Food and Drug Administration issued a proposed rule (6) in February 2019 to update regulatory requirements for 16 sunscreen ingredients including recommendations to assess their human systemic absorption. Historical assumptions posited sunscreen active ingredients were not absorbed and that the clinical studies to test this were not feasible. In collaboration with FDA’s Office of New Drugs, DARS led and conducted clinical research demonstrating that maximum usage sunscreen clinical trials are feasible, that sunscreen active ingredients are absorbed at levels that would trigger additional safety studies (Figure 1), and that further research is needed to fill in data gaps for sunscreen ingredients. Results from these studies were published in the Journal of the American Medical Association: JAMA (7, 8), one of which was the most viewed JAMA article in 2019 (8). DARS also published the detailed bioanalytical method it developed (9) to enable further study and collaborated with FDA’s Office of Pharmaceutical Quality staff on a study to assess sunscreen ingredient absorption with an in vitro model (10).

In 2019, FDA received a citizen petition (11) indicating that unacceptably high levels of N-nitrosodimethylamine (NDMA), a probable human carcinogen, were detected in specific ranitidine products. Ranitidine products were removed from the US market in April 2020, owing to the unacceptable amounts of NDMA in ranitidine products. Additionally, the petitioner postulated that ranitidine could convert to NDMA in people. In response, DARS, working with FDA’s Office of New Drugs and Office of Pharmaceutical Quality, designed a rigorous clinical study that carefully controlled for factors known to influence NDMA exposure and measurements including diet (12). DARS developed and validated a sensitive analytical method that was used to measure NDMA in study participants’ urine and blood. The study found no evidence of increased NDMA content in participants’ urine or blood (Figure 2) who were administered ranitidine (12). Additional in vitro investigations with simulated gastric fluid supported that ranitidine would not produce NDMA under physiologic conditions (13). With this information, FDA may consider allowing ranitidine products back on the market if they are manufactured to be stable with low, acceptable amounts of NDMA that do not increase over time.

In October 2020, FDA issued an Emergency Use Authorization for remdesivir to treat COVID-19. Subsequent adverse event reports received by FDA suggested a possible connection to acute kidney injury or hepatic injury. FDA consulted DARS to assess whether there was additional evidence or a potential mechanism for these adverse events. DARS used target prediction software, secondary pharmacology data analysis, quantitative structure-activity models, and structural similarity analysis to review remdesivir and its metabolites. DARS found that remdesivir and its metabolites were structurally similar to drugs associated with renal and hepatic toxicity and a description of the potential risks associated with these adverse events is now found in remdesivir’s product labeling (14).

Division of Applied Regulatory Science developed and published a mechanistic COVID-19 disease model that can be used to guide candidate drug selection and dosing strategies based on non-clinical data. The model strategy was developed using remdesivir as a proof-of-concept example and provides a well-calibrated and validated model for performing similar non-clinical to clinical translations for other potential COVID-19 therapies (15).

Division of Applied Regulatory Science is conducting in vitro, computational, and clinical research to optimize the use of existing opioid antagonists (naloxone) and to advance drug development tools for new opioid antagonists (Figure 3). Synthetic opioids, such as illicitly manufactured fentanyl, cause a large number of opioid overdose deaths in the U.S. (19). Moreover, a rise in illicit fentanyl derivatives (that vary widely in potency) make it difficult to identify an adequate naloxone dose. DARS, in collaboration with the University of Maryland (20, 21), applied an advanced in silico molecular dynamics method called metadynamics to elucidate the dissociation mechanism of fentanyl and its derivatives to calculate the residence times at the mu-opioid receptor. The simulations uncovered two distinct dissociation mechanisms, one of which involves a newly identified binding pocket that contributes to the long residence time and high binding affinity of fentanyl and its derivatives. This new method was used to predict the relative dissociation time of a newly identified opioid that emerged from illegal markets, helping to inform overdose prevention strategy (22).

Division of Applied Regulatory Science developed a quantitative systems pharmacology computational model (23) to assess naloxone dosing strategies and applied it to the regulatory review of a new naloxone formulation for use by military personnel and chemical incident responders (24, 25). The model connects the pharmacokinetics of different opioids and naloxone to in vitro receptor binding kinetics, and subsequently to effects on respiratory depression and cardiac arrest. DARS used this model to assess the impact of naloxone dosing administered as an emergency treatment or temporary prophylaxis following exposure to fentanyl or carfentanil using endpoints of respiratory depression and cardiac arrest (24, 25). With this information, FDA’s review team was better able to understand this naloxone formulation’s capacity to reverse opioid effects under different scenarios.

