Try a Demo

spinkerton

by

Nitrous oxide faces EU consumer ban under REACH proposal

By Scott Stephens, MPA

The European Commission is advancing a proposal to restrict consumer sales of nitrous oxide under the REACH Regulation, citing growing concerns over its misuse as a recreational drug. The measure would add the substance to Annex XVII, effectively banning consumer access while preserving limited uses, including certain food-related applications.

The proposal includes exemptions for controlled uses, along with restrictions on cartridge size, sales volume and age verification requirements. Member States are expected to vote on the measure in the coming weeks.

This AgencyIQ analysis examines the proposal’s regulatory implications and what it could mean for companies operating across EU chemical and food markets.

Fill out the form to read the full analysis.

Required *

by

Nitrous oxide faces EU consumer ban under REACH proposal

The European Commission is moving to restrict consumer sales of nitrous oxide under the REACH Regulation, aiming to curb its abuse as a recreational drug while preserving some food-related uses. Member States are expected to vote in April on a proposal that would add the gas to the regulation’s Annex XVII.

By Scott Stephens, MPA | Mar 18, 2026 2:55 PM EDT

Regulatory background

Nitrous oxide: Identity, uses, EU regulation

  • Nitrous oxide, or dinitrogen oxide, is a naturally occurring gas with a slightly sweet odor and taste that is colorless and nonflammable at room temperature. It is commonly known as “laughing gas” because of the euphoric effects of inhaling it, a quality that has led to its being abused as a recreational inhalant drug.
  • The compound is used in a wide range of applications. These include as an anesthetic in dentistry and surgery, a propellant gas in food aerosols such as whipped cream, an oxidant in organic chemical production, a component in rocket fuel formulations, and an agent to detect leaks in natural gas pipelines.
  • In the EU, nitrous oxide is regulated under several regulatory frameworks, depending on use or application. Under REACH, the EU’s cross-sectoral chemicals law to which all substances are subject with some exceptions, nitrous oxide is registered in the 1,000-10,000 annual tonnage band. Uses associated with the compound’s active REACH registrations include as a reactive processing aid in the manufacture of computers, electronic and optical products, and electrical equipment; a refrigerant gas in cooling systems; an oxidizer in engines; and a propellant in spray cans for technical applications.
  • Nitrous oxide has been classified under the CLP Regulation. In June 2025, Regulation (EU) 2025/1222 added an entry for nitrous oxide to the CLP’s Annex VI after the EU identified the substance as having several hazardous properties. Its Annex VI harmonized classifications are reproductive toxicity, Category 1B; specific target organ toxicity, single exposure 3; STOT, repeated exposure 1 (causing damage to the nervous system through prolonged or repeated exposure); ozone 1 (harms public health and the environment by destroying ozone in the upper atmosphere).
  • The substance is authorized in the EU for use as a food additive under Annexes II and III of the EU’s Food Additives Regulation. Nitrous oxide, assigned E942 under the EU’s nomenclature identifying permitted food additives, is allowed on the market and may be used in all food categories, in food additives, in food enzymes and in food flavorings at quantum satis, or the smallest amount necessary to achieve the desired result, rather than a strictly fixed quantity.
  • Nitrous oxide is regulated under the EU’s medicinal products framework. Falling under Directive 2001/83/EC’s Article 1(2) definition of a medicinal product, the compound is subject to obtaining market authorization before it may be placed on the market in any Member State.

Commission moves to restrict nitrous oxide

  • The EU executive introduced a draft regulation to amend REACH’s Annex XVII restrictions for discussion during the Feb. 25-26 meeting of the REACH Committee. The proposal includes adding nitrous oxide to Appendix 6 – the list of 1B reproductive toxicants associated with Annex XVII’s Entry 30 banning the supply of these substances to consumers. The committee is composed of Member State representatives and assists the Commission in deciding on REACH restrictions.
  • The proposal includes two exemptions under certain conditions to allow nitrous oxide’s use as a food additive “by the general public” as well as to allow for its presence in food, according to Recital 9 of the proposal. The authors explain that they’ve proposed conditions to “minimise the possibility of misuse,” including the sale of cartridges containing no more than 18 grams for consumer use as a food additive. Furthermore, the supply of these cartridges is limited to a total maximum content of 90 g per day, per individual. Likewise, the proposal conditions the sale of the cartridges on verifying the buyer is at least 18 years old, “checked by an effective age verification system.” In this way, the Commission intends to prevent would-be exploitation of the exemption for the sale of the substance as a drug. Second, the proposal stipulates that nitrous oxide may be used directly in food as well as in aerosols for food products, “where the substance is used as a propellant, and the dispenser is designed for use of the substance as food additive.” Here, the proposal justifies the exemption by stating that aerosol dispensers containing nitrous oxide “do not provide ready access to the substance and are therefore unlikely to be misused.”
  • Finally, the proposal specifies that any stricter measures established at the national level aimed at combating abuse of the substance as a drug would preempt these exemptions.

Next steps and analysis

  • WOPKE HOEKSTRA, the commissioner for climate, net zero and clean growth, discussed the proposed restriction with members of European Parliament at their March 12 plenary session. Hoekstra told MEPs in his address that discussions in the REACH Committee “are advancing well,” with a vote on the proposal expected at the upcoming April meeting of the committee. If it signs off on the proposed regulation, it will go before the Parliament and the Council of the EU, who will then have three months to scrutinize it. If neither objects, the regulation will be adopted and published in the Official Journal. The proposed restriction would become applicable starting Feb. 1, 2027.
  • The commissioner stated the importance of establishing the proposed restriction, considering the rise of nitrous oxide’s abuse as a recreational drug across the EU. In this context, he underscored that large canisters – containing up to several kilograms of the substance – “are being specifically placed on the market for the purpose of enabling abuse.” He said they are labeled as “allegedly serving for food purposes,” but that there’s “no evidence whatsoever” that they serve legitimate purposes.
  • Hoekstra also highlighted that several Member States “have already taken steps to address the issue through their national drug legislation” and that the Commission “very much welcomes exactly these efforts.” He emphasized that despite these actions, the sale of “these illegal and falsely-labelled canisters” seems to continue, highlighting that “the abuse of nitrous oxide is a Union-wide issue, which also requires a Union-wide approach.”
  • For the most part, MEPs voiced support for the Commission’s proposed measure during the debate. MARLENA MALĄG of the right-wing European Conservatives and Reformists group summed up succinctly what most MEPs expressed, saying “a ban on the sale of nitrous oxide to private individuals is a sensible and proportionate solution. It ensures continued access for hospitals and industry while closing the loophole that enables abuse.” The center-right European People’s Party’s NINA CARBERRY went further, urging the Commission to recognize nitrous oxide as an emerging drug of concern and saying the restriction needs to happen sooner. “Waiting until 2027 is simply too slow,” she said.

