By Riëtte van Laack –

On August 17, FDA published in the Federal Register the long awaited final rule, “Substances Generally Recognized as Safe.” The final rule is based on the proposed rule from 1997, and a reopening of the comment period with supplemental questions in 2010. . The final rule includes a number of tables that provide an overview of the differences between the proposed rule and the final rule. 

The final rule does two main things: 1) it clarifies the criteria for determining when the use of a substance in a food for humans or animals is “generally recognized as safe” (GRAS) for its intended use and therefore exempt from the food additive definition in the Federal Food, Drug, and Cosmetic Act (FDC Act), and 2) officially replaces the voluntary GRAS affirmation petition (GRASP) process with a voluntary GRAS notification (GRN) procedure.  FDA’s Center for Food Safety and Applied Nutrition (CFSAN) implemented the GRN process soon after publication of the 1997 GRN proposed rule, operating it successfully for nearly two decades, whereas FDA’s Center for Veterinary Medicine (CVM) did not establish a pilot process for GRAS notices for animal feed ingredients (AGRNs) until 2010 (see our previous post here).  As of Dec. 31, 2015, CFSAN had filed 614 GRNs for use of human food ingredients and CVM had filed a mere 18 AGRNs for use of animal food ingredients. 

A major driver for FDA’s shift from GRASPs which involve notice-and comment rulemaking to GRNs and AGRNs is the ability to respond in a timely manner. Under the final rule, FDA commits to respond to the notifier within 180 days of the date of filing the GRN or AGRN, with a possible extension by another 90 days.  Thus, the notifier may not receive a substantive response until 270 days after filing.  For purposes of comparison: during the 10-year period from 1990 through 1999, CFSAN completed the rulemaking process for 24 GRASPs, with an average elapsed time of approximately 7.9 years (median elapsed time of approximately 6.9 years).

In recent years, the integrity of GRAS determinations and FDA’s reliance on the voluntary GRN process has been under attack.  Critics allege that, among other things, GRAS determinations employ outdated science, are rife with conflicts of interest, and the data supporting GRAS determinations are secret.  As we previously reported, in October, 2014 the International Society for Regulatory Toxicology and Pharmacology (ISRTP) held a Workshop on GRAS determinations. The ISRTP Workshop provided a forum for experts in the food safety scientific community to respond to criticisms of the FDA’s GRAS processes.  The manuscripts from the Workshop are now available in a special GRAS Supplement of Regulatory Toxicology and Pharmacology Journal (2016), including an Overview of the Workshop co-authored by Diane McColl, Past President of ISRTP.  FDA’s 97-page final rule also addresses many of these critiques and explains the law and public nature of GRAS conclusions.

A main thread through the preamble is FDA’s emphasis on the public availability of information supporting GRAS status. With respect to GRAS status based on scientific procedures – the principal basis for GRAS conclusions in the modern era – the critical difference between a GRAS use and an approved food additive use is that the conclusion of GRAS status (FDA’s new terminology for what used to a GRAS determination)  must be based on safety information that are generally available and accepted.  As explained in the preamble, “[a]lthough general recognition of safety through scientific procedures may be corroborated by the application of unpublished scientific data, information, or methods, . . . to satisfy GRAS criteria, qualified experts must be able to conclude the substance is not harmful under the conditions of intended use without access to 'corroborative' information.”  As FDA further explains, there can be no basis for a conclusion of GRAS status if trade secret information (or other non-public information) is necessary for qualified experts to reach a conclusion that the notified substance is safe under the conditions of its intended use.  Also, if the public description of the method of manufacture that a notifier includes in a GRAS notice does not provide sufficient detail to evaluate the safety of the notified substance as manufactured, there would be inadequate basis to support a conclusion of GRAS status.

FDA acknowledges that there are differences between the “evaluation” of a GRAS notice and the “review” of a GRASP. Among other things, the data and information in a GRN are summary data and information whereas GRASPs would include the underlying data from studies described in the petition.  This does not mean, however, that FDA’s evaluation of a GRAS notice does not constitute a substantive evaluation, or that the safety standard for a GRAS notice is different from that of a GRAS affirmation.  Because of questions and potential confusion resulting from FDA’s standard language in a “no questions letter” (i.e.,“The Agency has not . . .  made its own determination regarding the GRAS status of the subject use of the notified substance”), FDA will use different language in the future, i.e., “the Agency has not affirmed the GRAS status of the notified substance under the conditions of its intended use in accordance with 21 CFR 170.35.”

In terms of data quality and quantity to support a conclusion of GRAS status, the final GRN rule does not impose any substantially different data requirements than did the GRN process operated under the proposed rule. However, the final regulations include more specific descriptions of what must be included in a GRAS notice. 

A GRN must include seven parts. For parts that may not be applicable (e.g., self-limiting levels of use (part 4), experience based on common use in food (part 5)), the GRN must include an explanation for the “omission.”  Under the new regulation, a GRN must include a certification that the GRN is “complete” in addition to “representative” and “balanced.”  FDA clarifies that it is the notifier’s responsibility to identify, discuss, and place in context, data and information that are, or may appear to be, inconsistent with a conclusion of GRAS status.  Such data need not be publicly available. 

Although the decision to submit a GRN is voluntary, the information included in a GRN is mandatory. The final rule stipulates that the data and information in a GRN are considered a mandatory, rather than voluntary, submission for purposes of its status under the FOIA and 21 C.F.R. Part 20.

FDA will make a GRN available for public disclosure immediately upon receipt, and will make certain information pertaining to filed GRAS notices readily accessible to the public so as to provide an opportunity for outside parties to make FDA aware of dissenting views about whether the available data and information support a conclusion that the notified substance is GRAS under the conditions of its intended use.

FDA’s final regulations do not specify the data and information that the Food Safety and Inspection Service (FSIS) of the U.S. Department of Agriculture will need to evaluate whether the intended use of the notified substance complies with applicable statutes and regulations enforced by FSIS.

A change from the past is that FDA no longer requires that a GRN provide the common or usual name for the notified substance. Instead, the GRN must provide an appropriately descriptive term.  What constitutes a common or usual name of the substance is not considered in the evaluation of the GRN.  

FDA’s final rule does not address potential conflicts of interest of an expert panel. FDA explains that an expert panel, often called a GRAS panel, is not mandatory.  “Convening a GRAS panel has historically been a way to provide evidence that generally available data and information are generally accepted by the expert scientific community, but convening a GRAS panel is not the only way to provide such evidence.”  Therefore, the issues of potential conflict of interest and ways to avoid a conflict of interest are not addressed by the final rule.  FDA plans to address these issues in guidance.

The regulations concerning AGRNs for animal food ingredients largely track the regulations for human food ingredients.

The final rule becomes effective on October 17, 2016.  As of the effective date of the final rule, FDA will close the docket for any pending GRASPs. There are 45 pending GRASPs.  The Agency plans to contact the affected petitioners and provide them with the option to submit a GRN that incorporates the GRASP.

As we previously reported, in 2014, FDA and the Center for Food Safety (CFS) entered into a consent decree that required FDA to issue the final rule by August 31, 2016. CFS reserved the right to “challenge . . . the merits of the final rule.”  As we indicated at that time, CFS’s posture suggests that FDA’s issuance of a final rule could beget additional litigation.  Litigation remains likely (see here). 

By Jennifer D. Newberger –

Perhaps the most interesting thing about the recently released draft guidance, Deciding When to Submit a 510(k) for a Software Change to an Existing Device, is its mere existence. Historically, all changes to 510(k)-cleared devices were analyzed under the same guidance, regardless of whether the change related to software, labeling, materials, or anything else. That FDA has now carved out software changes indicates the importance that software has come to play in a wide variety of medical devices, and the complexities associated with software modifications.

Much of the draft software guidance mirrors a companion guidance issued the same day, Deciding When to Submit a 510(k) for a Change to an Existing Device, which addresses non-software modifications, and on which we previously blogged here. For example, the emphasis on whether a change is “intended” to significantly affect the safety or effectiveness of a device, and the use of risk management to identify potential hazards associated with the proposed modification, feature prominently in both documents.

In order to encourage the security of medical devices, the draft guidance begins by explicitly stating that if the change is done solely to strengthen cybersecurity without otherwise impacting the software or device, then no 510(k) is needed. Given the recent allegations involving pacemakers and defibrillators, this text may be reviewed more closely than it would have been just a few weeks ago.

The draft also addresses the issue of “bug fixes.” Though the draft states that such fixes are to be considered design changes under 21 C.F.R. Part 820, it also states that “[w]hen a change to the software only restores the device to the specifications of the most recently cleared device, then a new 510(k) is likely not required.”

The draft also separates out whether the software modification creates a new cause of a hazardous situation versus creates a hazardous situation itself, or alters an existing hazardous situation. It is not clear how a software modification could create a hazardous situation without creating a cause of a hazardous situation, or vice versa.

As with the companion draft guidance, this draft guidance regarding software modifications includes a number of detailed examples regarding when a new 510(k) would be needed, and should be helpful to industry in assessing software modifications. As is the case with the other draft guidance, companies should review carefully, since seemingly subtle changes could have major impacts (see our previous post here).

By Riëtte van Laack –

On August 24, 2016, the Food Safety Inspection Service (FSIS) of the U.S. Department of Agriculture announced the availability of its compliance guidance regarding “Statements That Bioengineered or Genetically Modified (GM) Ingredients or Animal Feed Were Not Used in the Production of Meat, Poultry, or Egg Products.”  Although the title may suggest otherwise, the main change in FSIS policy is that the Agency will now allow the terms “GMO” and “genetically modified organism” when making claims for non-use of genetically modified or bioengineered ingredients (FSIS refers to such claims as negative claims).  Before the issuance of this guidance, FSIS did not allow these terms in the labeling for such products, except if the name of the third-party certifying organization contained these terms (e.g., “Non-GMO project”). 

As we reported previously, FDA also had indicated that it frowns on the term “GMO” because, according to FDA, it is scientifically inaccurate. However, FDA could not prevent such use.  Because FSIS, unlike FDA, approves labels for meat, poultry and processed egg products, FSIS could actually prevent the use of the term “GMO” on those products.

The recently enacted National Bioengineered Food Disclosure Standard includes a statement essentially defining non-GMO as not-bioengineered: “‘not bioengineered,’ ‘non-GMO,’ or any other similar claim describing the absence of bioengineering.”  Therefore, FSIS has decided that it no longer will object to use of the term GMO and, going forward, will allow the use of the term non-GMO in negative claims regarding the absence of genetically modified ingredients or animal feed.