Division of Applied Regulatory Science conducted a clinical study to characterize intranasal naloxone exposure after different repeat dosing strategies and predict the impact of the different dosing strategies on rescuing patients from fentanyl and fentanyl-derivative overdoses using the computational model described above (26). In addition, through a collaboration with Leiden University, an additional study is assessing the dynamics of how intranasal naloxone reverse respiratory depression from fentanyl in healthy opioid naïve participants and chronic opioid users (27). Together, these studies will inform intranasal naloxone dosing strategies in the community setting.

In 2016, FDA issued a warning about the increased risk of respiratory depression when combining opioids with benzodiazepines (28). Following the warning, DARS received a consult request concerning the potential for drug-drug interactions between opioids and drugs that might be co-prescribed in lieu of benzodiazepines. DARS conducted non-clinical in vivo studies to assess effects on respiratory depression for drugs given alone or in combination with an opioid (29). As a result of these studies, DARS designed a clinical study to assess the translatability of the findings to humans (27). This DARS-led clinical study assessed whether combining the selective serotonin reuptake inhibitor paroxetine or the atypical antipsychotic quetiapine with the opioid oxycodone, compared to oxycodone alone, decreased ventilation during hypercapnia (elevated carbon dioxide). The study found that paroxetine combined with oxycodone decreased ventilation (Figure 4), indicating the need for further study of the clinical implications (30).

Division of Applied Regulatory Science developed a Public Health Assessment via Structural Evaluation (PHASE) methodology that uses molecular structure to predict the biological function (31) of newly identified opioids (Figure 5). Results from PHASE have the potential to inform public health and law enforcement agencies with vital information regarding newly emerging illicit opioids in the absence of pharmacological data. In an example of using this method, DARS assessed kratom alkaloids (32), which was highlighted by then FDA Commissioner Dr. Scott Gottlieb in a statement about kratom’s potential for abuse, addiction, and serious health consequences (33).

While significant progress has been made in engineering biological products, the human immune system may still produce an unexpected or exaggerated immune response to a drug product resulting in poor efficacy or life-threatening adverse reactions. DARS is comparing the ability of different non-clinical models to predict immune-mediated adverse events from biological drug products. DARS tested the use of novel non-clinical models to predict cytokine release syndrome, a potentially life-threatening complication associated with biological products (36, 37) and showed that non-clinical models can effectively demonstrate this adverse event. Additionally, after successfully demonstrating immune-mediated activation in a non-clinical model (38), investigations are continuing for checkpoint inhibitor oncology treatments for which adverse events cannot presently be predicted using computational, in vitro, or conventional non-clinical methods.

Knowledge of a drug’s molecular targets can provide early identification of a drug’s effects and potential safety concerns. DARS leads multiple efforts collating information about a drug’s known and predicted targets to identify potential safety concerns. For example, DARS developed multiple computational methodologies, including with machine learning, to predict a drug’s adverse effects based on the biological receptors that the drug, or similarly structured drugs, are known to target (39–41). These computational methodologies demonstrated promising performance in predicting significant adverse events. These results may indicate which organ systems and adverse event categories to closely monitor during clinical trials or during clinical and non-clinical data review.

Additionally, DARS is analyzing and building a database (Figure 6) for secondary pharmacology activity submitted by industry as part of their Investigational New Drug application (42, 43). A drug developer typically conducts in vitro target binding and functional assays for 80–100 biological receptors to determine potential on-target and off-target effects. However, the targets chosen for the assays as well as the submission format are not currently standardized across the industry. Data from these assays have been manually extracted and curated into a database that will allow easier access to and analysis of these study results. Additionally, DARS is engaging in a public-private partnership with the Pistoia Alliance to determine the best methods for future regulatory submission of these studies.