To contact the author of this analysis, please email Scott Stephens ( sstephens@agencyiq.com).
To contact the editor, please email Jason Wermers ( jwermers@agencyiq.com).

Key documents and dates

by

How stakeholder feedback might shape a federal definition of ultra-processed foods

By Scarlett Salem, MS, MPH and Jared Rothstein

The FDA and USDA are moving toward establishing a federal definition of ultra-processed foods, with a proposal expected in the coming months. Public comments submitted in response to the agencies’ request for information reveal broad interest in defining UPFs — but little consensus on how that definition should be structured.

Stakeholders remain divided over whether to rely on processing-based frameworks like NOVA, incorporate nutrient criteria, or develop hybrid approaches grounded in labeling feasibility and existing nutrition policy. Many also raised concerns about unintended consequences for food access, labeling, and federal nutrition programs.

This AgencyIQ analysis examines key themes from stakeholder feedback and what they signal for the direction of a federal UPF definition.

Fill out the form to read the full analysis.

Required *

by

New FDA guidance encourages NAMs adoption, but keeps advice broad

By Ned Pagliarulo

The FDA has issued new draft guidance aimed at accelerating adoption of new approach methodologies (NAMs) as alternatives to animal testing in drug development. The guidance outlines general validation principles and encourages sponsors to incorporate NAM data into regulatory submissions.

While the document reinforces the agency’s broader push to reduce reliance on animal studies, it provides relatively high-level recommendations and leaves key questions around implementation, validation standards, and regulatory expectations unresolved.

This AgencyIQ analysis explores how the guidance fits into FDA’s evolving approach to nonclinical testing and what it means for sponsors considering NAM adoption.

Fill out the form to read the full analysis.

Required *

by

New FDA guidance encourages NAMs adoption, but keeps advice broad

FDA Commissioner MARTIN MAKARY has sought to accelerate adoption of new approach methodologies over animal testing. Newly issued recommendations could help advance that goal but may leave sponsors with unanswered questions.

By Ned Pagliarulo | Mar 18, 2026 7:17 PM EDT

Quick background on new approach methodologies

  • For many regulated medical products, sponsors must conduct nonclinical studies before beginning human testing. These studies are typically designed to demonstrate that a product is reasonably safe and to characterize its pharmacological effects and mechanism of action. This helps ensure the safety of subsequent clinical trial participants while providing some proof of concept for the product. Historically, nonclinical investigations have relied on animal testing. Researchers are working to develop alternative methods in response to the practical and ethical drawbacks of animal studies, however. As the FDA has explained: “Animal studies can be critical to evaluating the potential for toxicity, safety and, at times, effectiveness, of FDA-regulated products. However, premarket animal-based assessments are time and resource intensive and may not always fully predict or detect potential concerns of FDA-regulated products for proposed uses in humans or animals.”
  • New approach methodologies, or new alternative methodologies, are meant to replace, reduce or refine use of animal testing in medical product development. The FDA considers NAMs to include “a range of approaches, including AI-based computational models of toxicity and cell lines and organoid toxicity testing in a laboratory setting.” More specifically, NAMs can include engineered tissues formed using scaffolds and microphysiological systems such as organs-on-chips, but also chemical reactivity studies and in silico modeling.
  • The FDA has been encouraging NAM development for years. The agency stood up an Alternative Methods Working Group in 2019 to advance its Predictive Toxicology Roadmap, which sought to lessen animal testing. Two reports in 2021 further detailed the agency’s efforts around and investments in alternative methods. The next year, Congress updated statutory language allowing flexibility for nonclinical testing besides animal studies. And in 2024, the agency’s Science Board advanced six recommendations to boost adoption of NAMs.

Regulatory context: FDA’s new roadmap on NAMs

  • FDA Commissioner MARTIN MAKARY has prioritized NAMs. In April 2025, Makary announced a new roadmap to “phase out animal testing requirements.” As AgencyIQ described at the time, the roadmap singled out monoclonal antibodies, or mAbs, as a “promising area for reducing animal use in preclinical safety testing,” while outlining plans to “expand to include other biological molecules and eventually new chemical entities and medical countermeasures.” The FDA has required toxicology studies in animals for mAbs, although the agency wrote that “animal immunogenicity is not predictive of human immunogenicity,” and “some safety risks may go undetected in animals” while “animal testing of mAbs poses practical challenges.”
  • The roadmap outlined six activities the FDA plans to undertake through 2028 to “make animal studies the exception rather than the norm for pre-clinical safety/toxicity testing.” First, the agency said it will “explore pre-existing international data,” in which firms may collect and provide human toxicity data from other countries, if available, and allow manufacturers to submit these data in investigational new drug applications. Second, the FDA intends to “encourage sponsors to submit NAM data in parallel with animal data to build a repository of experience.” This may include “regulatory relief” for sponsors that voluntarily provide NAM data, as well as a series of pilot cases. Third, the agency will “develop an open-access repository with a comprehensive collection of international drug toxicity data from animals and humans.” Fourth, the regulator intends to “reduce the routine 6-month primate toxicology testing for mAbs that show no concerning signals in 1-month studies plus NAM tests to three months.” Fifth, the FDA will seek to expand this model for other types of drugs. And finally, the agency will track and quantify toxicity testing biannually to better understand the cost, science and safety of nonclinical testing.
  • The FDA took a step toward delivering on the fourth aim in December, publishing draft guidance that outlines “streamlined approaches” to nonclinical studies for developers of mAbs. The guidance explains that nonrodent animal studies longer than three months are “not warranted to evaluate toxicities from chronic administration of monospecific antibodies” when data are supplemented with a weight-of evidence risk assessment. Only in “exceptional cases” does the agency anticipate this combination to be inadequate for its purposes. And in some specific circumstances, even three-month toxicology studies may not be needed, according to the guidance. [Read AgencyIQ’s full analysis of the draft guidance here.]