The guidance is effective immediately. FSIS will begin approving negative claims that contain the terms “genetically modified organism” or “GMO” for meat, poultry and egg products that do not contain bioengineered ingredients or that are derived from animals that do not consume bioengineered feed.  In evaluating such claims, FSIS will use the definition of “bioengineering” from the new law, i.e., a food (A) that contains genetic material that has been modified through in vitro recombinant deoxyribonucleic acid (DNA) techniques; and (B) for which the modification could not otherwise be obtained through conventional breeding or found in nature.”  The requirements for a third party certification that the product is indeed non-GMO and related requirements (including the requirement to identify the third party’s website on the label and in labeling) remain the same.  For products that qualify for an “organic” claim under the National Organic Program, establishments need not provide FSIS with additional documentation for approval of negative claims.  In accordance with the new law, FSIS will not require additional documentation for non-GMO claims in labeling of products that qualify for an organic claim under the National Organic Program.

Comments on the guidance must be received by October 24, 2016.  AMS is responsible for the implementation of the National Bioengineered Food Disclosure Standard. So far AMS has established a webpage for tracking progress on its implementation of the new law. 

By Jennifer D. Newberger –

The long-awaited update to the 1997 guidance, Deciding When to Submit a 510(k) for a Change to an Existing Device, was released in draft on August 8. In contrast to its 39-page predecessor, this 73-page document is impressive in the level of detail provided, the much improved flow charts and more extensive examples, and discussion of documentation that should be retained if a manufacturer elects not to file a new 510(k). Whether the ultimate outcome of this draft will be a reduction or increase in the circumstances in which a 510(k) must be submitted for a modification to a cleared device is not entirely clear, but at least this draft guidance provides additional detail that should be useful to manufacturers making that determination.

The draft focuses on conducting risk assessments of the modified device to determine whether a new 510(k) is required, stating, “the assessment of risk in deciding whether to submit a new 510(k) should identify all possible risks, and then focus on risks whose existence and characteristics are supported by objective scientific evidence. It is not necessary to focus on hypothetical risks that are not supported by scientific evidence or those that are determined to be negligible due to both the low probability of occurrence and low severity of harm.” Though it may seem obvious that it is not necessary to consider “hypothetical” risks, this language indicates that manufacturers need not analyze all potential risks that may occur from use of the modified device, but only those “supported by objective scientific evidence.” This sounds like a fairly high bar, and should give manufacturers some level of comfort when contemplating potential risks. How this would be applied in practice is, of course, another question. For example, there may not be agreement what constitutes “low probability” or “low severity.”

This emphasis on risk assessment will cause many companies to rethink how they analyze changes. Under 21 C.F.R. § 807.81(a)(3), the issue is whether a change could significantly affect safety or effectiveness. While there is a linkage between that question and risk assessments, the two approaches are not identical.

The draft guidance also includes a discussion of how to properly document a modification to a 510(k)-cleared device, as well as a sample “Regulatory Change Assessment,” more commonly known in industry as a “Letter to File.” The inclusion of this information should prove useful to manufacturers in preparing these assessments. The draft guidance does make clear that simple “yes” or “no” answers will not suffice. Nor will it be enough just to complete a flowchart, a process many companies follow today.  

As for how to determine whether a modification requires a new 510(k), the draft guidance states that “the first question is always whether the change is being made with the intent to significantly improve the safety or effectiveness of the device, for example, in response to a known risk, adverse event, etc. If so, then the change likely ‘could significantly affect the safety or effectiveness’ and a new 510(k) likely must be submitted.” The language about the “first question” implies that this has always been the standard for determining whether a change requires a new 510(k). This is notable, because of course this standard is being introduced for the first time here. This language is not in the regulation, or the 1997 guidance.

The reference to “intent” is new. To the extent that the draft guidance interjects a subjective element – focusing on the manufacturer’s state of mind rather than purely objective factors – it introduces a new factor.

If FDA wishes to take the approach expressed above, further explanation will be needed for how this should work in practice. According to the draft guidance, a manufacturer who learns of a safety issue associated with a product and makes modifications to reduce that risk must submit a 510(k) for that modification. Of course, clearance for that improved, safer version will take at least 90 days, and probably more. So in the interim, is the manufacturer to continue distributing the less safe product, when it knows, through testing, that it has a safer product ready to be marketed?

This quandary would seem particularly troublesome if the marketed product is not per se unsafe, and is performing according to its specifications and operating in conformance with the 510(k) clearance. Manufacturers may be dissuaded from proactively seeking to improve the device since, under the plain language of the guidance, such improvement requires a 510(k), regardless of the results of the validation and verification testing.

The revised guidance is emphatic that a 510(k) is required if the change improves safety or effectiveness. That approach is consistent with the regulation, which refers to a “change” in safety or effectiveness, without regard to direction. Yet it is easy to imagine scenarios where companies will forego making modifications under the guidance because a new 510(k) would be needed. It may well be that the guidance will deter modest incremental innovations by device manufacturers if, for example, a risk assessment shows a modest but identifiable reduction in risk.

Second, unlike the 1997 guidance, this draft explicitly addresses the ever-complex area of a change from a general to more specific use—although not necessarily in a way that provides more certain direction to manufacturers. The draft states that “[m]anufacturers should carefully consider the potential effects on their device’s risk profile in making [a change from a general to specific indication], as they are among the most difficult to assess. If a change of this type has the potential to expand device use to different users, different use environments, use in or on a different type of joint, organ, bone, vasculature, or tissue, use in different patient populations, or new therapeutic or diagnostic uses, it should be evaluated using the guidance provided above.”

It is interesting that the above is framed in terms of “expanding” device use, when the issue often relates to “narrowing” the use—going from a general (more expansive) use to a specific (narrower) use. If FDA considers “narrowing” an indication to be an “expansion,” then there is likely to remain a fundamental disagreement between industry and FDA. FDA is clearly right about one aspect: the general-to-specific changes will remain “among the most difficult to assess.”

FDA’s choice of the word “different” is also worth noting. In common understanding, something that is “different” is something that is not similar to the original object. Promotion of a device for a more specific use would not generally be considered promotion for a “different” use, since the use is incorporated in the general clearance. For example, as we discussed in a prior blog post, in the Vascular Solutions case, the government alleged that the company’s “Vari-Lase” product line could only be promoted for the treatment of superficial veins, and that the company’s promotion for the ablation of “perforator” veins, which connect the superficial vein system to the deep vein system, was outside the scope of the cleared indication. The government lost, at least in part due to the fact that the jury found the specific use was within the scope of the general clearance, i.e., it was not a use for a “different” type of vasculature. It is unlikely that the discussion provided in the draft guidance will prove useful in assessing when a new 510(k) will be needed for a more specific use.

As noted, there are still areas that will require additional consideration before the guidance should be finalized. Nevertheless, this draft guidance appears to be an improvement with respect to the number of examples, flow charts, and more detailed explanations by FDA. At the same time, other elements deserve careful scrutiny, such as the use of subjective intent, the role of risk assessment, and the recurring general vs. specific issue, as well as FDA’s revamped expectations for documentation.

By Kurt R. Karst –      

Orange Book aficionados may have noticed an interesting entry that recently appeared as an addition to the Orange Book. It came out earlier this month with FDA’s publication of the July 2016 Orange Book Cumulative Supplement (page 1-2C), and also with an update to the electronic Orange Book.  Here it is on both formats:

Active Ingredient: CHLORDIAZEPOXIDE HYDROCHLORIDE; CLIDINIUM BROMIDE
Proprietary Name: LIBRAX
Dosage Form; Route of Administration: CAPSULE; ORAL
Strength: 5MG;2.5MG
Reference Listed Drug: Yes
TE Code: N/A
Application Number: N012750
Product Number: 001
Approval Date: Approved Prior to Jan. 1, 1982
Applicant Holder Full Name: VALEANT PHARMACEUTICALS NORTH AMERICA LLC
Marketing Status:  Prescription

That’s right! LIBRAX, approved under NDA 012750, is now officially listed in the Orange Book.  Of course, LIBRAX has been identified by name in the Orange Book for quite some time – since the 14th edition (1994) – but only in the second line of the Orange Book Preface: “Drugs on the market approved only on the basis of safety (covered by the ongoing Drug Efficacy Study Implementation [DESI] review [e.g.,Donnatal® Tablets and Librax® Capsules] or pre-1938 drugs [e.g., Phenobarbital Tablets]) are not included in this publication.” And it was just a few months ago that FDA identified LIBRAX as a drug product subject to a pending DESI proceeding in correspondence to Congress.  But change has been afoot, as we intimated in a post a couple of months ago with a link to Docket No. FDA-1975-N-0336.

For decades now, FDA has treated LIBRAX as a “DESI drug.”  After all, NDA 012750 was approved prior to the enactment of the 1962 Kefauver-Harris Drug Amendments . . . right? “No,” according to FDA.  In a bit of revisionist history, FDA now believes that the LIBRAX NDA 012750 was approved on September 1, 1966, instead of pre-1962.

By way of background, FDA initially failed to include LIBRAX under the DESI program. In 1975, FDA became aware that LIBRAX had not been included in the DESI review and published a Federal Register notice (40 Fed. Reg. 52,644, 52,645-46 (Nov. 11, 1975)) requesting that the NDA holder provide substantial evidence of effectiveness for LIBRAX (DESI 10837).  FDA explained that:

On September 1, 1966, the [NDA] was reinstated after [the NDA holder] submitted new data including new test procedures to detect the amount of impurities.  However, since this reinstatement approval was not based upon a complete review of the entire application and did not constitute a determination that all claimed indications are supported by substantial evidence of effectiveness, exclusion of Librax from NAS-NRC review was in appropriate [sic].  The clinical data included in the [NDA] have now been reviewed by the [FDA] and it has been concluded that the data do not provide substantial evidence of effectiveness of the fixed combination . . . .  There is therefore no substantial evidence that the addition of chlordiazepoxide to clidinium bromide contributes to the effectiveness of the latter in the adjunctive therapy of peptic ulcer disease.