The use of (quantitative) structure-activity relationship, or (Q)SAR, models has become an integral part of regulatory review as it can rapidly assess a compound’s toxicological and pharmacological properties based solely on chemical structure (5, 44). (Q)SAR models describe the association between chemical structural features and biological activity under the general assumption that similar chemical structures exhibit similar biological activities (45). (Q)SAR models and databases developed through DARS are used to support drug safety assessments and inform regulatory decisions at FDA. Models for several endpoints that comprise the standard genetic toxicity battery described in the International Council for Harmonisation (ICH) S2(R1) regulatory guidance (46) were recently updated and best practices on their use were also described (47–49). Additionally, a structure-activity relationship (SAR) profiler was constructed through an external collaboration to evaluate potential interactions caused by metabolites. Lastly, DARS recently developed a new model to predict a drug’s potential to cross the blood-brain barrier (50) and is in the process of developing a model to predict drug-induced cardiotoxicity using post-market safety data.

As patients often use more than one drug at a time, it is critical to know if drugs taken together interact leading to safety or efficacy implications. To collect this information, FDA requires drug-drug interaction studies from pharmaceutical sponsors (51). DARS evaluates in vitro and in vivo methods to identify best practices when conducting transporter- or metabolism-based drug-drug interaction studies. For example, conflicting information appears in the scientific literature regarding the extent of specific cytochrome P450 (CYP) involvement in metabolizing methadone (a treatment for opiate dependence). Drawing from previously submitted new drug applications, DARS constructed a database of drug-drug interaction studies between methadone and antiviral medications known to affect these CYP enzymes (52, 53). After analyzing these 29 studies, DARS formulated recommendations on how best to conduct future drug-drug interaction studies with methadone.

Many of DARS’ current initiatives (Figure 7) will inform the Agency’s thinking on the use of pharmacodynamic (PD) biomarkers to demonstrate biosimilarity that may streamline or negate the need for comparative clinical studies (56, 57). This included conducting three clinical studies (58–60) to define best practices on characterizing PD biomarkers for different classes of drugs and to develop general considerations applicable to all types of biomarkers for biological products. These studies included assessments to evaluate uses of proteomic and transcriptomic analysis of human plasma to identify novel biomarkers for biosimilar development (61). A joint FDA/Duke Margolis Workshop (62) discussed initial findings and facilitated a broader discussion on use of PD biomarkers for biosimilar development. Additional details will be reported in the January 2023 themed issue on Innovations in Biosimilars in the journal Clinical Pharmacology and Therapeutics (61, 63).

A factor influencing rapid development of biosimilars and generics is the ability to predict the potential risk of a stronger immune response (immunogenicity) in humans to a proposed biosimilar or generic drug than that seen with the innovator product. For most biosimilars this is still assessed through clinical trials. DARS is studying the ability of non-clinical approaches to predict immunogenicity risk without conducting a clinical trial. This includes assessing specific in vitro assays and cell types, in vivo models, and identification of useful controls. These methods have the potential to identify products with immunogenicity risk sooner and streamline the development of biosimilars and certain complex generic drugs.

Performing clinical studies in pediatric populations that involve frequent blood samples can be challenging if conventional methods are used that require relatively large blood volumes that can limit generic drug development for children. An alternative method is determining the concentration of drug in blood from dried blood spots (DBS) that only require 10 μl of blood compared to ∼5 mL (500 times more) with conventional methods. However, there has been a lack of analytical methodology metrics for using DBS analysis for generic drug bioequivalence studies. DARS developed and validated novel, sensitive, and specific analytical methods (Figure 8) to support pediatric pharmacokinetic studies using DBS cards (64).

In collaboration with CDER’s Office of Generic Drugs, DARS has conducted multiple studies to evaluate the potential of new approaches to advance the development of complex generic drugs such as injectable delayed release formulations and device-drug combinations. DARS conducted a non-clinical in vivo study to evaluate an in silico modeling approach for delayed-release risperidone injectable products that might be used to inform clinical trials for generic drug approvals. Another recent study investigated slow-release dexamethasone intravitreal implants used to treat inflammatory conditions in a wide variety of ocular diseases. Unfortunately, no generic implants are available on the market largely due to the challenges associated with measuring drug concentration within the eye to establish bioequivalence. DARS conducted a non-clinical in vivo study to measure and correlate intraocular and systemic dexamethasone concentrations after intravitreal implantation (65). This research will inform whether systemic dexamethasone exposure might serve as a surrogate for intra-ocular exposure in assessing the bioequivalence of generic dexamethasone implants. A similar approach may be possible for development of other generic intraocular products.