New guidance offers “general considerations” for NAMs’ use

  • The FDA gave drug developers a fuller sense of its views on NAMs with draft guidance published March 18, 2026. Titled “General Considerations for the Use of New Approach Methodologies in Drug Development,” the guidance provides general background and describes the Center for Drug Evaluation and Research’s recommendations for validating NAMs. “This guidance will facilitate the adoption of modern alternatives to animal testing in regulatory submissions,” Makary said in an agency statement.
  • The guidance clarifies that validation, in this context, is not the same as qualification, which the FDA associates with its existing program to establish drug development tools and addresses in other guidance. As a concept, validation indicates a process to confirm a procedure’s accuracy, reliability and relevance for a specific context of use. Qualification, on the other hand, describes the FDA’s determination that a drug development tool can be depended on for a specific regulatory interpretation and application within its context of use. However, the guidance indicates that validation is not necessarily required for a NAM to be considered for review. “A fit-for-purpose NAM, even if not validated, may adequately address specific toxicological concerns,” the agency wrote.
  • The guidance describes four “key features” of any approach meant to validate NAMs. Sponsors are encouraged to outline a NAM’s context of use, human biological relevance and technical characterization. NAMs should also be “fit-for-purpose,” which describes whether they can aid the FDA’s decision-making.
  • Context of use is a well-established regulatory concept. In this setting, a COU “should clearly define the intended use and regulatory purpose of the NAM to address a specific drug development decision context.” The guidance provides five examples of how a NAM’s COU can be used to support regulatory decision-making, including supporting dose selection, addressing an adverse event’s mechanism and justifying not using an animal species.
  • A NAM’s biological relevance is determined by how well the information it produces ends up impacting a drug’s assessment in human testing. “Adequate demonstration of the biological relevance of a NAM provides confidence in the methodology,” the guidance states. The FDA advised that sponsors describe the physiological features a NAM will assess, demonstrate how toxicological findings can be reliably evaluated, and describe how the biological mechanism studied via the NAM applies to clinical outcomes. For instance, a NAM designed to evaluate liver toxicity should contain relevant cell types, such as hepatocytes, and biological features that mimic liver physiology. A sponsor using this hypothetical NAM, the guidance adds, could then indicate whether it accurately measures liver markers such as albumin to simulate a repeated drug exposure scenario.
  • Technical characterization of a NAM, the guidance states, is “essential to establish scientific confidence” in the data the NAM produces. The guidance provides a list of high-level recommendations describing what sponsors should document and share with the FDA. These range from statistical methods to data showing the predictive performance of a NAM. Some are particular to certain types of NAMs, such as a range of characteristics for organ chips. These recommendations mirror the Organization for Economic Cooperation and Development’s guidance on Good In Vitro Method Practices, the FDA’s guidance notes.
  • Sponsors attempting to show a NAM is fit for purpose should consider three general points, according to the guidance. When traditional animal tests are available, sponsors should provide data that demonstrate a NAM can characterize risk “at least as well as” the established method, including on standard metrics such as specificity and sensitivity. Sponsors should also detail the benefits and limitations of a NAM and describe how resulting data contribute to the “overall determination of risks to human health and/or safety of the drug.”
  • Finally, the guidance outlines three drug development objectives that a fit-for-purpose NAM should accomplish. NAMs could be used to replace or provide alternatives to traditional methods, such as substituting an in vitro NAM for an animal study. NAMs might also fill a data gap by providing activity or safety information where traditional methods aren’t useful or available. They could also serve to confirm or complement findings from traditional techniques, such as by justifying why a specific animal species does not provide useful data.

Analysis and what’s next

  • As the title suggests, the guidance is more general than specific. About half of the document is dedicated to background information, while the recommendations outlined in the second half have to do with “general validation principles.” While the FDA provides modestly detailed examples for how a sponsor may want to demonstrate the biological relevance of a NAM, it offers comparatively less detail in examples of the three drug development objectives outlined later in the guidance. The guidance also does not attempt to address specific NAMs or how NAMs might be used in drug discovery, which may leave sponsors with unaddressed questions. The agency appears to expect this, as it encourages sponsors to reach out to review divisions if they are uncertain whether a NAM is suitable for regulatory use. “We anticipate that early engagement will focus on indication-, disease-, organ-, and endpoint-specific considerations that should include interactions with review divisions,” the guidance states.
  • The recommendations contained in the guidance are CDER’s. The Center for Biologics Evaluation and Research, which oversees cell and gene therapies and blood products, is not listed on the document’s title page. A footnote clarifies that the term “drug” or “drug product,” when mentioned in the guidance, refers to “products regulated by CDER, unless otherwise specified.” CDER’s leadership on this issue is expected, given its broad remit and role in reviewing products, such as monoclonal antibodies, that may be most amenable to incorporating NAMs into development. (CDER also maintains a list of settings in which streamlined nonclinical programs are permissible.) But CBER has participated in cross-agency work on NAMs, so its omission from this guidance is notable.
  • The guidance sets an encouraging tone. “NAMs hold great potential in terms of enhancing efficiencies in drug development and generating data to advance human health,” the FDA wrote. The guidance urges sponsors to include NAMs data in regulatory submissions and, when those methods have proven accurate, to adopt them. The FDA’s press release indicates that it’s already seen “several examples” of validated NAMs performing better than unvalidated animal models to predict drug responses. The tone matches the high priority HHS has put on shifting toward nonanimal studies. Simultaneous to the guidance’s release, the NIH announced more than $150 million in new grants to “reduce reliance on animal models.” These grants will fund “technology development centers” to advance NAM development in gynecological, cardiac, neurological and rare diseases. Quoted in the NIH’s release is NICOLE KLEINSTREUER, deputy director for program coordination, planning and strategic initiatives, who’s listed as first author on research on skin sensitization NAMs cited in the FDA’s guidance.
  • Qualification is not emphasized, however. The guidance makes clear that CDER will look at both qualified and validated NAMs and determine whether their use is appropriate. “While qualification is useful, validation is critical to establishing the reliability of NAM data for specific situations,” the FDA wrote. Qualification through the Drug Development Tools program is slow and, at least in two other settings, has proven challenging for sponsors. The guidance suggests that CDER is more interested in advancing NAMs via interactions between review divisions and sponsors.
  • What’s next? The FDA is currently studying more granular questions tied to NAMs development. CDER’s Division of Applied Regulatory Science, which researches “challenging scientific questions” that impede regulatory review, indicated in a recent annual report that it plans to take on two new projects related to NAMs. The first involves cardiac NAMs for assessing drug combinations, and the other features a neural NAM to examine certain opioid challenges. [Read AgencyIQ’s full analysis of DARS’s annual report here.] Feedback from industry on the FDA’s draft guidance, meanwhile, may indicate sticking points to NAMs adoption. Comments are due in mid-May.

