of these instances—Depomed serving as a prime example—Congress has not hesitated to step in.

Most recently, the Senate passed legislation amending the definition of “active ingredient” with respect to New Chemical Entity (“NCE”) exclusivity, presumably in an attempt to address the Vascepa litigation.   In a case that we followed closely in 2015, Amarin challenged FDA’s determination that its Vascepa Capsules (icosapent ethyl) was not entitled to New Chemical Entity exclusivity based on FDA’s theory that Vascepa’s “active ingredient” is a component of a mixture that makes up the “active ingredient” of Lovaza, a previously approved drug.  FDA opined that when a mixture is well-characterized and components of that mixture have been identified as both the active ingredient and the active moiety in another product, a component of that mixture is not eligible for NCE exclusivity because it contains a previously approved active ingredient or moiety.  The key legal issue in that case was framed as “whether the prior approval of a drug product, the active ingredient of which is a complex mixture of constituents, constitutes approval of each constituent as an active ingredient so as to preclude NCE exclusivity for a new drug product in which one of those constituents alone is the active ingredient.”  Central to this dispute was the definition of “active ingredient” as compared to “active moiety” in the context of mixtures.  While the statute offers NCE exclusivity for a new “active ingredient,” FDA regulations offer it for a new “active moiety.”  And while the distinction is often inconsequential, it becomes relevant for poorly characterized or well characterized mixtures.  As we explained back in 2015, Judge Moss was not too convinced by FDA’s conflation of active moiety and active ingredient, noting that it contradicts the plain language of the statute and is inconsistent with other provisions (and application of those provisions) of the statute.

Presumably with this decision in mind, Congress is granting FDA permission to conflate those definitions in a recent bill passed by the Senate in mid-March 2021.  S.415, entitled “An Act to amend the [FDCA] with respect to the scope of new chemical exclusivity” strikes the term “active ingredient” throughout the drug provisions set forth in section 505 of the FDCA and inserts “active moiety.”  Under this new definition, an active moiety approved in a mixture of an active ingredient can now block a component of that active moiety used alone as an active ingredient from obtaining NCE exclusivity.  This bill gives FDA the discretion it relied upon in the Vascepa litigation to make narrower NCE exclusivity decisions and therefore further restrict awards of such exclusivity and bolsters any denials by further review of an Advisory Committee.  Though, practically, the bill will have little effect, active ingredients composed of “mixtures” can no longer rely on Judge Moss’s decision.  Further, this new definition of “active ingredient” refers all new active moieties to an Advisory Committee for review unless FDA provides a reason not to do so.  Theoretically, this will ensure that all NCE determinations—well characterized active ingredient or otherwise—are consistent.

In a different but related bill, Congress attempts to address biological product pricing through further education about biosimilars.  In S.164, “Advancing Education on Biosimilars Act,” Congress provides authority for FDA to maintain and operate a website providing educational materials regarding biosimilars and interchangeable biosimilars.  The bill also requires FDA to publish the action package and summary review of each approved biologic and biosimilar.  It’s not clear how this bill would change the status quo.  Firstly, under the bill, the provision of educations resources is optional, as the bill uses the term “may” rather than “shall.”  Secondly, FDA already engages in this type of education on its website and as part of its Biosimilars Competition Action Plan.   And FDA already publishes the summary basis of approvals for (at least some) biologics and biosimilars, which you can access through the Purple Book, in its drug database.  Perhaps this a response to the rampant misinformation about the quality of biosimilars, but, again, it’s not entirely clear how this will help.

Both S.415 and S.164 have cleared the Senate and are up for review in the House of Representatives.

On March 24, 2021, Virginia became the latest state to enact a prescription drug price transparency law.  Similar to other state drug price transparency laws (see, for example, our coverage here and here), HB 2007 sets forth reporting requirements applicable to health carriers, pharmacy benefit managers (PBMs), wholesale distributors, and manufacturers.

Under the new law, health carriers are required to submit annual reports with information about the carrier’s spending on prescription drugs, including identifying the 25 most frequently prescribed outpatient prescription drugs, the 25 outpatient prescription drugs covered at the greatest cost, and the 25 outpatient drugs that experienced the greatest year-over-year increase in cost.  PBMs are required to submit an annual report with information about the manufacturer rebates received by the PBM, the rebates distributed to health benefit plans, and the rebates passed on to enrollees of health benefit plans.

Manufacturers are required to submit annual reports for each (i) brand-name drug and biologic with a wholesale acquisition cost (WAC) of $100 or more for a 30-day supply or a single course of treatment and an increase of 15% or more in the WAC over the preceding calendar year; (ii) biosimilar with an initial WAC that is not at least 15% less than the WAC of the referenced brand biologic at the time the biosimilar is launched; and (iii) generic drug with a price increase that results in a WAC increase that is equal to 200% or more during the preceding 12-month period, when the WAC of that generic drug is equal to or greater than $100 for a 30-day supply.  A price increase for a generic drug is defined as the difference between the WAC of the generic drug following an increase and the average WAC of the generic drug during the previous 12 months.  For each reportable drug, manufacturers must provide the following:

• Name of prescription drug;
• Identify whether brand name or generic;
• Effective date of the change in WAC;
• Aggregate, company-level research and development costs for the most recent year for which final audit data is available;
• Name of each new prescription drug approved by the FDA within the previous 3 calendar years;
• Name of each prescription drug that, within the previous 3 calendar years, became subject to generic competition and for which there is a therapeutically equivalent generic version; and
• A statement regarding the factor(s) that cause the WAC increase.

Upon a determination that the data received from the health carriers, PBMs, and manufacturers is insufficient, the law allows the Virginia Department of Health to request reports from wholesale distributors with information on WACs, rebates, discounts, price concessions, and other fees related to the 25 costliest drugs.