In January 1981, FDA issued a Notice of Opportunity for Hearing to withdraw approval of NDA 012750 on the basis that the data submitted to FDA were inadequate and did not provide substantial evidence that LIBRAX was effective for its intended uses.  FDA revoked the temporary exemption allowing LIBRAX to be marketed while the NDA was under review at FDA.  In July 1981, FDA, in response to a request from the NDA holder for a hearing on LIBRAX, stated that LIBRAX may continue to be marketed while the hearing request is under review at FDA pending a final ruling on that request.

On July 24, 2012, FDA published a notice in the Federal Register (77 Fed. Reg. 43,337) asking companies with outstanding hearing requests for pre-1962 DESI drugs to either affirm their hearing requests or withdraw them.  LIBRAX was one of the drugs discussed in the notice.  The NDA holder responded to the Federal Register notice on August 22, 2012 affirming its hearing request. 

And then things went quiet . . . . until May 23, 2016, when FDA posted a rather cryptic memorandum in Docket No. FDA-1975-N-0336 for the LIBRAX DESI proceeding (DESI 10837). According to the 1-page memo:

As set forth in the Stipulation for Dismissal in Hoffmann-La Roche, Inc. v. Richardson, et. al, Civil Action 11-73 (D.N.J. August 2, 1973), “[a] new drug application for Librax was approved by the Food and Drug Administration on September 1, 1966.  At that time, pursuant to the 1962 New Drug Amendments, the Food and Drug Administration determined that Librax was safe and effective for the indications set forth in its labeling.”  As such, Librax is not subject to review under DESI.

A few days later, our firm, on behalf of a client, sent a letter to FDA requesting that the Agency withdraw the May 23rd document and not take any action based on it. “The May 23 document is factually and legally wrong,” our firm explained in the letter.  But in an August 18, 2016 response to our firm’s May 27, 2016 letter, FDA doubles down on the Agency’s newfound position. 

At this juncture, FDA has determined that [the] Stipulation for Dismissal [in Hoffmann-La Roche, Inc. v. Richardson, et. al] is the controlling legal document on the question of whether Librax is subject to DESI review. In accordance with this Stipulation for Dismissal, FDA has concluded that Librax is not subject to review under DESI, as the drug was fully approved on the basis of safety and effectiveness in 1966. . . .

[W]e recognize that there were conflicting agency statements concerning the regulatory status of Librax. Nonetheless, FDA has concluded that it must look to the August 1973 Stipulation for Dismissal of the Roche lawsuit on the question of whether Librax is subject to review under DESI, as the Roche lawsuit was filed in January 1973 over this precise question. In our assessment, FDA is bound by the Stipulation for Dismissal, which was signed on behalf of, among others, the Commissioner of Food and Drugs, and which explicitly states that Librax was fully approved on the basis of safety and effectiveness in September 1966.

Right around the time the August 18, 2016 response from FDA was sent, the Orange Book was updated with the “NEWA” (“New drug product approval usually in the supplement month”) addition for LIBRAX. If FDA’s dubious determination stands, then that’s one less pending DESI review for the Agency to deal with. 

By John A. Gilbert, Jr. & Larry K. Houck –

We previously summarized (here and here) the Drug Enforcement Administration’s (“DEA’s”) recent actions related to marijuana and reviewed DEA’s notice that it will expand the number of marijuana cultivators for research. Today we analyze DEA’s denial of several recent petitions to reschedule marijuana.

On August 12, 2016, DEA issued a final rule denying two petitions to reschedule marijuana. As discussed in more detail below, while both the Department of Health and Human Services (“HHS”) and DEA raised concerns about the abuse potential of marijuana, the critical basis for this decision remains the same, neither HHS nor DEA found that marijuana has a currently accepted medical use in the United States. It is worth noting that a growing majority of states have passed legislation authorizing use of marijuana for medicinal purposes or limited use of low-THC oil (e.g., cannabidiol or “CBD”) for medical purposes. In addition, several jurisdictions, including Connecticut, the District of Columbia, Iowa and Oregon, have rescheduled marijuana for medical use. This is in addition to several states and the District of Columbia which have authorized some use of marijuana for recreational use. Thus, the issue remains under either state or federal law as to a finding of currently accepted medical use in the United States.

DEA received a petition from Bryan Krumm, a private citizen in New Mexico in 2009 (“Krumm Petition”), and the second from the then-governors of Rhode Island and Washington in 2011 (“Governor’s Petition”).  Denial of Petition To Initiate Proceedings to Reschedule Marijuana, 81 Fed. Reg. 53,767 (Aug. 12, 2016); Denial of Petition To Initiate Proceedings to Reschedule Marijuana, 81 Fed. Reg. 53,688 (Aug. 12, 2016) [hereinafter Denial of Petitions].

The Krumm Petition requested DEA to remove marijuana from schedule I of the Controlled Substances Act (“CSA”), asserting that:

1. Marijuana has accepted medical use in the United States;
2. Studies have shown that smoked marijuana has proven safety and efficacy;
3. Marijuana is safe for use under medical supervision; and
4. Marijuana does not have the abuse potential for placement in schedule. Id. at 53,767.

The Governor’s Petition requested removal of marijuana and “related items” from schedule I and reschedule as “medical cannabis” in schedule II because cannabis has an accepted medical use in the United States; is safe for use under medical supervision; and for medical purposes has a relatively low potential for abuse, especially in comparison with other schedule II drugs. Id. at 53,688. Neither petition appeared to provide adequate detailed scientific or medical support for the fact that marijuana has an acceptable medical use in the United States. Rather both made anecdotal references to use of marijuana in certain types of treatment.

As a threshold matter, DEA asserted that marijuana or cannabis could only be rescheduled to schedule II in order for the U.S. to comply with its obligations under international treaties. Next, DEA explained that while schedule I drugs have “no currently accepted medical use in treatment in the United States” and “a lack of accepted safety for use of the drug . . . under medical supervision,” schedule II drugs have “currently accepted medical use in treatment in the United States.” Id. at 53,689, 53,768. Thus, DEA found that any action to reschedule marijuana turns on whether marijuana has a currently accepted medical use in treatment in the U.S. Id.

As we have noted before, the federal CSA requires that DEA make certain findings before initiating any action to reschedule a drug. Placement of a drug in one of the four schedules (II-V) where there has been a finding of accepted medical use relative to its potential for abuse and potential for psychological or physical dependence. Moreover, prior to initiating rescheduling proceedings, 21 U.S.C. § 811(b) requires DEA to seek an opinion from HHS and determine whether the “scheduling recommendation, scientific and medical evaluation, and ‘all other relevant data’ constitute substantial evidence that the drug should be rescheduled as proposed.” Id. at 53,739, 53,820. HHS’ recommendations are binding on DEA as to scientific and medical matters. 21 U.S.C. § 811(b); Denial of Petitions at 53,739, 53,820. This analysis is commonly known as the “eight factor analysis.” The eight factors include:

1. The drug’s actual or relative potential for abuse;
2. The drug’s scientific evidence of its pharmacologic effect, if known;
3. The state of current scientific knowledge regarding the drug;
4. The drug’s history and current pattern of abuse;
5. The drug’s scope, duration, and significance of abuse;
6. The risk, if any, to public health;
7. The drug’s psychic or physiological dependence liability and
8. Whether the drug is an immediate precursor of a controlled substance. 21 U.S.C. § 811(c).

DEA requested a scientific and medical evaluation and scheduling recommendation from HHS as required. Denial of Petitions at 53,739, 53,820. HHS concluded, upon considering the eight factors, that marijuana meets the scheduling criteria for remaining in schedule I. Denial of Petitions at 53,688, 53,767. Most important, HHS found that marijuana does not have a currently accepted medical use in the United States. This scientific and medical finding is binding on DEA. Thus, while DEA also made specific findings relative to these factors, in effect, DEA’s hands are tied in that without a finding from HHS that marijuana does have an accepted medical use, DEA will have to deny any petition to reschedule the drug.

It is worth taking a closer look at the factors that HHS and DEA believe still indicate that marijuana does not have a currently accepted medical use in the United States.

Both DEA and HHS assert that a drug is considered to have a currently accepted medical use in treatment in the United States for purposes of the CSA if it is the subject of an approved new drug application (“NDA”) under the federal Food, Drug and Cosmetic Act (21 U.S.C. § 355). Denial of Petitions at 53,740, 53,821. FDA has not approved an NDA for marijuana for any indication. Id. In the alternative, DEA established a five part test in a 1992 marijuana rescheduling petition denial to determine whether a drug has a currently accepted medical use in the U.S. Id; Marijuana Scheduling Petition; Denial of Petition; Remand, 57 Fed. Reg. 10,499, 10,504-06 (Mar. 26, 1992). Under the DEA test, a drug is considered to have a currently accepted medical use only if it meets all five elements. Denial of Petitions at 53,740, 53,821. Those elements are:

1. The drug’s chemistry is known and reproducible;
2. There are adequate safety studies;
3. There are adequate and well-controlled studies proving efficacy;
4. The drug is accepted by qualified experts; and
5. The scientific evidence is widely available.

In the denying the petitions, DEA reviewed these five factors and found that marijuana did not meet this alternative test establishing a currently accepted medical use in the United States. DEA found that under Factor 1, chemical constituents including tetrahydrocabinnol (“THC”) and other cannabinoids vary significantly in different marijuana strains “chemical composition among different marijuana samples is not reproducible.” Id. at 53,761, 53,840. DEA notes, however, that chemistry may be consistent enough to derive standardized doses if a specific cannabis strain is processed under controlled conditions. Id. DEA also asserted that there are no adequate marijuana safety studies for use in specific, recognized conditions, nor are there adequate, well-controlled studies that determine marijuana’s efficacy. Id. Also, there is no consensus of expert opinions about the medical utility of marijuana for treating specific disorders. Id. Lastly, DEA stated that there is a lack of currently available data on marijuana that sufficiently addresses its chemistry, toxicology, and effectiveness. Id. In addition, the agency noted there is a lack of scientific evidence about marijuana’s chemistry to a specific cannabis strain that could be formulated into standardized and reproducible doses. Id. DEA, thus applying the five-part test, found that marijuana meets none of the five elements.

While denying the petitions for initiating proceedings to reschedule marijuana, the DEA’s findings offer a potential roadmap for future rescheduling petitions. Any future petitions would do well to limit the rescheduling request to a particular strain or strains from which standardized doses can be processed for a specific disorder. The strain should have been subjected to safety and efficacy studies for use in specific, recognized conditions. Petitioners should elicit expert opinions about the medical utility of the marijuana strain for treating those specific disorders. In addition, petitioners should provide data that sufficiently addresses the chemistry, toxicology, and effectiveness of that specific marijuana strain. In summary, the key to any future rescheduling of marijuana remains a finding, specifically by HHS, and supported by DEA, that marijuana does have an accepted medical use in the United States.