Featuring previous research by Laura DiAngelo.

To contact the author of this item, please email Ned Pagliarulo ( npagliarulo@agencyiq.com).
To contact the editor of this item, please email Jason Wermers ( jwermers@agencyiq.com).

Key documents and dates

by

How stakeholder feedback might shape a federal definition of ultra-processed foods

The Trump administration has signaled that a federal definition of ultra-processed foods will be issued this spring. Here, AgencyIQ analyzes public comments on the FDA and USDA’s request for information to identify UPF themes, recommendations and possible frameworks offered by food stakeholders.

By Salem Scarlett, MS, MPH and Jared Rothstein | Mar 20, 2026 4:08 PM EDT

Quick background on ultra-processed foods

  • Ultra-processed food has emerged as a topic of public health concern. UPFs are generally described as foods that undergo a high degree of industrial processing, contain a significant number of additives and have high levels of nutrients of concern such as saturated fat, sodium and added sugar. Other characteristics often cited in public discourse, policymaking and research are that UPFs have a long shelf life and high energy density and are ready to eat and highly palatable. Acknowledging limitations and evidence gaps, research has found associations between UPFs and obesity and diet-related disease. However, documented associations often depend on the UPF subgroups being evaluated, with sweetened drinks, processed meats and ready-to-eat meals linked to higher health risks, while cereals, whole grain breads and yogurt are associated with beneficial health outcomes.
  • Currently, there is not a uniform approach for defining UPFs. The most frequently cited scientific approach for defining UPFs, particularly for nutrition research, relies upon the NOVA classification system, which categorizes foods based on the extent and purpose of their industrial processing. Other scientific approaches focus on the presence of additives, ingredient lists, nutritional quality and other markers. Some classification methods also attempt to quantitatively score foods based on these cumulative factors. [Read AgencyIQ’s analysis of current approaches for defining UPFs here.]
  • Federal regulators have taken steps to establish a uniform definition. On July 25, 2025, the FDA and USDA issued a joint request for information seeking input on factors and criteria that should inform a standardized federal definition. The agencies wrote in their accompanying press release that a clear definition established by the federal government is “a critical step in providing increased transparency to consumers about the foods they eat” and will inform research and policy activities “to pave the way for addressing health concerns associated with the consumption of ultra-processed foods.” The comment period for the RFI closed on Oct. 23. [Read AgencyIQ’s analysis of the FDA and USDA’s request for information on UPFs here.]
  • Stakeholder feedback from the RFI will likely shape and influence the FDA and USDA’s proposal. AgencyIQ took a deep dive into the 5,000-plus comments in the regulations.gov public docket to identify common themes, recommendations and points of interest across stakeholders and interest groups. Those takeaways are summarized below.

Key theme from stakeholder feedback: Disagreement over using NOVA as the foundation

  • One of the clearest divides in the docket was whether the NOVA classification system should serve as the basis for a federal UPF definition. Some public health and consumer groups, including the Consumer Federation of America, Crohn’s and Colitis Foundation, and Tufts University’s Food is Medicine Institute, argued that NOVA should form the foundation because it underpins much of the existing research and provides a common framework for research and communication. The Consumer Federation of America wrote that “research on diet-related disease outcomes fairly equates the term ‘ultra processed food’ with the Nova classification system.” These commenters stated that adopting NOVA would allow regulators, researchers and clinicians to operate from a shared baseline rather than creating a new definition that could fragment the evidence base.
  • However, many other stakeholders cautioned that NOVA may not be suitable for regulatory use. Groups including the Academy of Nutrition and Dietetics, School Nutrition Association, U.K. Department of Health and Social Care, New York City’s Health Department, and several food science and industry organizations said the framework is too broad, subjective and difficult to operationalize using current labeling systems. These commenters wrote that NOVA groups foods with very different nutritional profiles into the same category and relies on processing information that is often not available to regulators or consumers. Several also said its application could lead to inconsistent classification outcomes across studies and policy contexts. The Academy wrote, “Defining UPFs remains a challenge due to limited evidence identifying specific aspects of UPFs that may be responsible for adverse health effects.”