The law states that the Virginia Department of Health shall engage a nonprofit data services organization to collect, compile, and make information submitted by the health carriers, PBMs, wholesale distributors, and manufacturers available on a public website.  The data and information will be made available in aggregate in a form and manner that does not disclose proprietary or confidential information of the reporting entities.  The law provides that information submitted pursuant to the new prescription drug price transparency requirements is excluded from disclosure under the Virginia Freedom of Information Act, and also specifies that a manufacturer’s reporting obligations shall be fully satisfied if the manufacturer reports the information and data that would be included in reports to the Securities and Exchange Commission or any other public disclosure.

The law specifies that regulations must be promulgated within 280 days of enactment – which would be December 29, 2021.  The Virginia prescription drug price transparency law will become effective on January 1, 2022, so the first annual reports will be due by April 1, 2022.  Entities that do not comply with the new reporting requirements may be subject to a civil penalty of up to $2,500 per day from the date that any such reporting was due.

FDA has long said that it would consider real-world evidence (RWE) in making regulatory decisions related to medical devices.  Those in industry know that FDA can be very critical of RWE, however, and it is not always clear why RWE was (or was not) acceptable in a particular submission.  Indeed, in 2017, when FDA released its guidance on acceptance of RWE, this author was skeptical as to whether it would actually result in an increase in Agency acceptance of RWE in medical device decision making (see earlier post here).

Earlier this month, FDA released a summary document describing 90 instances in which RWE was used in medical device decision making (see report here).  Interestingly, the examples in the report span between fiscal years 2012 and 2019, making it difficult to identify whether there has been any kind of shift in FDA acceptance of RWE following issuance of the 2017 guidance.

The report is helpful, however, in giving sponsors examples of where FDA has been accepting of clinical data in recent history.  In the press statement announcing release of the report, FDA stated, “When reviewing the use of RWE to support a regulatory decision, the FDA relies on scientifically robust methods and approaches to determine whether the submitted RWE is of sufficient quality to support the regulatory decision.”  (see March 16, 2021 announcement here) While the considerations for acceptance of RWE were made clearer in the 2017 guidance, the guidance included a modest list of six examples of where FDA had accepted RWE to illustrate the criteria in the guidance.  At 183 pages long, this new report can essentially serve as a robust appendix to the guidance giving sponsors a much more comprehensive and wide-ranging list of examples.

The report includes examples from all submission types, including 510(k)s, de novos, HDEs, and PMAs.  It also includes examples of RWE used in both pre- and post-market decision making.  The report is organized into six sections and is separated by device type (therapeutic devices, in vitro diagnostics) and RWE source (Registries, Administrative Claims Data, Medical Records, Other Sources).  The report also discusses each RWE example individually.  While the discussion of the examples does not give the specific reasons why FDA concluded that the RWE was acceptable, the examples should, nonetheless, provide sponsors with analogies upon which it can draw when arguing for its own RWE.

In the report, FDA encourages sponsors to continue to use RWE to support device regulatory submissions throughout the entire product lifecycle.  FDA also advises sponsors who are considering using RWE to consult with FDA via the pre-submission process, as needed, “to understand how to best utilize the RWE to support the marketing claims.”

We commend FDA for making these examples public, although caution sponsors that these examples only represent one-side of the story.  In our experience, there are plenty of examples of cases where FDA has not accepted RWE due to the quality and/or availability of details in the underlying data.  Thus, while this list looks robust and encouraging, sponsors should not consider this as an endorsement by FDA of any and all RWE.

Tooth whiteners – drugs or cosmetics or both?  In the case of certain tooth whiteners, the regulatory status remains gray, apparently with no prospect of illumination from the Agency.

As we reported back in 2014, FDA’s interest in the regulation of peroxide-containing tooth whiteners began in the early 1990s not too many years after these products were introduced to the market. In 1991, FDA sent Warning Letters to several manufacturers stating that it considered products which claimed to whiten teeth by bleaching were unapproved new drugs being marketed in violation of the Food, Drug, and Cosmetic Act.  One of the manufacturers (Den-Met Holdings, LLC) sued FDA asserting its product (Rembrandt Lighten Bleaching Gel) was a cosmetic, rather than a drug under the statute.  Den-Mat withdrew its case in 1992 after FDA agreed to review new information submitted and determine whether the new information changed FDA’s position. And then, nothing.

“Fast” forward to 2009 when the American Dental Association submitted a Citizen Petition (Docket No. FDA-2009-P-0566) requesting that FDA establish an “appropriate regulatory classification” for peroxide-containing tooth whiteners citing concerns about their safe use without benefit of professional consultation or examination.  FDA responded in 2014, denying the petition, but leaving open the question of the appropriate regulatory status for peroxide-containing tooth whiteners.   In the response, FDA described the different types of tooth stains and their causes, and concluded that the means by which whitening occurs may be different for different types of stains.  FDA denied the petition stating there was insufficient data to determine whether, as a group, peroxide-containing tooth whiteners that act by a chemical means met the definition of a drug (as well as a cosmetic) and specifically noting the need for data on specific products to determine on a case-by-case basis whether individual products are intended to affect the structure or function of the teeth and/or intended to mitigate or treat a disease.  In conclusion, FDA stated that without further data illuminating the mechanism of action, it could not answer the question of whether all tooth whiteners as a group met the definition of a drug.

FDA has said no more about it since 2014. Meanwhile, in the private sector, battles over tooth whiteners have continued over the years with a very recent case before the National Advertising Division about Optic White Toothpaste (NAD Case #6914, 2/23/21) and a number of false advertising cases about whitening products pending in the courts.  As a result of that recent activity, the trade journal HBW Insight (a subscription publication) reported recently that it had reached out to FDA for clarification on the regulatory status of these products.  US FDA’s Position Hasn’t Changed On Hydrogen Peroxide Teeth Whiteners – HBW News, 18 Mar 2021.  As described by HBW News, there does not appear to be any change in FDA’s position on peroxide-containing tooth whiteners and they appear to remain in the gray zone where they have dwelled for the past 30 or so years.