DEA stated in its notice to allow additional cultivators of marijuana for research that it might take a negative inference if an entity previously conducted activities in violation of federal law. We are left wondering how this position would play out if marijuana is eventually rescheduled and whether DEA will similarly scrutinize an applicant’s experience in conducting authorized activities in states where marijuana has been legalized for medical use. This would include practitioners, pharmacies, cultivators, distributors, researchers and testing facilities, all of whom are licensed and required to comply with regulatory requirements under state law. Some of these entities, including many practitioners who prescribe or authorize marijuana for authorized indications, are current DEA registrants.

By Kurt R. Karst –       

Q. Why did the chicken cross the road?
A. To get to the other side.

Q. Why did the turtle cross the road?
A. To get to the shell station.

Q. Why did the duck cross the road?
A. To prove he wasn’t chicken!

Q. Why did the cow cross the road?
A. To get to the udder side!

Q. What happened when the elephant crossed the road?
 A. It stepped on the chicken!

Q. Why did the . . . . .  Well, you get the drift.  The same old joke can be dressed up a bit, but the punchline ultimately remains the same. That’s kind of how we view FDA’s Eighth Annual Report to Congress on Delays in Approvals of Applications Related to Citizen Petitions and Petitions for Stay of Agency Action for Fiscal Year 2015.  The report, which is required by FDC Act § 505(q)(3), gives us the low-down on FDA’s experience during Fiscal Year 2015 (“FY 2015”) with citizen petitions subject to FDC Act § 505(q).  Other than some updated numbers for FY 2015, however, the report largely repeats both FDA’s concerns about petitioning expressed in previous reports (see our previous posts here, here, here, here, here, here, and here) and the trends the Agency has been seeing in petitioning.  (And to make our point on repeating the same punchline, the preceding sentence is almost identical to the sentence we used in our post on FDA’s FY 2014 Report to Congress.)

By way of background, FDC Act § 505(q) was added to the law by the 2007 FDA Amendments Act (“FDAAA”) and is intended to prevent the citizen petition process from being used to delay approval of pending ANDAs and 505(b)(2) applications.  The law was amended by Section 301 of Pub. L. No. 110-316 (2008), and again by Section 1135 of the 2012 FDA Safety and Innovation Act (“FDASIA”).  Among other things, FDASIA changed the original 180-day response deadline to 150 days, and made the law applicable to citizen petitions concerning biosimilar applications submitted to FDA pursuant to PHS Act § 351(k). In June 2011, FDA issued final guidance on FDC Act § 505(q).  That guidance was revised in November 2014 to account for changes made to the law by FDASIA.  In January 2012, FDA issued proposed regulations to amend the Agency’s citizen petition regulations to implement changes made to the law by Section 505(q). FDA appears to be nearing completion of a final rule (see here).

Under FDC Act § 505(q), FDA shall not delay approval of a pending ANDA, 505(b)(2) application, or 351(k) biosimilar application as a result of a citizen petition submitted to the Agency pursuant to 21 C.F.R. § 10.30 (citizen petition) or § 10.35 (petition for stay of action), unless FDA “determines, upon reviewing the petition, that a delay is necessary to protect the public health.” FDA is required to “take final agency action on a petition not later than 150 days after the date on which the petition is submitted.”  FDA may not extend the 150-day period “for any reason,” including consent of the petitioner.  Although the statute provides that FDA may summarily deny a petition submitted with the primary purpose of delaying ANDA, 505(b)(2) application, or 351(k) biosimilar approval, the Agency has never done so, as noted above by FDA.

FDC Act § 505(q)(3) requires that each annual report to Congress specify: “(A) the number of applications that were approved during the preceding 12-month period; (B) the number of such applications whose effective dates were delayed by petitions . . . during such period; (C) the number of days by which such applications were so delayed; and (D) the number of such petitions that were submitted during such period.”  FDA says in its Eighth Annual Report to Congress that:

During the FY 2015 reporting period, the Agency approved 492 ANDAs, 45 505(b)(2) applications, and 1 biosimilar biological product application. The approval of one 505(b)(2) application was delayed because of one 505(q) petition, and the approval of one ANDA was delayed because of two 505(q) petitions. No approvals for biosimilar biological product applications were delayed because of a 505(q) petition in this reporting period.

The ANDA and 505(b)(2) approvals were delayed by 141 and 44 days, respectively. “FDA was concerned that if it approved the ANDA and 505(b)(2) applications before resolving the issues raised in the petitions and later concluded that one or more of the arguments against approval were meritorious, then the presence on the market of a drug product that did not meet the requirements for approval could negatively affect public health,” says FDA in the report (in what is now boilerplate language. FDA does not identify by name or application number the particular approvals delayed.

As to the number of 505(q) citizen petitions submitted in FY 2015, the Report says that 15 of the 74 citizen petitions (or 20%) handled by the Center for Drug Evaluation and Research (excluding ANDA suitability petitions and petitions that raise only OTC monograph issues) were 505(q) petitions.  That’s a 7% decrease compared to FY 2014 when FDA received 28 505(q) petitions.  “From FY 2008 through FY 2015, FDA received a total of 175 petitions subject to section 505(q). Over this 8-year period, FDA responded to all but 11 of the 505(q) petitions within the statutory time frame that was applicable during that period” (i.e., 180 days or 150 days), comments FDA in the report. The report includes tables showing the percentage of 505(q) petitions received during FYs 2008-2015, and the outcomes for the 167 petitions that have been resolved under FDC Act § 505(q) as of September 30, 2015.     

Most of FDA’s 505(q) petition decisions – about 68% of them – have been denials. Often, these denials are without comment (see, e.g., here). According to FDA, “[t]here is no evidence that in enacting section 505(q) of the FD&C Act, Congress intended to bypass the application review process or to lessen the procedural rights of an ANDA or NDA applicant by requiring that the Agency make decisions that constitute final Agency action regarding the approvability of certain aspects of pending applications on a piecemeal basis outside of the process established under the FD&C Act and FDA regulations.”

As to 505(q) petitioning trends and FDA concerns, although the Agency pared down comments that appeared in previous reports, the bottom line in the FY 2015 Report is still the same. FDA says that “[i]t is difficult to determine whether section 505(q) is discouraging the filing of citizen petitions aimed at blocking generic or biosimilar competition. . . . There are no clear trends in the data over time.” But FDA’s primary concern – and one expressed over and over in previous reports – remains resources:

The Agency continues to be concerned that section 505(q) may not be discouraging the submission of petitions that are intended primarily to delay the approval of competing drug products and do not raise valid scientific issues. The statute requires FDA to prioritize these petitions above other matters, such as safety petitions, that do raise important public health concerns. As a result, FDA remains concerned about the resources required to respond to 505(q) petitions within the 150-day deadline at the expense of completing the other work of the Agency.

Two specific concerns identified in previous reports are absent from the FY 2015 Report. Specifically, in the FY 2014 Report, FDA included the following comments:

• FDA continues to receive serial 505(q) petitions, frequently from the same petitioner, about the same specific drug or class of drugs, sometimes requiring several separate responses about different issues regarding the same product.  Responding to such serial petitions requires the use of substantial FDA resources, on a repeated basis, over a protracted period of time.

• 505(q) contains a provision that permits FDA to summarily deny a petition at any point if FDA finds that it was submitted with the primary purpose of delaying the approval of an ANDA, 505(b)(2) application, or 351 (k) application and the petition does not “on its face” raise valid scientific or regulatory issues (FD&C Act, section 505(q)(l)(E)).  As FDA previously noted in its Report to Congress, “Encouraging Early Submission of Citizen Petitions and Petitions for Stay of Agency Action” dated February 2009, we believe that the statutory language requires that both preconditions be present, and we believe this statutory standard would be extremely difficult to meet.  To date, FDA has never applied this provision to summarily deny a petition, despite the fact that, in FDA’s estimation, many 505(q) petitions do not in fact raise persuasive scientific or regulatory issues when those issues have been reviewed by FDA (as previously noted, approximately two-thirds of these petitions are denied in full).  Accordingly, it is FDA’s view that this provision has neither curbed the filing of petitions submitted with the primary purpose of delay nor has it permitted FDA to dispose of such petitions without expending substantial amounts of resources.

Although these two concerns are not specifically identified in the FY 2015 Report, that’s not to say they are no longer concerns for FDA. (After all, serial petitioning still happens today, and FDA still has not used FDC Act § 505(q)(l)(E) to deny a petition).  Both concerns appear in the FY 2015 Report in spirit as they come down to issues concerning a strain on Agency resources and FDA’s view that FDC Act § 505(q) has not discouraged petitioning.

By Kurt R. Karst –  

Last month, while this blogger was on vacation enjoying the fine beers of Germany during a tour of that country, the U.S. District Court for the District of Columbia issued a Memorandum Opinion in a rather interesting Hatch-Waxman Patent Term Extension (“PTE”) dispute. As we previously posted, the case at bar offers up a new twist on the question: “When is an NDA subject to a rolling review initially submitted to FDA?”

As we previously reported (here and here), in April 2015, Boehringer Ingelheim Pharma GmbH & Co. KG and Boehringer Ingelheim Pharmaceuticals, Inc. (collectively “Boehringer”) filed a Complaint alleging that FDA and the PTO unlawfully shorted by about two months a PTE for U.S. Patent No. 6,087,380 (“the ‘380 patent) covering Boehringer’s PRADAXA (dabigatran etexilate) Capsules (NDA 022512). The Complaint was prompted by a December 2014 PTE regulatory review period decision in which FDA refused to revise the Agency’s previous determination published in the Federal Register.   In that May 2012 Notice, FDA commented that:

[Boehringer] claims December 15, 2009, as the date the new drug application (NDA) for PRADAXA (NDA 22-512) was initially submitted. However, FDA records indicate that NDA 22-512, received December 15, 2009, was incomplete.  FDA refused to file this application and notified the applicant of this fact by letter dated February 12, 2010.  The completed NDA was then submitted on April 19, 2010, which is considered to be the NDA initially submitted date.