Broad support for incorporating nutrient criteria

  • Across stakeholder groups, many commenters emphasized that nutrient composition should play a central role in any UPF definition. Researchers at the NIH stated that adverse health associations were primarily driven by UPFs high in nutrients of concern, while other UPFs showed neutral or even protective associations. They suggested that regulatory approaches should focus on lower-nutritional-quality UPFs rather than the full category. They presented their findings at a conference in June 2025. A draft of the manuscript reporting these results has been posted to a preprint server, and the paper is currently under peer review, according to the NIH’s comment. “Our findings support an approach that classifies foods based on both processing and nutritional quality, such that only ultra-processed foods of low nutritional quality are subject to regulations or other efforts to reduce their intake,” the researchers wrote.
  • Similarly, groups including the Environmental Working Group, School Nutrition Association, Dairy Council of California and Plant Based Foods Association wrote that relying solely on processing could misclassify nutrient-dense foods. As the Dairy Council of California put it, “Not all processed foods are created equal, and some are shown to be beneficial to health.” Frequently cited examples include yogurt, whole grain breads and cereals, fortified foods, and plant-based alternatives that may be categorized as UPFs but still align with dietary recommendations.
  • Many commenters also pointed out that foods commonly categorized as UPFs often overlap with those already identified in nutrition policy as high in fat, sugar or salt. This overlap led some to say that existing nutrient-based frameworks may already capture the foods most associated with health risks. However, there was little consensus on how nutrient criteria should be incorporated. Some groups, such as the Center for Science in the Public Interest, recommended specific thresholds based on models used in Chile or by the Pan American Health Organization, while others urged alignment with the Dietary Guidelines for Americans without adopting new numeric cutoffs.
  • At the same time, a smaller group of commenters cautioned against incorporating nutrient criteria at all, arguing that doing so could undermine the purpose of a processing-based definition and allow reformulated products to meet thresholds while remaining highly processed. “The use of nutrient criteria provides a loophole to industry, who will create new ‘Franken-food’ UPF that meet the nutrient thresholds but have little nourishing value,” the Food is Medicine Institute wrote. “This has been clearly seen, for example, in school meals, which meet USDA nutrient criteria but are too often UPF contributing to the epidemic of chronic disease in American children.”

Concerns about unintended consequences for nutrition programs and access

  • Many commenters warned that a broad UPF definition could have unintended consequences if applied to food assistance programs, school meals or specialized nutrition products. Health care nutrition and infant formula groups said products such as medical foods, infant formula and oral nutrition supplements are intentionally formulated to meet specific medical or life-stage needs. As these groups wrote, such products are “intentionally formulated and processed to provide nutrition in a form that is tailored to the consumer’s or patient’s specific dietary needs and/or health condition.”
  • School nutrition stakeholders raised similar concerns, noting that some processed or prepared foods are operationally necessary in K-12 settings because of constraints related to staffing, equipment and cost. They warned that misclassifying these foods could reduce menu options or increase costs for school meal programs.
  • Produce stakeholders also warned that minimal processing used to preserve fruits and vegetables could inadvertently trigger UPF classification. They said practices such as washing, coating or packaging are used to maintain freshness and reduce waste but could be misunderstood or misclassified under a processing-based definition.
  • Several commenters also articulated broader concerns about food access, affordability and equity, noting that overly restrictive definitions could limit availability of affordable and shelf-stable foods for vulnerable populations. The International Fresh Produce Association urged the FDA and USDA not to classify “value-added fruits and vegetables,” such as bagged lettuce or sliced apples, “as processed simply because they undergo minimal handling that can improve fruit and vegetable consumption. Such misclassification based on any level of processing could inadvertently restrict access to nutrient-dense produce and undermine dietary guidance.” Taken together, these comments suggest that regulators may face pressure to incorporate carveouts, phased implementation timelines, or program-specific exceptions.

Skepticism about the scientific basis for a UPF definition

  • Several industry and food science organizations questioned whether the current evidence base is sufficient to support a regulatory definition. Groups including the American Frozen Food Institute, International Food Additives Council, Institute of Food Technologists and Institute for the Advancement of Food and Nutrition Sciences argued that existing classification systems rely heavily on ingredients or formulation as proxies for processing and lack clear causal evidence linking processing itself to adverse health outcomes.
  • The American Frozen Food Institute stated that “the lack of scientific agreement and evidence indicates that establishment of any regulatory definition to inform policy is premature and futile at this time.” Similarly, IAFNS highlighted the need for classification systems to rely on “properties that are empirically testable and linked to health-related endpoints,” suggesting that current approaches may not meet that standard.
  • These commenters wrote that much of the existing research is observational and may not isolate processing as an independent driver of health outcomes. They generally supported continued focus on nutrient density, food groups, and overall dietary patterns rather than establishing a standalone processing-based classification.

Calls for practical, label-based definitions

  • A number of commenters asserted that any definition must use information available on food labels. Groups such as CSPI, New York City’s Health Department, and Guiding Stars highlighted the importance of using ingredient lists, nutrient disclosures and other label-based information rather than relying on proprietary or undisclosed processing methods.
  • CSPI posited that a definition should “be based on information found on product labels, using evidence-based nutrition, ingredient, and food category criteria.” Similarly, Guiding Stars wrote that the Nutrition Facts Panel and ingredient list provide a “transparent, consistent, and scalable foundation for evaluation without relying on proprietary formulation details or subjective judgments about certain product characteristics like processing.”
  • These stakeholders said definitions tied to measurable and accessible criteria would be more practical for enforcement and consumer understanding. They also wrote that relying on information not currently disclosed on labels could limit the feasibility of implementing a processing-based definition.

Alternative frameworks and international approaches

  • Some commenters suggested that regulators consider alternatives to NOVA or hybrid approaches. Proposals included the IUFoST NOURISH framework, which evaluates both formulation and processing using measurable indicators. Others recommended consumer-facing systems such as traffic light labeling, which categorizes foods based on overall healthfulness rather than processing level.
  • Several commenters also encouraged the agencies to review international policy models. These included Brazil’s use of NOVA in dietary guidelines, Mexico’s front-of-package warning labels, and Australia’s approach to discretionary foods. These examples were cited as potential models or reference points for how a U.S. definition could be structured or applied in practice.