Perhaps at some point FDA will utilize the new processes established under OTC Monograph Reform (see our blog here) to force the issue and issue a proposed order that peroxide-containing tooth whiteners or some subset of them are non-monograph unapproved new drugs, but little in the past decades or recent responses to HBW News suggests it is a priority.

On March 24, 2021, the U.S. Department of Health and Human Services Office of Inspector General (“OIG”) issued its first advisory opinion of the year.  Advisory Opinion 21-01 addresses whether the provision of a specific drug at no cost by a pharmaceutical manufacturer to a physician or healthcare facility violates the Federal health care program antikickback statute (“AKS”) or the civil monetary penalty (“CMP”) provision prohibiting certain inducements to beneficiaries.  OIG concluded that it would not impose administrative sanctions or civil monetary penalties in connection with the program proposed by the pharmaceutical manufacturer.  OIG’s conclusion was, however, highly dependent on specific facts surrounding the drug at issue.

Background

The drug is intended to be administered one time only per patient and is potentially curative with that one dose.  The FDA subjected the drug to a Risk Evaluation and Mitigation Strategy (REMS) to ensure its safe use.  The Elements To Ensure Safe Use (“ETASU”) under this REMS include that it may only be administered at a certified healthcare facility and prescribed only by a physician trained to meet the requirements of the drug’s REMS.  The pharmaceutical manufacturer is responsible for certifying the health care facilities where the drug can be administered.

Typically, the drug is administered only once.  In order to receive the drug free of charge, a patient must have an on-label prescription, be uninsured or have insurance that does not cover the drug, and have a household income not exceeding a limit of $75,000 for a single person plus $25,000 for each additional person.  While patients with any government or commercial insurance are eligible for the free drug program, the manufacturer represented that Medicare beneficiaries have not, and likely will not, qualify for the program.  However, beneficiaries of other Federal health care programs, including Medicaid and TRICARE, may qualify for the program.  The prescriber and/or certified healthcare facility are prohibited from billing any Federal health care program for the cost of the drug.  However, administration and other ancillary costs can be billed.

OIG’s Decision

Based on these facts, OIG concluded that the free drug program implicates the AKS because provision of the drug at no cost constitutes remuneration and may induce prescribers and hospitals to prescribe the drug and patients to select the drug for treatment.  Prescribers and hospitals would be incentivized to do so because they could bill for administration and other services involved with injecting the drug into the patient.  Patients would be incentivized to select the drug because it would be free and, therefore, may cost less than other alternatives.  Despite these incentives, OIG determined that the risk of fraud and abuse was low, for several reasons.

First, OIG stated that the risk of seeding (i.e., inducements for future use of a federally reimbursable drug) is minimal because, unlike most drugs, this drug is administered only once and is individually made for each patient.  OIG also noted that, although providers can bill for administration fees, the risk that the program will cause the drug to be overused is negligible because, besides being used only once, it must be administered on label, and it is not indicated for first line therapy.  Therefore, in order to receive this drug cost-free, a patient would have to fail at least two other therapies.  This reduces the likelihood that this free drug program will induce prescribers to use this drug versus other, cheaper options.

OIG next determined that the free drug program did not implicate the CMP, which prohibits inducements to beneficiaries that the offeror knows, or should know, is likely to influence the beneficiary to select a specific provider, practitioner, or supplier.  OIG repeated its long-held policy that a drug manufacturer is not a “provider, practitioner, or supplier,” so an inducement to use a manufacturer’s drug does not violate the CMP, even if it does violate the AKS.  However, a pharmaceutical manufacturer can violate the statute if it offers remuneration to a beneficiary that the manufacturer knows (or should know) will influence the beneficiary to use a particular provider (such as a hospital), practitioner (such as a physician), or supplier (such as a pharmacy).  The OIG found no violation under these specific facts because the free drug program places no limits on which prescriber or healthcare facility a patient may use, other than those imposed by FDA’s REMS requirements.

Future Implications

This OIG advisory opinion is one of several approving of free drug programs in specific factual circumstances.  For example, Advisory Opinion 15-11 approved of a manufacturer’s program to provide a one- to two-month supply of a drug cost-free where a new patient was experiencing reimbursement delays (see our blog post here).  Another series of advisory opinions approved of free drug provided to Medicare Part D enrollees in financial need outside of the Part D program, in accordance with OIG guidelines.  However, caution must be exercised in extrapolating these opinions to other free drug programs, because the opinions are highly fact specific.  Nevertheless, certain questions are consistently considered by the OIG in these opinions:  whether the free drug could induce use of the same or other drugs that are federally reimbursable, whether it could have an adverse effect on patient care, and whether it could influence the use of particular providers or suppliers.  The new opinion is unique in that the drug is intended to only be administered once and is individually made for each patient.   However, as more biologics come to market, along with other forms of personalized medicine, certain aspects of this opinion can be useful in providing a roadmap to pharmaceutical manufacturers regarding the design of a program to provide important drugs to patients that otherwise could not afford them.

For more than five years now, FDA has pursued action regarding cannabidiol (CBD) products.  FDA has taken the position that CBD is not a lawful food or dietary ingredient.   But despite the agency’s strong statement about CBD, FDA has acted, primarily through Warning Letters (WLs), only when the claims for CBD are egregious.   Last week, it shifted focus to CBD as ingredient in over-the-counter (OTC) drug products.

On March 22, FDA announced that it had issued warning letters, here and here, to two companies for selling OTC drugs for pain relief containing CBD.  FDA asserts that CBD not only is an impermissible active ingredient in such products, it also is an impermissible inactive ingredient because it has known pharmacological effects with demonstrated risks in humans with demonstrated risks.  In other words, according to these WLs there is no place for CBD in OTC drug products.

Citing numerous examples of marketing claims, FDA asserts that, although CBD is listed as an inactive ingredient in the OTC topical analgesics, the labeling and marketing claims for the Companies’ CBD-containing topical pain-relieving products represent CBD as an active ingredient.  Since CBD was not an active ingredient in any applicable final OTC drug monograph or tentative final monograph, the CBD-containing drug products are unapproved drugs and do not meet the requirements of FDC Act § 505G.