Boehringer alleged that FDA violated the Administrative Procedure Act (“APA”) and the Hatch-Waxman Amendments and regulations – as well as inconsistency with a PTE precedent from concerning TONOCARD (tocainide HCl) Tablets – when the Agency unlawfully relied on the NDA “filing” standard instead of an “initially submitted” standard in the Agency’s PTE calculation for the ‘380 patent.

In her 26-page July 2016 ruling, Judge Colleen Kollar-Kotelly granted FDA’s Motion for Summary Judgment and denied Boehringer’s Motion for Summary Judgment.  “Ultimately, the question before the Court is narrow,” wrote Judge Kollar-Kotelly: “whether the agency’s determination of the date on which the approval phase begins is lawful under the [FDC Act], as amended, not whether Congress has chosen the best statutory scheme or whether the agency has implemented that scheme through the best set of regulations” (emphasis in original). 

With respect to Boehringer’s first argument – “that [FDA’s] determination of the start date of the approval phase of the FDA’s regulatory review period is inconsistent with the statute” – the court addressed that argument through the familiar two-step Chevron framework, ultimately concluding under Step Two that FDA’s determination is consistent with the statute.

Plaintiffs argue that the phrase “the date an application was initially submitted” is unambiguous. Defendants respond that the phrase is ambiguous because the statute does not specify exactly what must be “submitted” to satisfy that requirement.  The Court agrees with Defendants.  Plaintiffs emphasize the use of the word “initially” in the phrase “initially submitted,” suggesting talismanic importance to the use of that word.  But the fact that the statute refers to the initial submission—as opposed to a later or final submission—does not resolve the question at hand.  Notwithstanding Plaintiffs’ arguments to the contrary, the Court concludes that the statute simply does not specify what must be submitted initially to satisfy the statutory requirements.  It is clear that the application must include several components—enumerated immediately above. See 21 U.S.C. § 355(b)(1).  But it is not clear from the statutory language what must actually be submitted to qualify. That is, the statute is ambiguous as to whether a deficient application qualifies as an application that was “initially submitted” or whether materials submitted must be sufficient for substantive review in order to qualify as an application that was “initially submitted”. . . .

[T]he agency’s interpretation of the statutory language, as stated in the regulations, warrants deference and is reasonable. As a regulation duly promulgated with notice and comment there is no dispute such an interpretation is the sort of agency pronouncement that warrants deference.  Similarly, this interpretation is plainly reasonable.  The agency determined that, in order to qualify as “initially submitted,” an application must have sufficient information to allow the agency to commence a substantive review.  This is sensible. The purpose of the agency’s review process is for the agency to be able to review the application to determine whether the proposed product satisfies the several complex statutory criteria.  Any application that would not allow the FDA to begin a substantive review would not fulfill the purpose of the substantive review process.  Therefore, it is proper not to consider any such deficient application to have been “initially submitted.” [(Emphasis in original)]

Boehringer also argued that FDA’s action was inconsistent with the Agency’s regulations, and specifically with 21 C.F.R. § 60.22, which provides that the approval phase of the PTE regulatory review period begins “on the date [the application] contains sufficient information to allow FDA to commence review of the application.” But like Boehringer’s statutory argument, Judge Kollar-Kotelly deferred to FDA’s interpretation of the PTE regulation, concluding that “it was reasonable for [FDA] to determine that the materials provided to the agency as of December 15, 2009, did not ‘contain[] sufficient information to allow FDA to commence review of the application’ because they were fatally deficient and would not allow a full substantive review of the proposed product.”

Finally, Judge Kollar-Kotelly dismissed Boehringer’s argument that the Agency’s determination with respect to the ‘380 patent covering PRADAXA violated the APA in light of the TONOCARD precedent (which we discussed in a previous post). That instance appears to have been a one-off for FDA. “Ultimately, the question comes down to whether, in each individual case, the applicant has provided sufficient information for the agency to commence its substantive review,” wrote Judge Kollar-Kotelly.

Interestingly, Judge Kollar-Kotelly notes in her Opinion, and which FDA raised in the Agency Motion for Summary Judgment, that Boehringer did not avail itself of the opportunity – see 21 C.F.R. § 314.101(a)(3) – to have the PRADAXA NDA filed over protest and reviewed by FDA.  Instead, the company accepted FDA’s refuse-to-file letter and responded to the deficiencies identified therein a couple of months later, after which FDA accepted the NDA as filed.  In hindsight, would that have been the better option?  Perhaps – at least for PTE purposes – but who knows what delays (if any) might have resulted in final NDA approval.  In any case, it highlights the importance of communication between a company’s regulatory and intellectual property counsel to maximize a product’s protections. 

By Riëtte van Laack –

On August 15, 2016 FDA simultaneously issued two guidance documents, the Draft Guidance, Calorie Labeling of Articles of Food in Vending Machines and the Final Guidance, Small Entity Compliance Guide (SECG), Calorie Labeling of Articles of Food in Vending Machines, to help Vending Machine Operators comply with the December 1, 2014 final rule, Food Labeling: Calorie Labeling of Articles of Food Sold in Vending Machines (“the rule”) codified at 21 C.F.R. 101.8 (see our prior posting here).  Except for certain gums, mints and roll candy products in glass-front machines, the effective date is December 1, 2016. 

The SECG is intended to help small businesses comply with the rule and restates the requirements of the final rule in plain language. The draft guidance provides additional information and clarification regarding certain aspects of the rule, and responds to frequently asked questions.

Although the SECG is intended to help small businesses, it provides useful information for any business, e.g., pictures of various formats used to declare calorie content of the product with details about font size requirements.

Vending Machine Operators are Subject to the Rule

The draft guidance clarifies what constitutes a vending machine operator subject to the rule.  The labeling requirements apply to vending machine operators who own or operate 20 or more vending machines that dispense food items (the term food includes dietary supplements); the number of vending machines that dispense only non-food items is not relevant. Vending machine operators who own or operate less than 20 vending machines that dispense food items may voluntarily register with FDA.  According to FDA, voluntary registration exempts a vending machine operator from “State or local nutrition labeling requirements for foods sold in vending machines that are not identical to the [federal] requirements.”  However, this voluntary registration does not mean that the operator is exempt from state and local licensing requirements.  Vending machine operators that own or operate 20 or more vending machine do not need to register because they are automatically covered by the rule.

FDA explains that in situations where vending machines are owned or leased, and another entity is contracted to control or direct the function of the vending machine (including selecting the foods to be vended), and the owner or lessee is not compensated for the control or direction of the vending machine, then that owner or lessee would not fit the definition of a vending machine operator and therefore would not be responsible for posting the calorie declarations.

Total Calorie Declarations Must be Declared and Visible Prior to Purchase

Calories must be declared prominently in a clear and conspicuous manner for the total calories in the vended food not per serving if the vended food constitutes more than one serving. If the vended food includes condiments the total calories must include the condiment.  If, however, the condiment is provided next to the vending machine, the labeled calorie content does not include the condiment.  If the food comes in variations, such as coffee with options for cream, sugar, or sugar substitute, then the calories must declared per option, or as the total calories of each variation in the final vended product. 

Calorie declarations may be based on various sources, including the Nutrition Facts Label provided by the manufacturer or supplier, nutrient databases (i.e., USDA database, or FDA Nutrition Facts found in Appendix C of 21 C.F.R. 101 for fruits and vegetables), cookbooks, or laboratory analysis. If a food varies in calorie content, such as an apple, then the approximate standard size for calorie declarations may be used or different size calorie declarations may be used. 

Calorie declarations may be on or near the food, or on or near the vending machine (e.g. using a sign). Because the side of the vending machine usually is not visible to the consumer, the information may not be placed on the side.  The calorie declarations that are displayed must be for foods that are currently or typically sold from that machine.  Calorie declarations may be presented in any format as long as they are clear and conspicuous.  Although not required, if a sign adjacent to the vending machine is used for calorie declaration of a number of foods, FDA encourages including the calorie information for all foods sold in the vending machine, even for foods which have calorie labeling on the food (i.e., front of package),

The compliance date is December 1, 2016. Any vending machine operator that becomes subject to the rule after that date, e.g., by adding vending machines to their operation, must comply immediately.  FDA recently extended the compliance date for type size requirements and calorie disclosure requirements for certain gums, mints, and roll candy products in glass-front machines until July 26, 2018.

Comments to the draft guidance are due September 30, 2016. Comments to the SECG may be submitted any time. 

By Jennifer D. Newberger –

Though the UDI compliance date for Class II devices is just over one month away, one key question remains unanswered: are private label distributors “labelers” for purposes of the UDI rule? This turns out to be much less clear-cut than it might seem. FDA has publicly stated that they are, without providing any supporting rationale for its position. The original economic impact analysis to support the final rule, however, paints a very different picture.

The UDI regulation defines a “labeler” as a person who “causes” a label to be applied to a device. FDA notes that the labeler is usually the manufacturer, but takes the position that it also may include specification developers, repackagers, or relabelers. See 78 Fed. Reg. 58786, 58787 (Sept. 24, 2013).

Neither the rule nor the preamble to the rule specifically addresses the role of private label distributors. Private label distribution is common in the medical device industry. Companies often contract with a third party to have their name and brand on the label of a device over which they do not exercise control. The private label distributor does not own the 510(k), is not the specification developer, does not repackage or relabel, and does not manufacture the product. Therefore, private label distributors are not “manufacturers” subject to FDA regulation and are not required to register with FDA under 21 C.F.R. Part 807.

In informal advice, FDA officials have expressed the view that private label distributors are “labelers” required to comply with the UDI rule. For example, in a RAPS Regulatory Exchange Open Forum on July 8, a commenter posted a response received from FDA’s UDI Help Desk regarding private label distributors. The response stated, in relevant part:

Example 2: Company A manufactures a device under contract with Company B.  As part of the contract, Company A places the label with Company B's brand on the immediate container of the device.  Company B, as the private label company, is the labeler. Company A as the contract manufacturer or the contract packaging company, is not the labeler for purposes of UDI compliance.  However, UDI Rules allow third parties to submit GUDID.  Therefore, Company B can contract with Company A and authorize Company A to open a GUDID account on behalf of Company B and to submit the required information.

Example 3:  Company A manufactures a device and commercially distributes the device under its own brand.  Company A also manufactures the same version or model of the device under contract with Company B, Company C, and Company D.  As part of the contract, Company A places the label with Company B, C, or D's brand, respectively, on the immediate container of the device.  Company A, B, C, and D are all labelers and required to submit information pertaining to their respective brands to GUDID.  In addition, despite the fact that the devices are all the same version or model, each brand is required to have its own UDI. . . .