What’s next

  • Taken together, the large number of stakeholder comments reflects broad interest in establishing a federal definition of UPFs but little consensus on how that definition should be structured. The comments highlight several challenges the FDA and USDA will need to address, including balancing scientific consistency with practical implementation, determining whether processing alone is sufficient to identify foods of concern, and ensuring that any definition aligns with existing nutrition policy frameworks. Commenters also repeatedly raised questions about how a definition would be applied in practice, including whether it could be operationalized using existing food labeling, how it would interact with federal nutrition programs, and whether it could create unintended incentives for reformulation or product classification. The most consistent signal across the docket is that regulators may face pressure to avoid a definition based solely on processing intensity.
  • At the same time, the docket suggests that the agencies are unlikely to satisfy all camps with a single simple framework. Commenters who favored NOVA often viewed a processing-based definition as necessary to capture harms they believe are not explained by nutrients alone. Other stakeholders argued just as strongly that nutrient density, food category and intended use must remain central to the definition, especially for products that contribute meaningfully to diet quality or serve important medical, developmental or operational functions. This split suggests that the FDA and USDA may ultimately look for a middle-ground approach that draws from processing-based concepts while incorporating additional criteria or implementation guardrails.
  • The comments also show that different food industries are watching the proposal closely. Meat, dairy, frozen food, plant-based food, produce and specialized nutrition groups did not all recommend the same policy approach, but many shared a similar concern that a broad UPF definition could sweep in foods that are processed for safety, shelf life, fortification, convenience or functionality rather than because they have low nutritional value. That repeated concern may increase pressure on regulators to explain clearly what the definition is – and is not – meant to capture with evidence-based justifications.
  • Another major question is whether the agencies are developing a definition primarily for research consistency, consumer-facing policy or future regulatory use. Some commenters treated the exercise as a threshold scientific question, while others clearly viewed it as a precursor to front-of-package labeling, nutrition standards, procurement policies or school meal restrictions. This distinction matters because a definition that is workable for research may not be easily translated into labeling or enforcement, particularly if it relies on information not currently disclosed on food packages.
  • In the near term, the agencies may face pressure to show that any proposed definition can be operationalized using information already available to regulators, manufacturers and consumers. Multiple commenters pointed to the Nutrition Facts panel and ingredient list as the most practical foundation for implementation, while others said current labels do not reveal enough about industrial processing methods to support a meaningful definition. That tension could shape whether the agencies pursue a narrower label-based approach, a broader conceptual framework for future policymaking or a phased strategy that starts with food categories or nutrients of concern and evolves over time.
  • The timing of a proposed definition remains somewhat uncertain but appears to be approaching. The FDA’s 2026 priority deliverables do not specify a timeline, only saying that the “HFP will continue to collaborate with USDA and other federal government partners on gathering information, data and research, and analyzing comments to the 2025 RFI to develop a federal government definition of UPFs.” The language indicates that the timeline may extend beyond this year if the focus remains on interagency discussions, coordination and public comment review. However, recent public remarks from agency leadership suggest the process is accelerating. HHS Secretary ROBERT F. KENNEDY JR. said in late February 2026 that the administration plans to release a federal UPF definition in April, and FDA Commissioner MARTIN MAKARY provided an April or May timeline during remarks a week later at the National Food Policy Conference. These signals suggest that a proposal could emerge in the coming weeks, potentially reflecting the hybrid and operational approaches described across stakeholder comments.

To contact the authors of this item, please email Scarlett Salem ( ssalem@agencyiq.com) and Jared Rothstein ( jrothstein@agencyiq.com).
To contact the editor of this item, please email Jason Wermers ( jwermers@agencyiq.com).

Key documents and dates

by

FDA emphasizes study quality in shift to one-trial default for drug approval

By Amanda Conti and Ned Pagliarulo

FDA leadership has announced that one adequate and well-controlled trial, supported by confirmatory evidence, will now serve as the agency’s “default standard” for approval of novel drugs. The policy shift, articulated in a New England Journal of Medicine article, formalizes a trend the agency has increasingly adopted in practice.

The move places new emphasis on study quality, control selection, statistical methodology, and postmarket data collection — while raising questions about implementation, consistency across review divisions, and how sponsors currently in development should respond.

This AgencyIQ analysis examines the regulatory, evidentiary, and strategic implications of the new one-trial default and what it could mean for drug sponsors in the near and long term.

Fill out the form to read the full analysis.

Required *

by

FDA emphasizes study quality in shift to one-trial default for drug approval

The FDA will no longer regularly ask for two pivotal studies to support approval of novel drugs, Commissioner MARTIN MAKARY and top deputy VINAY PRASAD wrote in a medical journal article. They announced that one adequate and well-controlled trial, together with confirmatory evidence, will be the agency’s “default standard.” The new policy, which Makary previously signaled was forthcoming, is thin on some details, however.

By Amanda Conti and Ned Pagliarulo | Feb 19, 2026 6:47 PM EST

Catching up: Makary’s hints at policy change

  • In December 2025, FDA Commissioner MARTIN MAKARY told STAT that he planned to make a single pivotal trial the agency’s default option for drug approvals, shifting its longtime two-trial standard. “You can achieve the same statistical power with one trial as you would with two trials when it’s designed and controlled appropriately,” Makary said in his interview with the publication. [Read AgencyIQ’s initial analysis of the plan here.]
  • This change reportedly contributed to tensions between former Center for Drug Evaluation and Research Director RICHARD PAZDUR and Makary, The Washington Post reported on Nov. 21. “Pazdur and Makary have disagreed over the commissioner’s plan to reduce the number of research studies needed to justify more drug-related decisions, such as changes to drug labels,” two people granted anonymity told the Post. “Makary has argued that his plan to reduce the burden of proof to a single study would allow the agency to more quickly respond to new evidence. Pazdur has countered that multiple studies may be necessary in many cases and would better protect public health.” Subsequent reporting by the Post indicated that this clash, among others, contributed to Pazdur’s decision to retire weeks after taking the CDER director job.
  • Makary referenced the policy change again in a January 2026 interview on the All-In podcast. “I had one pharma executive tell me, ‘Hey, I love you’re going from two to one pivotal trials. That means we can run twice as many drugs through large pivotal trials,’” Makary said. “That’s what we want. We want to see more.” He later contextualized the policy as helping address the long development times required to bring a new drug to market, which he said “makes no sense” today.
  • The FDA has advanced other policies in parallel that could align with a one-trial standard. The agency in January released draft guidance on the use of Bayesian methodology in clinical research, which could help sponsors more easily design the kind of adaptive or seamless trials that may be favored under a one-trial standard. Makary has also repeatedly hinted that a major new proposal on these study types is in the works. [Read AgencyIQ’s analysis of the Bayesian guidance here and of Makary’s plans for seamless trials here.]