Somewhat surprisingly, FDA also asserts that even if the products were properly labeled and the claims did not imply that CBD is an active ingredient, the products would still be unapproved drugs because CBD does not qualify as an inactive ingredient.  Specifically, FDA asserts that CBD is not a valid inactive ingredient because it does “not conform with the general requirement in 21 CFR 330.1(e) that inactive ingredients must be safe and suitable;” CBD has no known functional role as an inactive ingredient in a finished drug product and is not safe.  It is not suitable because “a suitable inactive ingredient generally provides a beneficial formulation function, such as a tablet binder or preservative, or improves product delivery (e.g., enhances absorption or controls release of the drug substance),” and FDA does not know of such a beneficial function for CBD in a finished drug product.  In support, FDA cites guidances regarding excipients in FDA approved drugs, which arguably do not apply to OTC drugs marketed under FDC Act 505G.  Even though, as the Agency acknowledged in the WLs, FDA does not know whether the levels in the products at issue have pharmacological activity, it concludes that the mere fact that CBD has pharmacological activity in an approved oral drug  causes the ingredient to be unsafe as inactive in the topical OTC drugs.

It is not clear what inspired FDA to take this action and it remains to be seen if these two letters are just the beginning.  We will be monitoring FDA activity regarding CBD.

On March 17, 2021, FDA granted BioFire Diagnostics’ De Novo, making it the first COVID assay originally authorized on a temporary basis for this public health emergency to be given permanent access to the US market. While FDA has only released the signed letter affirming their permanent marketing status, we can infer a few key points:

1. This De Novo was product of collaboration between FDA and industry over the course of months.

In the letter granting the De Novo it is important to note the date that the De Novo was “Received” by FDA May 19, 2020. This is significant as it means that it is very likely that the candidate device changed from the initial submission to the De Novo being granted. The BioFire® COVID-19 Test was initially authorized by FDA on March 23, 2020 which is well in advance of the DEN200031 being received by FDA.

The BioFire Respiratory Panel 2.1 was submitted for FDA review under EUA202392 meaning that this EUA was originally submitted for FDA review sometime between mid-July and mid-September.  As this device was going to ‘set the bar’ for future clearances FDA likely allowed BioFire to add data and indications to the device from the original submission in March through at least October 2, 2020 when the BioFire Respiratory Panel 2.1 was originally authorized as an EUA assay.  It is reasonable to assume that if BioFire did not have enough data till mid/late summer to submit an EUA for the BioFire Respiratory Panel 2.1, they did not have enough data for a complete submission back in May 2020.

While this is not typically allowed under normal circumstances it can be viewed as an action to everyone’s benefit as is explained below.

2.  A more comprehensive De Novo submission allows for more EUA devices to declare this test as a predicate. The new regulation (21 CFR 866.3981) is as follows:

Device to detect and identify nucleic acid targets in respiratory specimens from microbial agents that cause the SARS-CoV-2 respiratory infection and other microbial agents when in a multi-target test.  A device to detect and identify nucleic acid targets in respiratory specimens from microbial agents that cause the SARS-CoV-2 respiratory infection and other microbial agents when in a multi-target test is an in vitro diagnostic device intended for the detection and identification of SARS-CoV-2 and other microbial agents when in a multi-target test in human clinical respiratory specimens from patients suspected of respiratory infection who are at risk for exposure or who may have been exposed to these agents. The device is intended to aid in the diagnosis of respiratory infection in conjunction with other clinical, epidemiologic, and laboratory data or other risk factors.

As shown by the bolded language, this regulation means that not only can the multitudes of single analyte RT-PCR tests claim this device as a predicate, but also the multiplexed assays will be able to use this De Novo as a predicate and proceed down a less burdensome 510(k) pathway.  Furthermore, the regulation is worded to encompass the expected mutation of the virus by defining the measurand as simply “nucleic acid targets” from “microbial agents that cause the SARS-CoV-2 respiratory infection.”

Throughout the pandemic FDA has been very particular about the specific respiratory specimens a device has validated for use.  FDA had previous stratified the clinical matrices between the “upper” and “lower” respiratory tracts requiring, in most cases, a representative sample from each.  While the BioFire assay is only indicated for nasopharyngeal swabs, the regulation is crafted broadly to encompass samples from all portions of the respiratory tracts.

It appears that this new regulation is only applicable to assays where the measurand is a nucleic acid.  This means that both antigen and antibody tests will likely need their own regulations created via the granting of De Novos for their respective technologies.

3.  The controversial ‘FDA Reference Panel’ will likely survive the pandemic and fight on to plague industry for years to come. Item 5 of the Special Controls list states the following:

When applicable, performance results of the analytical study testing the FDA recommended reference panel described in paragraph (b)(4)(vi) of this section must be included in the device’s labeling under 21 CFR 809.10(b).

The reference panel was FDA’s attempt in the early stages of the pandemic to gain an understanding of assay performance across IVD manufacturers via a single standardized panel.  The panel was met with controversy as some test manufacturers were obtaining inconsistent or otherwise confounding results when using the test samples.  It appeared to many that the issue was not the tests but the panel.  Apparently, there are tests with perfectly good performance that, for some reason or other, are not optimized for the panel.

This use of this panel began showing up in “Conditions of Authorization” for PCR EUAs making the testing compulsory when directed by FDA if a manufacturer wished to stay on the market.  In another surprise move, FDA used the performance on their reference panel to rank assays by sensitivity on a public facing website.  This website has not been maintained in recent months and is only current as of December 7, 2020.

A manufacturer’s performance with this panel could yield both regulatory and business advantages.  If an assay performed well, it afforded the manufacturer the opportunity to pursue an asymptomatic claim for the assay before completing the required clinical study. The device would be considered a ‘more sensitive’ test that would potentially drive new business as a comparator for other assays.  Links to the website containing the data from FDA’s reference panel can be found in many of the issued EUA templates:

From FDA’s “Molecular Diagnostic Template for Commercial Manufacturers” regarding “adding population screening of individuals without symptoms or other reasons to suspect COVID-19 to an authorized test”:

If your assay is highly sensitive as determined by testing with the FDA SARS-CoV-2 Reference Panel or a recognized international standard, a post-authorization study may be appropriate.