Along the same lines, our firm asked the Help Desk whether private label distributors are “labelers” for UDI purposes. We received the following response:

In general, our current position is that where the name and place of business of the private label distributor is on the device label to meet the requirements of 21 CFR 801.1, the private label distributor is the labeler for UDI purposes.

To our knowledge, FDA has not publicly cited a rationale for this position. We have learned that one basis for the position that private label distributors are labelers for UDI is that the concept of labeler for UDI was borrowed from its use in the drug space, although this position was not publicly articulated during the course of the rulemaking. The idea is that the manufacturer of a product—be it a drug or device—for a party may change over time, and neither party may want to publicly divulge that relationship. Therefore, the labeler is essentially the “public face” of the product.

Though this position is not without merit, private label distributors still find themselves in something of a bind—their obligations under the UDI rule were not clearly discussed in either the proposed or final rule, and they now find themselves responsible for regulatory obligations for which most are unprepared. Moreover, this position appears to be contrary to what FDA told the public during the UDI rulemaking. For instance, FDA’s economic impact report relies upon an analysis developed by Eastern Research Group, Inc. (ERG). See Unique Device Identification System; Final Rule, Docket No. FDA-2011-N-0090, Final Regulatory Impact Analysis, Final Regulatory Flexibility Analysis, Unfunded Mandates Reform Act Analysis. The ERG report includes a section, “Profile of the Affected Entities,” that discusses the entities considered to be labelers:

Based on FDA’s definition of labeler in the proposed rule [which is the same as that in the final rule], the affected entities are expected to include manufacturers, single-use device reprocessors (which take single use devices, sterilize them, and return them to the end user), specification developers (who oversee the manufacture of devices by contract manufacturers), and relabelers/repackagers (R/Rs).

May 2012 ERG Report, at 3-1. The May 2012 report also includes the following table (emphasis added):

Both the language from the ERG report and the table above indicate that the only entities that meet the definition of a labeler are those that are required to register with FDA as the specified types of medical device establishments in the table above. It is clear from the ERG report that the table above was intended to be a complete list of all “affected entities,” i.e., labelers. Notably absent from the list are private label distributors.

In fact, the only discussion of private label distributors in the ERG report specifically excludes those entities from the definition of a labeler:

Some specification development is done under the specification developer’s brand; other times, specification developers are contracted by private label distributors to provide the specifications and coordinate the manufacturing of the device to those specifications. Distributors are not allowed to list devices in FDA’s database, however, and do not keep the device records. In the first case, where the specification developer and the contract manufacturer are the only parties involved, we assume that the costs of labeling are ultimately borne by the specification developer. In the second case, where a third party private labeler is involved, it is still assumed that the specification developer is the UDI labeler of record and incurs the immediate costs of any UDI requirements. Some of these costs might be passed to the third-party private labeler or the labeler might handle the actual label application and thereby incur costs. Nevertheless, we assume that the specification developers bear the costs and impacts because there is no way to clearly determine the extent of third-party interactions.

May 2012 ERG Report, at 3-13 (emphasis added).

Thus, in addressing the situation where there is a specification developer and a private labeler, ERG expressly allocated the UDI compliance costs to the specification developer—not the private label distributor. In its own cost analysis, FDA does not specifically mention private label distributors, but accepts the analysis provided by ERG.

We understand that it is FDA’s position that most of the costs for UDI compliance will in fact be borne by the manufacturer or specification developer, and not the private label distributor, since most of the costs associated with compliance relate to label printing. Under this view, the economic impact report is correct in stating that the costs of UDI compliance will be borne by the manufacturer or specification developer. The flaw in this argument, however, rests with the actual words in the report—not only does it say that costs will be borne by the specification developer, it also says that the “specification developer is the UDI labeler of record.” If costs could be separated out from UDI labeler of record, that should have been specified. But it was not, and the economic report therefore is clear that private labeler distributors do not bear the costs associated with UDI compliance, nor are they the “UDI labeler of record.”

Based upon the language of the proposed rule, private label distributors had no reason to believe that they would be subject to UDI regulatory obligations. If FDA believed that private label distributors were subject to the UDI regulation, the agency should have said so explicitly. FDA’s subsequent application of UDI regulations has therefore come as quite a surprise, and without input from this large segment of industry. FDA should remedy this situation by publicly renouncing its position that private label distributors are “labelers” for purposes of the UDI rule.

We would finally note that this is just one of the many confounding aspects of the UDI regulation. FDA may have initially thought the UDI rules were straightforward. They are not. A recent report found that “just 15% of respondents indicate that they are already compliant with the next phase of the regulation, and of those who are currently working towards compliance, nearly 40% will be taking it right to the due date of Sept. 24th.” Taking the Pulse of UDI Compliance, A Survey of the Medical Device Industry on Compliance and Labeling.  With that September 24 compliance date looming, it may be too late to fix the multiple ambiguities of the UDI regulation. However, the UDI regulation should stand as an abject lesson in the need for greater clarity whenever CDRH considers imposing a major new regulatory obligation.

By John A. Gilbert, Jr. & Larry K. Houck –

Last week we summarized several new Drug Enforcement Administration (“DEA”) actions related to marijuana.  We now take a more in depth review of one of those actions related to DEA registration to cultivate marijuana for research.

For almost 50 years the DEA has granted only one DEA manufacturer registration for marijuana and thus restricted all marijuana production for research to one entity, the University of Mississippi, under contract with the National Institute on Drug Abuse (“NIDA”). DEA’s longstanding policy was based on belief that “having fewer registrants of a given controlled substances [sic] tends to decrease the likelihood of diversion.” Applications to Become Registered Under the Controlled Substances Act To Manufacture Marijuana To Supply Researchers in the United States, 81 Fed. Reg. 53,846, 53,847 (Aug. 12, 2016) [hereinafter Applications to Become Registered]. To our knowledge DEA has never provided any data to support this position. In addition, the government considered that controlled arrangement as the best way to comply with U.S. obligations under the Single Convention on Narcotic Drugs, 1961 (“Single Convention”). Id. at 53,846, 53,846-47. Then last week the DEA, concurrent with affirming the status of marijuana as a federally-controlled schedule I substance and reporting the U.S. Department of Agriculture’s statement of principles on industrial hemp under the Agricultural Act of 2014, abrogated its long-held policy limiting marijuana cultivation for research to a single grower.

DEA explained that the prior arrangement worked for so long because the single grower could meet the limited U.S. demand for research-grade marijuana. Id. at 53,846. DEA asserted that with the Food and Drug Administration (“FDA”) and NIDA, it “fully supports expanding research into the potential medical utility of marijuana and its chemical constituents.” Id. DEA stated that it recognized the recent increased interest in the research of certain cannabinoids including cannabidiol, and concluded, again based upon discussions with NIDA and FDA, “that the best way to satisfy the current researcher demand is to increase the number of federally-authorized marijuana growers.” Id. While the prior system was geared towards federally-funded and academic research, not commercial product development, the new approach will also foster private sector commercial endeavors for product development independent of NIDA. Id.  While the policy change is to be applauded, the cynical among us might surmise that DEA loosened its marijuana growth policy to blunt criticism of its recent decision to deny the petitions to reschedule marijuana.

Authority to Cultivate Marijuana Under the CSA

The federal Controlled Substances Act (“CSA”) requires all controlled substance manufacturers and distributors to apply for and obtain a DEA registration authorizing those activities, but the registrations must be consistent with the public interest and with U.S. obligations under the Single Convention. 21 U.S.C. §823(a)(1); Id. at 53,847. DEA cautions that while it intends to increase the number of registered marijuana cultivators for supplying researchers (assuming it receives qualified applicants), it must register only the number of bulk manufacturers necessary to “produce an adequate and uninterrupted supply of these substances under adequately competitive conditions for legitimate medical, scientific, research, and industrial purposes.” Applications to Become Registered at 53,847. Thus, DEA will evaluate every application submitted to determine whether each registration is necessary to provide an adequate and uninterrupted supply of marijuana, its extracts and derivatives to U.S. researchers. Id. Again, it will be interesting to see how DEA conducts this evaluation given the precedent established in its decisions related to importers of narcotic raw materials.

The CSA establishes six criteria that DEA must consider in evaluating as to whether an application for a manufacturer registration for schedule I substances is consistent with the public interest. 21 U.S.C. § 823(a). DEA indicated that it intends to place specific emphasis on factor 3, that is, whether the applicant has previous experience handling controlled substances in a lawful manner and whether the applicant has engaged in illegal activity involving controlled substances. In this context, DEA stated that illegal activity incudes any activity in violation of the CSA (regardless of whether such activity is permissible under State law) as well as activity in violation of State or local law.” Id. DEA observes, though, that while “past illegal conduct” does not automatically disqualify an applicant, “it may weigh heavily against granting the registration.” Applications to Become Registered at 53,847. This is an interesting position to say the least given that the statutory criteria also requires DEA to consider whether an applicant has complied with “State and local laws.” 21 U.S.C. § 823(a)(2). So in this case DEA appears to be stating that they will ignore this criteria in favor of compliance with federal law.

DEA appears to be directing these statements at potential applicants who are currently providing state authorized marijuana-related services including cultivation, distribution and testing. Such activities without a DEA registration violate the CSA. So these entities may be on the outside looking in as DEA reviews applications for registrations for marijuana cultivation and other activities. This position seems inconsistent with the U.S. Department of Justice (“DOJ”) policy as stated in the “Cole Memo” that it would not preempt state law in states that legalized marijuana for medical or recreational use, and the DOJ and DEA were unlikely to take action against a marijuana-related business operating in compliance with state law as long as the activity did not implicate one of the enforcement priorities. Memorandum for all United States Attorneys from James M. Cole, Deputy Attorney General, Guidance Regarding Marijuana Enforcement, Aug. 29, 2013. However, it now appears that marijuana-related activities authorized under state law but which violated the CSA may constitute grounds for DEA to deny an applicant’s registration to manufacture marijuana for research. So the question is whether DEA in registering a limited number of marijuana manufacturers, given the choice between registering applicants who conducted state-authorized marijuana activities and those that have not, all else being equal, will choose to register the latter over the former.

It will be interesting to see if Congress reacts to DEA’s position given that Congress has recently been active in passing legislation to limit DEA’s authority to take adverse action against marijuana-related entities operating within the scope of state law.