Context: The FDA’s evidence standards in principle and practice

  • When seeking approval from the FDA, companies must demonstrate “substantial evidence,” as defined by the Federal Food, Drug, and Cosmetic Act, that their product is safe and effective. Traditionally, the FDA has interpreted “well-controlled investigations” to mean at least two clinical trials for applications for new drugs or supplemental indications.
  • With the help of litigation and legislation, the legal definition has been updated so that “one adequate and well-controlled investigation” can be sufficient to demonstrate safety and effectiveness if supported by “confirmatory evidence.” The decision to accept two trials or a single trial with confirmatory evidence belongs to the FDA, though statute directs the agency to “implement a structured risk-benefit assessment framework.” The FDA has published and revised guidance offering its perspective on both foundational issues (last updated in December 2019) and the types of confirmatory evidence it is willing to consider (last updated in September 2023). The second guidance document states that, in general, the volume of confirmatory evidence needed is inversely proportional to the strength of the single adequate and well-controlled study. [Read full AgencyIQ analysis here.] Both documents currently remain in draft form.
  • Over time, the FDA has increasingly accepted a single pivotal trial and confirmatory evidence to support novel drug approvals. In a July 2024 analysis of original labeling documents, AgencyIQ found that the majority of new molecular entities approved by the agency since 2020 relied on a single pivotal trial. While the median number of pivotal trials held steady at two until 2020, the median has since dropped to one. AgencyIQ’s most recent analysis found that 27 of 46, or 59%, of novel products approved in 2025 listed just one pivotal clinical trial in their original labeling.

The FDA’s new bar for drug approval

  • The FDA will make one pivotal trial its “default standard” for approval of new medicines, Makary and Center for Biologics Evaluation and Research Director VINAY PRASAD declared in a “Sounding Board” piece published by The New England Journal of Medicine on Feb. 18, 2026. “Going forward, the FDA’s default position is that one adequate and well-controlled study, combined with confirmatory evidence, will serve as the basis of marketing authorization of novel products,” they wrote. While they acknowledged that the agency has often granted approvals on a similar basis in the past, they argued that “there remains confusion from manufacturers regarding settings in which a single trial will be accepted.”
  • Prasad and Makary justified this change by pointing to the many ways modern drug development can establish the “clinical credibility” of an experimental drug. They compared the current moment in clinical science to decades past, when two trials were necessary to reduce the risk of obtaining a lucky false positive with an ineffective or inert compound. “Yet in the modern world, as drug discovery becomes increasingly precise and scientific, the FDA considers not just effects on survival, but biochemical and intermediate changes that tell a complete biologic story: does this drug actually work?” Prasad and Makary wrote. “In this setting, overreliance on two trials no longer makes sense.” They outlined a list of other factors that the FDA will scrutinize to assess clinical credibility, including a trial’s control group, endpoints and statistical power, as well as the resulting data’s “concordance with biologic correlates” and “alignment” with intermediate endpoints.
  • Many of the factors they listed are already points of focus for FDA reviewers assessing submissions based on two trials. However, the way Prasad and Makary presented them in NEJM suggests that they could push the agency to apply higher standards on one trial. For instance, in noting how a trial’s control arm can contribute to a drug’s clinical credibility, they suggested that the FDA wants to see sponsors use the best available treatment as a comparator and concurrent controls instead of historical controls. Another factor they listed is “whether post-protocol therapy is on par with the U.S. standard of care,” citing a 2023 paper Prasad coauthored that criticized the use of “substandard” post-progression treatment in many cancer drug trials and called for stricter regulatory rules. Later in the NEJM piece, Prasad and Makary added that their proposal “may even enhance the FDA’s standards and reduce the risk of approval of products that may later need to be withdrawn, since greater attention will be placed on the one trial.” As an example, they noted that some past FDA approvals were based on trials comparing novel drugs to controls “no longer used by the American people,” again citing a prior paper coauthored by Prasad. The FDA’s focus moving forward, in other words, seems to be on trial quality rather than trial quantity.
  • Bayesian statistics can help capture many of these clinical factors, Prasad and Makary wrote. The FDA’s draft guidance, which is currently out for comment, details how sponsors should structure Bayesian trial analyses, with a particular note on controlling the Type 1, or false positive, rate. This is done differently under Bayesian methodology than the “frequentist” approach historically been most used in drug trials, requiring careful planning on the part of sponsors. In a world in which the FDA requires only one pivotal study, sponsors’ statistical assumptions may get a closer look by reviewers who lack a second study to provide reassurance on Type 1 error control.
  • The FDA may still ask sponsors for two trials. “If an intervention has a nebulous, pluripotent or nonspecific mechanism of action; if it affects a labile, short-term, or surrogate outcome; or if a trial has some underlying limitation or deficiency,” Prasad and Makary wrote, “additional adequate and well-controlled studies may be required.” While previously, sponsors were expected to justify why a single trial was sufficient, the agency’s leaders are now indicating that the FDA will tell them when it’s not. The outlined criteria are specific, but they’re also subject to interpretation. The FDA’s view of a trial limitation or the durability of a drug’s effect may be different than a sponsor’s, for instance. Also left undescribed is when the FDA might communicate whether additional studies are necessary.
  • “This reform is being rolled out synchronously with the agency’s postmarket initiative to collect robust data on all drugs and devices,” Prasad and Makary wrote. In a June 2025 JAMA Viewpoint article that defined their joint regulatory philosophy, the duo previously argued that the infrastructure to support regulatory use of large, heterogeneous data sources should be expanded to enhance postapproval monitoring and adverse event reporting. “Current systems both wrongly condemn and wrongly exonerate medical products,” Makary and Prasad wrote. [Read the full AgencyIQ breakdown of the piece here.] An October contract notice then previewed plans to enhance and expand the FDA’s existing postmarket surveillance system, the Sentinel Initiative, consolidating systems for drugs, biologics and devices into a centralized framework. However, the status of this project remains unclear, and the agency recently indicated in user fee negotiations that it is not looking to significantly expand Sentinel’s capabilities to include medical devices. [Read AgencyIQ’s January 2026 deep dive on Sentinel here.] Recent meeting minutes from negotiations for the medical device user fee program reauthorization do indicate a larger role for the device-specific surveillance system, known as the National Evaluation System for Health Technology, or NEST, but primarily as an evidence-generation mechanism rather than for surveillance. A September 2025 contract notice separately outlined an entirely new adverse event monitoring system – called “FDA Adverse Event Management (AEM)” in the notice – that would “centralize” and potentially replace the agency’s existing product-specific safety surveillance systems. However, the notice showed that the proposal was still in planning. Prasad and Makary’s reference in NEJM to a synchronous rollout, therefore, raises questions regarding how “robust data on all drugs and devices” will be collected and by whom.
  • Makary and Prasad defended their proposal against expected criticism. They denied that the FDA “is relaxing its standards and will now permit ineffective or even harmful products on the U.S. market,” arguing that “the FDA has never been perfect” and that multiple trials can still produce erroneous conclusions if there are deficiencies in trial design. The leaders added that increased attention on one trial could enhance standards, especially because FDA reviewers may have more time to assess an individual protocol.
  • The authors also addressed skepticism that the proposal’s impact could be limited given the agency’s consistent and increasing acceptance of these approaches in recent years – a point previously raised by AgencyIQ. “A wide body of literature suggests that default options anchor individuals and institutions psychologically, and we believe that formally articulating the FDA’s new position will spur biomedical innovation,” Makary and Prasad wrote, citing DANIEL KAHNEMAN’s popular 2011 psychology book, Thinking, Fast and Slow. This reference stands apart from other sources cited in the article, which mostly consist of research articles (the majority of which feature Prasad as an author).
  • Throughout the article, Makary and Prasad outlined their expectations for the announcement’s benefits. “Our move to change the FDA’s default position from two clinical trials to one will substantially reduce costs for sponsors and will speed drugs to market,” they wrote. Citing a March 2025 paper Prasad coauthored, they estimated that a single pivotal study could cost “30 million to 150 million dollars.” The 2025 paper references a 2020 paper reporting a study that found significant variation in costs associated with trials, noting, “From an individual clinical trial perspective, the largest single influence on estimated cost was the number of patients needed to establish the treatment effects.” Industry has historically cited much higher costs for drug development than Prasad’s research; recent studies have produced widely divergent estimates.