Division of Applied Regulatory Science led regulatory reviews of in vitro data to expand the approval of rare disease treatments. The scientific framework enables the evaluation of drug approval proposals in disease population subsets with very rare genetic variants who may not be well represented in clinical trials (70, 71). In 2017, DARS was consulted to determine if electrophysiology and other in vitro data were adequate to support the expanded approval of ivacaftor to treat patients with cystic fibrosis (CF) who have such rare variants that the patient population is too small for an adequately powered clinical trial. An in vitro cell-based approach was used to assess the response of the mutated or dysfunctional proteins in the presence of drug to make inferences about the potential for response in patients. As a result, the 2017 expanded approval of ivacaftor has allowed ∼1,500 new patients access to the drug based on in vitro data that predicted the clinical responsiveness of patients not included in clinical trials (72). Soon after, DARS evaluated in vitro data to support the expansion of approval for the combination drug ivacaftor/tezacaftor/elexacaftor (73). In 2018 DARS was also consulted to determine if in vitro functional data was adequate to support the approval of migalastat to treat patients with Fabry disease. An estimation about the extent of patient access to migalastat based on the original approval is not yet available but a similar increase in access to the drug for patients with 348 amenable galactosidase alpha gene (GLA) variants is expected (71). In 2021, in vitro data was also evaluated by DARS to support addition of two new amenable GLA variants to the label.

Induced pluripotent stem cells (iPSCs) may serve as a renewable source of differentiated cell types with patient- or population-specific properties (74–76). This model could replace, reduce, or help refine clinical trials for rare diseases. A collaboration with Stanford University is evaluating and comparing in vitro models with patient-specific iPSCs from patients with Duchenne Muscular Dystrophy and iPSCs genetically engineered to have the same genetic variants using CRISPR. This project seeks to inform the development of general standards, quality control criteria, and best practices for iPSC-based models to assess the efficacy for rare disease treatments.

Pediatric cancers are rare diseases with limited treatment options (77). To facilitate drug development in accordance with the Research to Accelerate Cures and Equity (RACE) Act, FDA developed the Pediatric Molecular Target List to provide guidance to pharmaceutical developers when planning new drug and biologic submissions that may be relevant to pediatric cancer treatment. To assist with maintaining the Pediatric Molecular Target List, DARS is developing natural language processing algorithms to identify evidence in peer-reviewed literature and external databases for molecular targets associated with pediatric cancer. These algorithms will inform regular updates to the Pediatric Molecular Target List including identifying new references for existing targets, emerging evidence for new targets, as well as supporting the review of initial pediatric study plan submissions.

As a part of FDA’s mission to ensure the safety and efficacy of human drugs, FDA reviews drug developer submitted data to establish under what conditions a new drug can be safely administered to patients and whether the new drug carries an increased risk of various adverse effects. This involves assessing end points that cannot be ethically obtained in humans, such as histopathological analysis of all major organs. Animal studies have played a critical role to meet this need and bring safe and effective therapies to patients. At the same time, FDA has a long-standing commitment to replace, reduce, and refine animal testing (the “3Rs”) with successes to date in harmonizing regulatory guidelines with international regulators and accepting alternatives to animal testing in certain areas (78, 79). Recent advances in systems biology, iPSCs, engineered tissues and mathematical modeling present new opportunities to improve our ability to predict risk and efficacy (Figure 10). However, multiple steps are required to translate these new technologies into regulatory use and maintain the same standard of safety, efficacy, and quality of FDA-regulated products. While we are nowhere near being able to replace all animal testing, there are opportunities for new alternative methods to make additional inroads in addressing the 3Rs for specific contexts of use. FDA has proposed funding for an Agency-wide New Alternative Methods Program, which was presented to the FDA Science Board in June 2022 by DARS (80).

In line with the Agency’s broader goals, DARS is performing applied research on complex in vitro models, including with iPSCs and microphysiological systems, to fill information gaps about the potential utility of these assays for regulatory contexts of use and to assess reproducibility, quality control and performance criteria. This has included studies on liver microphysiological systems (81, 82) for assaying drug toxicity, metabolism and accumulation, and additional studies with iPSC-derived liver and cardiac cells (74–76). Furthermore, DARS is assessing drug permeability and metabolism using the lung, gut, and a gut-liver interconnected MPS. Recently, DARS contributed to two white papers, which focused on in vitro methods for assessing drug-induced effects on cardiac contractility (83, 84). Additionally, DARS is studying the reproducibility of three-dimensional engineered heart tissue models. This work will be used to inform policy and guidance around qualifying alternative methods for regulatory use.