Requirements Under the Single Convention

As discussed above, DEA must also ensure that any registration to cultivate marijuana is also consistent with U.S. obligations under the Single Convention. Articles 23 and 28 of the Single Convention require signatory countries that allow the cultivation of cannabis for lawful uses such as manufacturing for research to:

1. Designate the areas and plots of land where it will permit cannabis cultivation;
2. License cultivators;
3. Specify through licensing the extent of the land on which cultivation is permitted;
4. “Purchase and take physical possession of all cannabis crops from all cultivators as soon as possible, but not later than four months after the end of the harvest;” and
5. Have the “exclusive right of importing, exporting, wholesale trading and maintaining stocks of cannabis.” Applications to Become Registered, at 53,847.

DEA’s registration system carries out the functions of article 23, paragraph 2, a-c above, while NIDA carries out functions related purchasing marijuana and maintaining a wholesale distribution monopoly (d and e above). Id. At first blush it would seem from the Single Convention requirements that NIDA, DEA or another government agency must purchase, take possession, and control wholesale distributing, importing, exporting and maintaining marijuana stocks. This would not be much different from the current system except there are more growers. DEA stated that that it would be consistent with the purposes of the Single Convention for DEA to register marijuana growers outside of the single grower NIDA system to supply researchers, if the growers agree to distribute marijuana only with prior, written approval from DEA. Id. at 53,847-48 This is consistent with other CSA requirements such as import and export permits for schedule I and II substances and all narcotic substances. Of course DEA will have to implement a system to review and issue approvals in a timely manner. DEA will authorize registered marijuana growers to operate independent of NIDA if they execute a written memorandum of agreement with the agency to distribute marijuana with prior, written approval from DEA only to DEA-registered researchers who have HHS-approved protocols. Id. As with all schedule I manufacturers, marijuana cultivators will be subject to all CSA schedule I requirements including quota, inventory, recordkeeping and reporting, order form and security requirements.

It remains to be seen how many applications DEA will receive to cultivate marijuana for research and, of more interest, how many registrations DEA will actually issue. While, DEA asserted that its prior policy was based on the precept that fewer registrants meant less likelihood of diversion, the agency provided no data to support this position. In an interesting juxtaposition, this issue was widely argued and debated in a line of cases related to the import of narcotic raw materials. See, e.g., Johnson Matthey, Inc.: Conditional Grant of Registration to Import Schedules II Substances, 67 Fed. Reg. 39,041, 39,043-44 (June 6, 2002) (finding that “DEA is required to register an applicant . . . if the Administration determines that registering another manufacturer will not increase the difficulty of maintaining effective controls against diversion.” So it will be interesting to see how DEA applies this standard in the context of registering marijuana manufacturers.

Registrants can apply for a DEA registration to manufacture marijuana for research by filling out a DEA Form 225, available on the DEA Office of Diversion Control website. Applications to Become Registered at 53,848.

By Jennifer D. Newberger –

On August 11, 2016, the Office of Combination Products (OCP) announced a “pre-RFD” process. A Request for Designation (RFD) is the formal process for seeking FDA’s assessment of product classification, for example, determining whether a product should be regulated as a drug or a device. The key issue for designation is determining the primary mode of action (PMOA) of the product. If the PMOA cannot be determined, OCP is supposed to designate based on the Center that regulates other combination products that present similar questions of safety and effectiveness. If there are no such combination products, then OCP will assign jurisdiction to the Center with the most expertise related to the most significant safety and effectiveness questions presented by the product.

OCP has earned a reputation for designating products as drugs rather than devices, and OCP is well aware of industry’s frustration with the unwritten presumption that products should be regulated as drugs. OCP has supposedly been listening to industry, and trying to find ways to ease these frustrations. Earlier this year, FDA announced creation of the Combination Product Policy Council, which is intended to be “a senior-level, agency-wide forum for discussing, resolving, and implementing product and policy issues.” We are not yet aware of any activities by the Council, or how, precisely, its activities will ensure a fair review of combination products and assignment of the proper jurisdictional designation.

Now, OCP has announced the pre-RFD process, intended to provide a more “interactive approach” to product designation. OCP acknowledges that the information to be provided to OCP as part of the pre-RFD process is similar to that in a formal RFD—sponsors must provide “a complete, clear, and detailed product description, which includes the product’s indication for use, its composition/ingredients, and an explanation of how it works.” OCP also states that both pre-RFD and RFD “require input from the product jurisdiction officers in the relevant Centers and, if necessary, legal perspectives from the Office of Chief Counsel.”

The following are the differences between the pre-RFD and RFD processes:

1. Sponsors are not required to provide a recommendation for classification and assignment of their product along with a corresponding rationale (e.g., bench studies; clinical studies) for that recommendation;
2. Sponsors are not required to discuss the classification of currently marketed products that they believe to be similar to their product; and,
3. Sponsors can receive preliminary feedback and information from the Agency that is derived from a structured and efficient process. The feedback will ultimately help lead to better decision-making and development of products for the sponsors.

For anyone who has ever dealt with OCP, the idea of providing information about a product, without also having the opportunity to provide a recommendation as to product jurisdiction and relevant precedent, should be unsettling. Even when provided with a jurisdictional recommendation, and information about precedent, OCP is likely to assign the product to CDER. Without the sponsor advocating for its position, the odds of a CDER jurisdictional determination is nearly a foregone conclusion in any submission where the product might arguably be a drug.

As for interactive feedback, we have heard this line before. OCP previously offered what it called an “informal RFD” process with the promise of interactive feedback and an opportunity to address questions as they arose during the course of OCP’s review. The reality, however, was quite different. The process still took 60 days and there was no interaction between OCP and the sponsor. The end result? An “informal” decision that the product was—you guessed it!—a drug.

The only possible advantage to engaging in a pre-RFD or informal RFD before submitting a formal RFD is that it will give the company insight into what OCP is thinking and the basis for why it believes the product is a drug. This information could prove useful to the sponsor when developing its formal RFD submission, as it would allow the company to address the issues raised by OCP in the earlier process. The downside, of course, is that it takes 60 days to reach an informal or pre-RFD decision, and the decision reached is not binding. If a company needs a formal designation, it must then submit a formal RFD and await a formal Letter of Designation—another 60 days. And if it wants to appeal to the Office of Special Medical Programs, there is no deadline for a decision, dragging out the process even longer.

The problem with OCP is not the RFD process per se—it is the outcome of that process, particularly the inherent bias that most combination products can and should be regulated as drugs. Until that bias is fixed, no process, formal or informal, will alter industry’s frustrations with OCP.

In recent years, the level of scrutiny on foreign pharmaceutical manufacturing facilities in countries like China and India has skyrocketed—along with the demand and dependence of western countries on the supply of goods coming from those facilities. Furthermore, increased enforcement aimed at the medical device industry and the safety and effectiveness of medical device production have become a key focus of global regulators.

With the latest updates to the global standard for medical device quality management systems and the proposed new EU Medical Device Regulations, device manufacturers, service providers, and supply chain partners will need to closely assess their product lifecycle risk management systems to ensure they are complying with requirements.

This webinar will discuss these recent pharmaceutical and medical device supply chain developments and what steps companies need to take to prevent and remediate compliance issues in the context of:

• The uptick in investigations in China and India.
• Speed vs. Safety: the approval process differences between the EU and U.S.
• Increased scrutiny of non-EU-based manufactures and requirements for importers and distributors.
• The potential impact of adverse regulatory actions against nonconforming facilities.
• Threats posed by bribery and corruption.
• Anti-corruption compliance issues from U.S. and Asian perspectives.
• “War stories” from recent investigations.

When

Wednesday, September 14, 2016
1:30 p.m. – 2:30 p.m. BST

Where

This presentation will be simulcast via Webex as a webinar.  Please click here to register. 

Speakers

Douglas B. Farquhar
Director
Hyman, Phelps & McNamara PC
Washington, D.C.

Mark I. Schwartz
Of Counsel
Hyman, Phelps & McNamara PC
Washington, D.C.

Jeremy B. Zucker
Partner
Dechert LLP
Washington, D.C.

Lewis Ho
Partner
Dechert LLP
Hong Kong

Kareena Teh
Partner
Dechert LLP
Hong Kong

Application for accreditation of this program for Continuing Professional Development (CPD) in the United Kingdom is currently pending. 

For questions, please contact Reiko Tate (reiko.tate@dechert.com).

By Jennifer D. Newberger, Anne K. Walsh & Jeffrey N. Gibbs –

The recent issuance of FDA’s draft guidance, Deciding When to Submit a 510(k) for a Change to an Existing Device, which, when finalized, will replace the guidance of the same name issued in January 1997, was undoubtedly met with both excitement and trepidation for companies with products marketed under a 510(k) clearance. Trepidation because in 2011, FDA issued a controversial draft guidance on this issue that would have significantly increased the circumstances under which a 510(k) would need to be submitted for a modification. We blogged on that draft guidance here.  And excitement because not only is an update to the 1997 guidance well overdue, but also because in 2012 Congress held FDA accountable for the positions it took in the 2011 draft, required FDA to withdraw that draft and operate under the 1997 guidance, and mandated that FDA provide to Congress a report describing when a new 510(k) is needed for a modification to a cleared device.  FDA is not permitted to finalize the draft guidance for one year after receipt of the report by Congress.

With this background in mind, medical device enthusiasts likely reviewed the draft guidance eager to see how this revised draft modifies the 1997 guidance, looking at it from the regulatory perspective of when a new 510(k) is needed. (Stay tuned for a follow-up post providing an in-depth analysis from this viewpoint.) But while the details will be important, companies should not consider them as solely affecting regulatory decisionmaking. In fact, the consequences could be far more significant: the failure to obtain a new 510(k) for a modification could expose a company to civil and criminal charges brought by the Department of Justice (DoJ).

FDA has often disagreed with a company’s determination that no new 510(k) is needed for a modification to a cleared device, including a labeling change. Such disagreement was usually handled by negotiations between FDA and the company, and the company’s agreement to submit a 510(k) within a specified timeframe. Often, FDA did not require the company to take the product off the market during the pendency of the review, so long as there was no known safety issue associated with use of the modified device, and the company kept up its end of the deal to timely submit a 510(k). It was strictly a regulatory issue, and handled appropriately as such.