From FDA’s “Antigen Template for Test Developers” regarding “POC Clinical Evaluation”:

The comparator method should be one of the more sensitive RT-PCR assays authorized by FDA. We encourage you to review the results from the FDA SARS-CoV-2 Reference Panel available here.

Many manufacturers will likely hope that when FDA states “when applicable” in the special controls with respect to the reference panel that FDA means sometime after the eventual heat death of the universe.

4.  This submission set the bar for all the other manufacturers wishing to stay on the market post-pandemic.

A key parameter of substantial equivalence is the performance in the clinical agreement study. While OHT-7 typically posts the decision summary for public consumption it has yet to do so for this De Novo (we expect it to be posted in the coming days).  While the performance in the decision summary will be the true bar for substantial equivalence, we can look to the labeling for the EUA authorized device to get a preview of what to expect. On page 24 of the labeling, you can find the overall Sensitivity of this device for SARS-CoV-2 is 98% with a Specificity of 100%.  We do not know what this performance means for assays that had acceptable clinical agreement as an EUA authorized test (Sensitivity ≥95% and Specificity ≥98%) but do not meet this new bar for performance. As more COVID assay manufacturers scramble to convert their temporary marketing authorization afforded by the EUA pathway into a permanent clearance, these initial devices will be compared to BioFire’s to determine substantial equivalence.  If FDA clears a device with a lower sensitivity and specificity than BioFire’s this new device can then be used a predicate for others thereby lowering the bar for substantial equivalence for the rest of industry.

This incremental lowering of acceptable performance standards is a permanent fixture in FDA’s thinking when it comes to clearing device via the 510(k) pathway. Typically, OHT-7 is more intractable than Captain Picard when facing the Borg when it comes to clearing a test with lower clinical performance than the declared predicate. We do not yet know how FDA will apply the substantial equivalence paradigm in the future but granting a De Novo to a device with very high clinical agreement could be used as a gating mechanism to weed out devices that were adequate for the pandemic, but not suitable as a permanent fixture in the US market.

Over the past several years, FDA has faced criticism stemming from high-profile device issues related to materials, including the Essure permanently implanted birth control device and metal-on-metal hip implants. Given this, it is not surprising that device biocompatibility has received greater focus in FDA premarket reviews for medical devices.  FDA  initially issued Use of International Standard ISO 10993-1, “Biological evaluation of medical devices – Part 1: Evaluation and testing within a risk management process” (Guidance) in 2016 to provide guidance to sponsors of premarket applications on evaluation of biological risk and biocompatibility testing for devices that are in direct or indirect contact with the human body.  However, many device sponsors have continued to face challenges demonstrating device biocompatibility, with many sponsors receiving additional information requests related to the biocompatibility data necessary to support a marketing application.  The Guidance describes a risk-based approach where justification may be provided to support the overall testing strategy and interpretation of results.  Many additional information requests that we have seen relate to disagreements between FDA and the sponsor with respect to justifications provided by the sponsor on what tests are applicable to the device, how the device tested is representative of the device intended for commercialization, the impact of differences between device tested and the device intended for commercialization, test methods, especially for extraction, and interpretation of results.

On March 19, 2021, FDA announced the launch of an online Biocompatibility Assessment Resource Center to provide more clarity for sponsors navigating biocompatibility requirements.  The new resource is intended to explain terms and concepts for use in conjunction with the Guidance.  The site is divided into four steps:  1) Biocompatibility Basics, 2) Evaluation Endpoints, 3) Test Articles, and 4) Test reports.

Biocompatibility Basics provides a very high-level description of the basis for biocompatibility and a glossary of biocompatibility terms.  In some ways, it is a repackaging of the Guidance.  The Basics page describes devices for which biocompatibility information is needed, what FDA assesses or evaluates, how FDA assesses or evaluates biocompatibility and biocompatibility factors of interest to FDA.  In the glossary, there is quite a bit of overlap in terms defined compared to the Guidance, with some new terms added, most of which come from the ISO 10993 series of biocompatibility standards.

Evaluation Endpoints provides the same information as Attachment A of the Guidance in an easy-to-use format.  One statement of note is that FDA indicates that device categorization information can be obtained informally via e-mail or as part of the pre-submission process.  Information on Test Articles includes the same information as Attachment F of the guidance, providing example text to use when relying on biocompatibility data from a test article that is not the final, finished device, e.g., when using an identical material to that used in another cleared device.  Information on Test Reports includes the same information as Attachment E of the Guidance.

While these resources may prove helpful to those new to biocompatibility, for those familiar with the Guidance, there is not likely much value other than the convenient format and it is not likely to help resolve the biocompatibility challenges faced by sponsors during premarket review.  It is nice to see informal e-mail mentioned as a way to get questions answered, but use of the Guidance and the pre-submission process will continue to be the best way to determine testing necessary to support a device’s biocompatibility before conducting studies.

On August 25th, 2020 Hyman, Phelps & McNamara, P.C. released a blog post covering experiences and lessons learned from our interactions with FDA and what was at the time the new and unfamiliar EUA pathway (FDA, Testing, and COVID-19: A “Mid-Mortem”). In the six months since that post, we have seen new products hit the market for non-laboratory use, new technologies, and what I assume is attributable in some small part to my offerings to Jupiter, Son of Saturn, an effective vaccine.

Since the early days of the pandemic, FDA has made it possible for manufacturers to market and distribute certain diagnostic tests to high-complexity CLIA Labs prior to EUA authorization provided that the validation testing is complete at the time of marketing and an EUA is filed with 15 business days. The “Policy for Coronavirus Disease-2019 Tests During the Public Health Emergency (Revised) Immediately in Effect Guidance for Clinical Laboratories, Commercial Manufacturers, and Food and Drug Administration Staff” Issued on May 11, 2020 (the “FDA Testing Policy”).