Division of Applied Regulatory Science has pursued a multi-year collaborative effort to improve the assessment of drug-induced cardiac toxicity from abnormal heart rhythms. Multiple drugs were removed from the market in the 1990s to 2000s and the ICH of Technical Requirements for Pharmaceuticals for Human Use Guidelines were implemented in 2005 to require specific non-clinical in vitro and in vivo studies, as well as dedicated clinical trials to assess the risk. However, these methods lacked specificity, resulting in drugs being dropped from development, sometimes unnecessarily. In response, the Comprehensive in vitro Proarrhythmia Assay (CiPA) initiative was implemented by FDA in collaboration with public-private partnerships including industry, academics, and other global regulators (Figure 11) (85–87). Through this, DARS led collaborative studies to assess:

The initiative prompted the ICH to open an Implementation Working Group with the purpose of developing a new guideline through a series of questions and answers to the existing ICH cardiac safety guidelines for an integrated strategy on using non-clinical data to inform clinical decision making. DARS staff served as lead (rapporteur) in developing new Questions and Answers (Q and As) to the ICH E14/S7B “Clinical and Non-clinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential” guideline (108). This new ICH guideline contains best practice recommendations for in vitro ion channel and human induced pluripotent stem cell assays to enable use as follow-up studies in place of potential animal studies and principles for validating in vitro and in silico proarrhythmia models and qualifying them for regulatory use, which can reduce animal use.

The ICH M7(R1) international regulatory guideline recommends the use of (Q)SAR predictions as a state-of-the-art alternative to experimental testing to qualify a drug impurity for mutagenic potential (112). DARS receives consults to review mutagenicity (Q)SAR predictions in pharmaceutical industry submissions and to generate in-house (Q)SAR predictions using a battery of complementary models and expert knowledge.

Extractable/leachable compounds are industrial chemicals that may be present in a drug product from manufacturing, packaging, drug delivery, and/or container closure systems. For non-mutagenic compounds with limited general toxicity data, a manual structure-activity relationship, or “read-across,” assessment based on structurally similar surrogate compounds may be used to inform setting a permissible daily exposure limit. DARS staff review applicants’ proposed surrogates for extractable/leachable compounds and, if needed, recommend alternative surrogates based on structural, metabolic, and physicochemical considerations.

Nitrosamine impurities constitute a special class of potentially mutagenic impurities that are controlled to very low levels because of their high carcinogenic potency. Due to the limited availability of experimental data for these impurities, read-across and (Q)SAR models may be used to inform an acceptable intake based on structurally related surrogate compounds. DARS staff conduct read-across and (Q)SAR analyses and review structure-based justifications for proposed nitrosamine limits to support internal decision-making.

Within drug labeling, the overdosage section describes the signs, symptoms, and laboratory abnormalities that occur with drug overdose as well as recommendations on treatment, if known (113). Although FDA has implemented initiatives to update this section, it is typically not revised after initial approval (114). Using natural language processing informatics tools, DARS is identifying outdated or inaccurate information within the overdosage section. Thus far, the project has evaluated drugs from 10 different drug classes highly associated with drug overdose fatalities to inform potential label updates.

Antimicrobial resistance (AMR) remains a significant global public health threat. FDA, in collaboration with internal stakeholders, develops approaches to detect, prevent, and limit the impact of AMR (115). In particular, DARS developed a cell-based method to measure the rate at which antibiotic resistance appears. This hollow fiber bioreactor system (116) administers drugs to cells using human pharmacokinetic profiles to identify drug combinations that slow the development of resistance. Next-generation sequencing is being used to investigate the genetic and epigenetic biomarkers of antibiotic resistance and to investigate the effects of combinations of oral antibiotics on bacterial genomes and the gut microbiome (117) high-through put screening assay to identify optimal drug combinations for different organisms.

Division of Applied Regulatory Science generates data and evidence to help the Agency make the most well-informed regulatory decisions and take the most appropriate regulatory actions. Examples include issuing Drug Safety Communications, drug approvals, and as in this case, updating drug labeling. In 2018, information emerged showing a disproportionate number of neural tube defects in infants born in an HIV patient population treated with dolutegravir as part of a clinical trial (118). Over time, DARS received four consults on this issue. DARS evaluated dolutegravir in comparison with four related drugs using structural similarity, secondary pharmacology, and computational toxicology. The work informed updates to dolutegravir’s label and the labels of related drugs.