What’s next?

  • “The FDA expects a surge in drug development in response to our initiative,” Prasad and Makary concluded. How should sponsors respond? In the short term, the new policy creates additional regulatory uncertainty just as the agency is under fire for a perceived lack of predictability. For the roughly 55% to 70% of novel products already consistently approved based on one trial and confirmatory evidence, how will increased scrutiny of design aspects such as control and endpoint selection play out? For the roughly 30% to 45% of novel products generally approved based on multiple trials, what proportion will have the opportunity to leverage the new default setting? Will the new standard be applied consistently across review centers and divisions? This uncertainty is particularly pronounced for sponsors currently in the middle of development. Could the new policy lead to course changes or scrapped second trials?
  • Still, if successfully implemented, the one-trial default could meaningfully lower the cost and time of development for some sponsors. In notes published after the NEJM piece was released, equity analysts covering the industry highlighted drug developers that could benefit, including companies working in psychiatry, immunology and cardiology – fields that have historically seen fewer drugs approved based on a single trial. But the specifics matter. A more stringent singular trial may be more expensive or lengthy than two that are less rigorous. In CAR-T cell therapy, for example, Prasad recently outlined a new approach to approving new products that demands randomized evidence proving superiority on time-to-event endpoints. All seven currently licensed CAR-T therapies won approval based on single-arm studies that, for six of the seven, measured response rates for the primary endpoint.
  • The policy shift comes after Prasad reportedly overruled agency staff and refused review of an experimental influenza vaccine developed by Moderna. In a refuse-to-file letter published by the biotechnology company, Prasad cited Moderna’s choice of control in a large efficacy study that enrolled nearly 41,000 volunteers. The licensed, standard-dose flu shot Moderna used as a comparator “does not reflect the best-available standard of care in the United States at the time of study,” Prasad wrote. Citing prior FDA communication, Moderna countered that the regulator had previously signed off on its trial plans. [Read AgencyIQ’s full analysis of the back-and-forth here.] While the agency reversed course on Feb. 18 and agreed to review Moderna’s submission, the episode raised questions about whether sponsors can rely on FDA feedback remaining consistent. It also showed how study design issues can trip up sponsors even when their principal trial otherwise appears robust. Both points are likely to be top of mind for sponsors deciding how to act on the FDA’s new policy.
  • The NEJM piece is another example of FDA leadership issuing high-profile policy changes through podcasts and medical literature rather than traditional channels. In addition to gaps in practical information, this format has important implications for the staying power of the announced changes. To implement this change, the agency will likely need to issue revisions or finalize the two draft guidance documents pertaining to foundational issues and confirmatory evidence. The term “substantial evidence” also appears in dozens of guidance documents, indicating that adjusting the threshold for meeting this standard could have wide-ranging implications. Until such steps are taken, it’s fair to ask whether sponsors will be comfortable basing development decisions on a three-page paper written by embattled senior leaders.
  • The article does not offer any plans to formally evaluate the policy change. Novel product applications currently under review or poised to be submitted in the next year reflect conversations and planning under the previous trial paradigm and years of drug development. This means that it may take a few years for any new trends to surface – a timeline that could outlast the tenure of the authors.

Featuring additional research by Laura DiAngelo and previous research from Alexander Gaffney.

To contact the authors of this item, please email Amanda Conti ( aconti@agencyiq.com) or Ned Pagliarulo ( npagliarulo@agencyiq.com).
To contact the editor of this item, please email Jason Wermers ( jwermers@agencyiq.com).

Key documents and dates

by

FDA clarifies applicability of food traceability rule with new draft Q&A guidance

By Jared Rothstein

The FDA has issued draft Q&A guidance clarifying how the food traceability rule applies across the supply chain, including manufacturers, farms, retail food establishments, and restaurants. The 31-page document addresses implementation challenges ahead of the revised compliance timeline and responds to questions surrounding exemptions, transformation events, recordkeeping responsibilities, and traceability lot codes.

The guidance provides additional detail on issues such as pallet breaking, intracompany shipments, ad hoc restaurant purchases, continuous processing, and the distinction between fresh-cut and trimmed produce — all areas that have generated uncertainty as industry prepares for enforcement.

This AgencyIQ analysis examines key clarifications, enforcement signals, and open questions companies should evaluate as they refine compliance strategies.

Fill out the form to read the full analysis.

Required *