FDA does not intend to object to a commercial manufacturer’s development and distribution of SARS-CoV-2 test kits for specimen testing for a reasonable period of time, where the test has been validated and while the manufacturer is preparing its EUA request, where the manufacturer gives notification of validation to FDA as described below, and where the manufacturer provides instructions for use of the test and posts data about the test’s performance characteristics on the manufacturer’s website.

On the surface, this statement sounds fantastic.  FDA appeared to understand that the pandemic is an all-hands-on-deck situation and is trying to find a middle ground to get tests on the market while manufacturers play catch-up compiling their EUA submission. But, in the words of the many great home shopping spokespeople, “But Wait! There’s More!”  The FDA Testing Policy states:

FDA believes that 15 business days is a reasonable period of time to prepare an EUA submission for a test that has already been validated. Soon after receiving the EUA request, FDA will perform a preliminary review to identify if there are any problems with the performance data. If a problem is identified, FDA intends to work with the manufacturer to address the problem (e.g., through labeling or bench testing). If the problem is significant and cannot be addressed in a timely manner, and the manufacturer has already distributed the device, FDA would expect the manufacturer to suspend distribution and conduct a recall of the test.

Per this policy, FDA is affording companies at least 3 weeks to compile completed data into a functional EUA submission and submit while also committing to a triage review with a tiered approach to deficiencies.  Hundreds of companies took advantage of this program and notified FDA of their intent to distribute test kits in accordance with FDA’s policy. This seemed too good to be true, and it most certainly was.

Many companies viewed the notification pathway as a way to stand out from the crowd by proactively signaling to FDA that their submission was complete and that their EUA should garner additional attention from the Agency. For many companies the notification pathway was only about visibility on FDA’s radar. In our experience, many manufacturers never availed themselves of the opportunity to distribute as nearly all potential customers were seeking out tests that had successfully obtained EUA authorization.

Weeks in the EUA queue turned into months with little to no action being taken by FDA for many EUAs on the notification list.  When companies finally received the reviewer introduction email, many were lifted from the morass of uncertainty and despair and filled with new hope for the future.  This life-reaffirming high would be snuffed out quickly by the all too common subsequent “Inadequate Validation” email.

This email would inform companies that they had not provided complete information to demonstrate that the studies performed are adequate to validate the test and support the claimed performance characteristics of the device.  The email would go on to list the deficiencies in the familiar fashion of a hold letter for a 510(K), De Novo, or PMA.  As you’ll recall, the FDA Testing Policy stated that, “If a problem is identified, FDA intends to work with the manufacturer to address the problem (e.g., through labeling or bench testing).” Most in industry expected that this meant FDA would allow time to address concerns via new labeling,  bench testing, or simply the submission of additional data.  The veritable chasm between industry’s expectations as to how this process would play out and the reality espoused by FDA can be found in the close of the “Inadequate Validation” letter, which typically, allowed a sponsor a mere 24 or 48 hours to respond:

[I]f the information you provide does not demonstrate that your device is adequately validated for its intended use and performing consistently, we may take steps as described in the guidance referenced above, including removing the test from the website listing of notifications, and may determine that the criteria for emergency use authorization have not been met. If such steps are taken and we make that determination, we would expect you to suspend distribution of your test, and we may request that you take additional actions to protect the public health as appropriate.

One can imagine the feeling of despair when after months of waiting for a reviewer to be assigned to your EUA submission FDA returns with questions but only provides you with 24 or 48 hours to respond with the rationale that the timeline was in the “interest of public health.” Once on the notification list FDA has no way of checking to see if devices were actually introduced into the market, therefore, FDA approaches each review as if they have allowed the company to exist in the marketplace unchecked for months.  It seems that FDA is content with the idea that by limiting response windows to one or two days they are fulfilling their mandate to promote and protect public health.  Many companies are at a loss, however, as they are not prepared for this over-correction by FDA and simply cannot operate within such arbitrary timelines with draconian enforcement.

We have to ask the question, “Is FDA intending for these companies to fail?” By now FDA should be well aware of industry’s plight.  Companies are waiting months only to scramble for 48-hours straight to try and deliver FDA the moon.  It behooves the Agency to publicly issue new instructions clarifying their expectations for companies on the notification list.

In comparison, for a company that has submitted an EUA without being added to the notification list, FDA may not even specify a deadline to respond to the review team’s questions.  Furthermore, we have not observed an acceleration in the review timeline for companies on the notification list. It appears that you have everything to lose by being added to the notification list with nothing to gain.

Prof. Hubert J. Farnsworth said it best “I’m sciencing as fast as I can.”  Only so much can be done in the span of 48 hours and many times FDA requests additional data or analysis. Bottom line: Companies should think twice before choosing to add their tests to the notification list.

The FDA is legally established as a law enforcement agency.  But its structure and activities have also generated its own “branding” as a trusted independent validator of medical devices.

This public trust gives FDA tremendous power.  If a device has not undergone FDA’s premarket review (even if lawfully exempt), it often causes consternation among potential customers.  Likewise, if FDA issues negative public statements about the safety of marketed devices or the compliance status of a company, there can be a significant and lasting market impact.

The use of this power is not new for FDA.  Historically, FDA has made public a variety warning letters about violations available to the public (and it continues to do so).  It also publicizes adverse event reports and  recalls, market withdrawals and safety notices. In Class I recalls, the agency may issue its own press release.  FDA issues device safety communications.  FDA makes public some information on inspection results.

During the pandemic, however, CDRH has qualitatively increased its use of public lists and other information to leverage the FDA brand to protect the public health.

A few examples:

CDRH allows commercial manufacturers to distribute diagnostic test kits to perform assays to detect SARS-CoV-while the manufacturer is preparing its EUA request, provided that the manufacturer provides instructions for use of the test and posts data about the test’s performance characteristics on the manufacturer’s website.  CDRH is relying upon this “[t]ransparency to mitigate potential adverse impacts from a poorly designed test by facilitating better informed decisions by potential purchasers and users.”  The manufacturers and their tests are listed here, and the list expressly notes whether a test is “unauthorized” or “authorized.”