Now, however, a different approach is afoot, and one that carries with it the potential for far more serious consequences than the need to obtain a new 510(k) clearance. For those of you who are regular readers of this blog, you know that FDA has suffered several significant losses with respect to its attempts to regulate companies that promote their products for off-label uses (see our previous posts here, here, and here). Those cases have demonstrated the power of the First Amendment, and that companies may distribute truthful, non-misleading information about off-label uses of their products without running afoul of the Federal Food, Drug, and Cosmetic Act.

To counter this new defense, the government has adjusted the way it views an off-label promotion violation for medical devices. Typically the government has considered a company engaged in off-label promotion as creating a new intended use for its product, and has alleged that the conduct misbrands the drug or device because the product does not bear adequate directions for its “off-label” intended use. (The government also has alleged that the product’s labeling is false or misleading, which of course would not be subject to First Amendment protection.) Now, to avoid the First Amendment “issue,” the government is taking a new tack: charging medical device companies engaged in off-label promotion with a failure to have a new 510(k), and criminalizing what has traditionally been considered a regulatory violation. The difference is subtle but important, by trying to shift the allegation from speech to conduct—the lack of a new 510(k).

This approach has recently been tested in the cases of Vascular Solutions, Inc., and Acclarent, Inc., with differing outcomes. In Vascular Solutions, the government alleged that the company’s “Vari-Lase” product line could only be promoted for the treatment of superficial veins, and that the company’s promotion for the ablation of “perforator” veins, which connect the superficial vein system to the deep vein system, was outside the scope of the cleared indication. The government charged the company and its CEO with one count of conspiracy and eight counts of introducing adulterated and misbranded devices into interstate commerce. Before trial, the government dropped the adulteration charges, leaving only the misbranding charges related to the company’s failure to have a 510(k) clearance for the perforator use and the typical “adequate directions for use” charge. The company’s defense hinged on whether the specific use for perforator veins was within the general indication cleared by FDA. The jury instructions appear to follow the lines of the 1997 Guidance:

A device is misbranded if the manufacturer was required to, but failed to provide, a 510(k) notification or information related to that notification as required by law before distributing the device. When a device is already in commercial distribution but is about to be significantly changed in its design or intended use, the manufacturer must submit a new 510(k) notification to the FDA at least 90 days before the manufacturer introduces the modified device into interstate commerce. Significant changes that require a premarket notification include: (1) a change in the device that could significantly affect the safety or effectiveness of the device (e.g., a significant change or modification in design or manufacturing process); or (2) a major change in the intended use of the device. The 510(k) notification must include appropriate supporting evidence to show that the manufacturer has considered what consequences and effects the changes might have on the safety and effectiveness of the device. A 510(k) notification also must contain proposed labeling sufficient to describe the device, its intended use, and the directions for its use. If a 510(k) notification does not contain sufficient information, the FDA may request additional information. If the additional information is not provided within 30 days of the request, the 510(k) notification is considered withdrawn.

After a four-week trial, the jury acquitted the company and the CEO of all counts.

The executives in the Acclarent matter, however, when faced with the same charges for similar conduct, were not as fortunate. In that matter, the government alleged that the Relieva Stratus Microflow Spacer was cleared to be used with saline to maintain sinus openings following surgery, but that the company promoted the device to deliver steroids to the sinus cavity. The government charged the company’s executives with “Misbranding by Failure to File Premarket Notification,” and the court instructed the jury that “a medical device is [ ] misbranded if the manufacturer introduces the device into interstate commerce for an intended use that is significantly different from the use covered by its 510(k) clearance and without submitting a new premarket notification to the FDA regarding the different intended use.” Although the jury acquitted the company’s executives of all felony charges based on a lack of evidence of intent to defraud or mislead, the executives were convicted of 10 misdemeanor counts related to the regulatory status of these devices. The company—which is now owned by J&J— also agreed to pay $18 million to resolve civil allegations related to the same conduct under the False Claims Act.

In light of the above approach to civil and criminal matters, companies now need to consider not only whether FDA could deem the failure to file a new 510(k) for a labeling change to be a regulatory violation that could be remedied through appropriate regulatory mechanisms, but also whether the failure to file could land the company with civil and/or criminal subpoenas from DoJ. Perhaps as a result of the loss in Vascular Solutions, FDA includes in the new draft guidance a specific example of when a more specific use may require a new 510(k), while the 1997 guidance was silent on that issue. Knowing this, medical device manufacturers should review the draft guidance not only to inform their regulatory activities, but also knowing that DoJ may come knocking. And, as companies prepare their comments on the draft guidance, they should be aware that the stakes could be much higher than receiving a Warning Letter.

By Etan J. Yeshua –

Last week, FDA issued for the dietary supplement industry a draft guidance document that largely doubles-down on controversial positions the Agency has previously taken. The draft guidance – “Dietary Supplements: New Dietary Ingredient Notifications and Related Issues” – replaces the draft of the same name that was issued in July 2011. As we blogged about at the time (here and here), FDA’s 2011 draft guidance reflected a regulatory approach that many considered to be inconsistent with the requirements and intent of the Dietary Supplement Health and Education Act (“DSHEA”), which amended the Federal Food, Drug, and Cosmetic Act (“FD&C Act”). The 2011 draft elicited a host of comments (including two from us, here and here) about key issues in FDA’s approach to implementing DSHEA. The “updated” draft guidance is likely to elicit much the same criticism, as it largely reiterates the Agency’s previous positions on key issues (which we describe below).

In a Federal Register notice announcing the updated draft guidance, FDA identified the “major topics on which we have revised or added questions and answers,” including:

• manufacturing changes that create a new dietary ingredient (“NDI”);
• synthetic substances;
• the definition of “chemical alteration”;
• the definition of “NDI” and the list of “grandfathered” dietary ingredients;
• structuring notifications efficiently and relying on data from prior notifications and master files;
• identity information to include in an NDI notification;
• electronic submission;
• PDF form for NDI notifications submitted on paper; and
• safety information to include in an NDI notification.

For the most part, however, the Agency’s revisions and additions restate controversial positions that FDA took in the 2011 draft guidance—positions about which the Agency received numerous comments from industry and Congress. These positions include:

• “Grandfathered” ingredients. The Agency narrowly construes the category of “grandfathered” (or “old”) dietary ingredients, i.e., those that are not NDIs, and therefore do not require NDI notification, because they were marketed in the United States before October 15, 1994. In the updated guidance, FDA maintains its position that, in order to be grandfathered, an ingredient must have been marketed before that date as a dietary ingredient (rather than, e.g., as a conventional food or an over-the-counter drug) further narrowing the class of “grandfathered” ingredients. FDA also continues to maintain that certain changes in manufacturing methods may cause an otherwise grandfathered ingredient to become an NDI.
• In addition, FDA states in the new draft guidance that it is open to compiling a (non-exclusive) list of grandfathered ingredients, but that the burden for establishing the grandfathered status of an ingredient would largely fall to industry: “[W]e are prepared to develop an authoritative list of [grandfathered] ingredients, based on independent and verifiable data. Because FDA does not generally have access to marketing records for dietary ingredients and dietary supplements, the documentation of pre-DSHEA marketing would have to be supplied by industry. . . . The mere fact that an ingredient is not on the list would not, however, establish that the ingredient is an NDI . . . . Rather, the omission of an ingredient from the list would be regarded as neutral and would not affect the ingredient’s regulatory status.” (Emphasis in original).
• Definition of “dietary ingredient.” Section 201(ff)(1)(E) of the FD&C Act includes as a dietary ingredient “a dietary substance for use by man to supplement the diet by increasing the total dietary intake.” FDA maintains its position that this category includes only substances that are “commonly used as human food or drink,” thereby excluding many novel substances, as well as many dietary substances that are not “commonly used,” from inclusion in a dietary supplement.
• NDIs exempt from the notification requirements. The FD&C Act exempts from the notification requirement NDIs that are “present in the food supply as an article used for food in a form in which the food has not been chemically altered,” and FDA continues to interpret this exemption narrowly. For example, the updated draft guidance again considers an NDI to have been “present in the food supply” only if it is present as an ingredient in conventional foods; presence in a dietary supplement would not exempt the ingredient from the notification requirement. 
• A new concept in the draft guidance is the option to submit a confidential “NDI master file.” The master file would contain “the manufacturing, specifications and other identity information needed to completely describe the ingredient” and may be incorporated by reference into later NDI notifications.
• Responsibility for submitting NDI notifications. FDA maintains its position that, except in limited circumstances, a dietary supplement manufacturer must submit a separate notification for each supplement that contains an NDI and generally cannot rely on a previously-submitted notification. The Agency again asserts that a notification must be submitted for a dietary ingredient even if the ingredient’s manufacturer or another supplement manufacturer previously submitted a notification for that ingredient, except when virtually all conditions of use are identical (including, for example, that “[t]he new supplement does not combine the NDI with other dietary ingredients that were not included in [the] original NDI notification,” the target consumer population is the same, etc.).  
• Synthetic ingredients. The Agency continues to consider most synthetic copies of botanicals (and synthetic constituents and extracts of botanicals) as falling outside the definition of “dietary ingredient” and, thus, as impermissible in dietary supplements, even if the synthetic ingredient is chemically identical to the natural ingredient.
• Although the new draft guidance explicitly states that “synthetic vitamins, minerals, and amino acids qualify as dietary ingredients . . . regardless of [their] source,” it reiterates that “a synthetic copy of an herb or other botanical does not qualify as a dietary ingredient under section 201(ff)(1)(C) of the FD&C Act” and that “[f]or more than a decade, FDA has consistently interpreted section 201(ff)(1)(F) of the FD&C Act as not including synthetic copies of botanical constituents, extracts, and concentrates.”
• The new draft guidance provides a limited exception: “a synthetic copy of a botanical ingredient [or constituent] may qualify as a dietary ingredient under section 201(ff)(1)(E) if the synthetic copy has been used as a lawfully marketed ingredient in the conventional food supply.”  In other words, if a synthetic botanical or synthetic botanical constituent is on the market in conventional foods (e.g., vanillin), then that synthetic ingredient could be used as a dietary ingredient in a dietary supplement.  Of course, in order to lawfully be on the market in a conventional food, the ingredient would have to be either an FDA-approved food additive or Generally Recognized As Safe (“GRAS”) in accordance with FDA regulations.

FDA has stated that comments to the draft guidance should be submitted by October 11, 2016 in order to ensure that the Agency considers them before it begins work on the final version of the guidance.