This approach has not been an unalloyed benefit to manufacturers.  Some customers are unwilling to purchase products based upon notification and insist on waiting for an EUA to issue.  That reflects the power of FDA’s brand.  But other customers are perfectly happy to do so.  For instance, some sophisticated clinical laboratories may feel that they have an adequate ability to validate the tests and do not need to rely on FDA’s validation.

As another example, CDRH maintains a list of tests that affirmatively should no longer be used or distributed here.

As a third example, CDRH issued enforcement discretion guidance allowing thermographic cameras to be commercially distributed for detecting fever during the pandemic.  Upon finding that some firms are marketing the cameras in violations of the guidelines, FDA has issued warning letters.  That much is not unusual (except for the demand for a response within 48 hours).  But FDA also has redeployed these technical and legalistic warning letters into a plain English warning against using the products:

FDA is advising consumers not to purchase or use certain products that have not been approved, cleared, or authorized by FDA and are being misleadingly represented as safe and/or effective for the mitigation, prevention, treatment, diagnosis, or cure of COVID-19. Your firm will be added to a published list on FDA’s website of firms and websites that have received warning letters from FDA concerning the sale or distribution of COVID-19 related products in violation of the Act. This list can be found at https://www.fda.gov/consumers/health-fraud-scams/fraudulent-coronavirus-disease-2019-covid-19-products. Once you have taken actions to address the sale of unapproved, uncleared, and unauthorized products . . . and any appropriate corrective actions have been confirmed by the FDA, the published list will be updated to indicate that your firm has taken such corrective actions.

Other examples:

SARS-CoV-2 Reference Panel Comparative Data | FDA (list of molecular tests and their performance against a reference panel).

EUA Authorized Serology Test Performance | FDA (list of serology tests and their expected performance).

Independent Evaluations of COVID-19 Serological Tests (Frederick National Laboratory for Cancer Research results of serology test performance – authorized and not authorized).

As a final example, in response to a great deal of (often deliberate) misinformation circulating in the public domain about the meaning of “FDA registered” and “FDA certified,” CDRH has published an explainer.

FDA took these steps in the crush of responding to the pandemic.  But they point to a broader potential after the pandemic.  We have gotten to the point where there has been wide and deep integration of internet-based research into business and consumer decision‑making.  It is now easy to quickly find FDA’s lists and other information they may put out.  For this reason, although the pandemic may have acted as a catalyst, FDA’s more aggressive use of publicity is likely to expand in the years to come.

Congress should keep an eye on how FDA evolves its use of publicity in order to ensure that this power is used responsibly.  Still, Congress will no doubt appreciate that this approach permits FDA to achieve a public health impact more quickly and at lower cost than might otherwise be the case.

Finally, the song title quoted in the title of this post is from Gilbert & Sullivan’s The Mikado.  The pertinent lyrics:

“I’ve got a little list.

I’ve got a little list.

Of society offenders

Who might well be underground

Who never would be missed – who never would be missed!”

On March 16th, 2021 CDRH announced a major policy change for the EUA program in an effort to expedite screening testing for the pandemic. Screening is the testing of asymptomatic individuals who do not have known or suspected exposure to COVID-19 in order to make individual decisions based on the test results.  This new supplemental template provides recommendations for device manufacturers with the goal:

to streamline the authorization of screening tests with serial testing. The recommendations apply to test developers who seek an EUA from the FDA for certain screening tests prior to conducting certain performance evaluations with asymptomatic individuals.

This is a major departure from previous policy as testing prior to authorization was only permissible in very limited cases for devices used only in High Complexity CLIA labs. It is important to note that this new pathway is only available to manufacturers of molecular and antigen tests.  Serology tests are not mentioned in this announcement.

Serial testing as defined by FDA is the “testing the same individual multiple times within a few days.”  The thinking behind this strategy is the testing of a single patient across multiple days would increase the chances of detecting infection even when using devices with lower sensitivity.  FDA’s expectations for a device’s Sensitivity/Specificity increase when you traverse across the continuum of indications.  We provide an example for antigen tests below (Table 1).

Table 1 – Performance Expectations by Indication in FDA’s EUA Templates (Antigen)

The key section of this announcement is where FDA identifies what indications may be eligible for this new program:

For example, in certain circumstances, a [point-of-care] test or an at-home test could be authorized for over-the-counter (OTC) use without the need for validating its use in asymptomatic individuals prior to authorization.

This statement has the potential to be both very powerful and very frustrating for industry.  In Table 2 below there is a breakdown of the data expectations for Point-of-Care, Prescription Home Use, and OTC Home Use with an asymptomatic claim.  What you can see from the information, in the various EUA templates laid-out in this fashion, is that ‘Prescription Home Use’ and ‘OTC Home Use’ are very similar in their requirements.  It is, therefore, reasonable to expect FDA to blur the line between ‘Prescription Home Use’ and ‘OTC Home Use’ when considering asymptomatic testing.  It was very surprising, however, to read that this new policy may also apply to devices validated for Point-of-Care.  The differences in data requirements between ‘Point-of-Care’ and ‘OTC Home Use’ are substantial.  Devices authorized for Point-of-Care are not validated to the same degree with respect to Robustness (Flex Studies), Human Factors (Human Usability), the data required to substantiate performance for Point-of-Care (POC) may have been largely retrospective testing.  This difference is due, in part, to the intended users – tests intended for POC are typically used by healthcare providers whereas home tests are used by lay people.

The inclusion of Point-of-Care in this new policy implies that a manufacturer can take their authorized test, supplement the EUA with the additional information describing serial testing and obtain an OTC Home-Use indication. As of March 16, there are twenty (20) molecular EUAs and one (1) Antigen EUA with an attribute of ‘screening’ according to FDA’s website.  Of these tests, eight (8) are Direct-to-Consumer (DTC) and two (2) are Over-the-Counter (OTC). To date, there are no Point-of-Care EUAs with a screening attribute.

Table 2 – Tests Recommended by FDA Stratified by Indication

Supplemental Template for Developers of Molecular and Antigen Diagnostic COVID-19 Tests for Screening with Serial Testing (Example Template)

The administrative section at the top of this new template echoes the announcement indicating that this policy applies to Point-of-Care indications:

This template is intended to provide supplemental recommendations for developers of molecular and antigen tests seeking claims for screening with serial testing without studying asymptomatic individuals prior to authorization, including for point-of-care (POC) and at-home tests.

While the intention of the policy is to increase testing availability FDA does not want manufacturers to misconstrue this as a paradigm shift with an opportunity to resurrect denied EUAs from the grave.

These recommendations will generally not be applicable to developers with tests for which data has already demonstrated poor performance (e.g., less than 80% PPA) for testing asymptomatic individuals.

FDA later goes on to seemingly contradict themselves on the next page of the template by stating:

As discussed in the Antigen Template for Test Developers, strategies for serial testing with less sensitive tests (i.e., PPA <80%) may be able to be support authorization

FDA gives the following example as a modification to the intended use:

…individuals without symptoms or other epidemiological reasons to suspect COVID-19 infection, when tested twice over two (or three) days with at least 24 hours (and no more than 36 hours) between tests.

As written, this implies that there is no expected modification to the intended use setting or user as a result of this policy change.

The template does indicate that FDA intends to push as much of the clinical validation as possible to a post-marketing commitment in order to expedite the authorization of new tests.  While this recommended study size is small, 20 asymptomatic positives individuals, it has been increasingly more difficult to find asymptomatic positives naturally within the population, and that challenge is likely to continue to grow.  It is not clear from this template how long FDA will give these companies to compete these post-authorization studies.

The template closes with a final recommendation for labeling:

Proposed labeling should clearly identify the population in which the test’s performance has been validated, and clearly identify any populations included in the intended use for which the test’s performance has not yet been established and will be established during the above referenced post-authorization study.

This final section further confuses what is expected of manufacturers as it implies that the post-authorization study is the only data requirement to expand the EUA claim to a new patient population.

This announcement from FDA generates a rollercoaster of emotion starting with cautious optimism, to elation, and finishing with confusion. This policy raises major administrative questions that need to be answered in full if this new program is to have a chance at success.

Can a manufacturer take an existing Point-of-Care or Rx-Only EUA and convert it to an OTC Authorization, if they commit to conducting an additional clinical study post-authorization?

It is our review of the policy that, as written, it appears that the answer is yes. However, the policy never mentions the non-clinical tests that FDA typically requires for Home-Use tests and whether those will also need to be performed, either pre- or post-market, to support the expanded indication Point-of-Care authorized tests.  Without additional clarifications on the part of FDA, manufacturers will be left scratching their heads and unsure as to what FDA expects in EUAs moving forward.

The next “FDA Virtual Townhall” is Wednesday March 24^th at 12:15 pm (ET).  See you there.

On Friday, March 12th FDA posted a letter to healthcare providers about performance concerns regarding the Roche Molecular Systems, Inc. cobas SARS-CoV-2 & Influenza Test for use on cobas Liat System.  This public letter appears to be the culmination of a dialogue between Roche and FDA.  Roche’s root cause analysis investigation has identified two potential causes for the false positives:

• Roche identified that the assay tubes may sporadically leak, causing an obstructed optical path in the Liat analyzer, producing abnormal PCR growth curves. This could lead to invalid or erroneous positive results, particularly for the Flu B test. If a tube leak occurs, later testing runs may have an increased likelihood of false positive Flu B results.

• Roche determined that abnormal PCR cycling in the reaction tubes may also produce abnormal PCR growth curves, leading to erroneous results. The issue is sporadic and may be caused by multiple factors happening at the same time, such as hardware positioning, volume movement, and curve interpretation. This issue may cause false positive results for multiple analytes (Influenza A, Influenza B and/or SARS-CoV-2) in a single testing run.

In response, FDA recommended three actions by users:

1. Monitor for unexpected clusters of positive Flu B results, as this may indicate the cobas Liat System has experienced a tube leak.
2. Repeat tests when two or three analytes are positive. Different results on the repeat test may indicate abnormal PCR cycling.
3. Stop using the cobas Liat System and contact Roche if you suspect either of these two issues has occurred.

This letter is not the first FDA has posted for Clinical Laboratory and Point-of-Care staff.  It is actually the fourth one since October 2020 and the eighth such notice flagging in‑vitro diagnostic performance issues or concerns during the pandemic.

These notices have hit each of the three major classes of in-vitro diagnostic products used for the nation’s pandemic response, PCR, Serology, and Antigen Tests.  Here are three relatively recent examples:

PCR

• Genetic Variants of SARS-CoV-2 May Lead to False Negative Results with Molecular Tests for Detection of SARS-CoV-2 – Letter to Clinical Laboratory Staff and Health Care Providers (Posted 01/08/2021)

Antigen

• Potential for False Positive Results with Antigen Tests for Rapid Detection of SARS-CoV-2 – Letter to Clinical Laboratory Staff and Health Care Providers (Posted 11/03/2020)

Serology

• Important Information on the Use of Serological (Antibody) Tests for COVID-19 – Letter to Health Care Providers (Posted 04/17/2020)

The full list of FDA’s “Letters to Health Care Providers” can be found here.

As we proceed through this pandemic, it is becoming more critical that manufacturers be as vigilant as ever with post-market surveillance to flag performance issues early in order to be able to investigate and identify potential root causes.  Prompt investigation and correction is always preferable to pulling a test that works out of the marketplace.  It is in no one’s best interest to lose testing capacity and we hope that FDA is affording companies of all sizes the same opportunity that Roche has been given to investigate and correct rather than be forced to pull an assay from distribution.