New Drug Testing Methods to Consider

💡 Beyond Animal Models: Embracing New Approaches (Organoids, AI & More) for Safer Drug Development Encouragingly, the FDA is attempting to modernize drug development by reducing reliance on animal testing and prioritizing effective, human-relevant methods (https://lnkd.in/eTsDsfVH). The NIH made a similar announcement (https://lnkd.in/e24qvyaZ). What does this mean in practice? The FDA's roadmap (https://lnkd.in/eSAVsCDv), initially focused on monoclonal antibodies (mAbs), outlines concrete steps: 📄 Encouraging sponsors to submit New Approach Methodology (NAM) data (cell assays, organoids, computational models) for INDs, potentially reducing or replacing animal data. 🌍 Considering existing human safety data from comparable international approvals to reduce duplicative tests. 🔬 Piloting programs for primarily non-animal testing strategies under FDA consultation. ⏱️ Potentially shortening required animal studies where NAMs and early data show no concern. ✅ Developing guidance and robust validation pathways for NAMs (via retrospective/prospective studies, standardization, benchmarking, qualification programs, and ICCVAM collaboration). 🎯 Ambitiously aiming for animal testing to be the exception, not the norm, within 3-5 years. This shift aligns well with computational advances. Machine learning/AI offers powerful tools for predicting safety issues early. Our work has explored this: In Gayvert et al., 2016 (https://lnkd.in/ehV-iFvG, see attached cover), we used a data-driven 'moneyball' approach (analyzing often-overlooked features/stats for predictive insights, like sabermetrics in baseball) with PrOCTOR. It predicted clinical trial toxicity by integrating drug structural/target properties, significantly outperforming traditional methods like Lipinski's Rule of 5 or QED. Following that, in Madhukar et al., 2018 (https://lnkd.in/eVdxt9X8), we developed MAESTER, an ML approach predicting specific, tissue-level adverse events by combining structure, target, phenotypic, and genomic data with high accuracy. Kudos to these government agencies for attempting to change the status quo on safety assessment and over-use of animal testing. The possibility of integrating predictive computational methods with the innovative NAMs championed by the FDA and NIH represents a significant and exciting step forward. It will lead to accelerated development of safer, more effective therapies focused on human-relevant data.

What if one test could replace dozens—and make next-generation cancer drugs safer and more reliable at the same time? Antibody–drug conjugates (ADCs) are a powerful new class of cancer medicines that work like “smart missiles,” delivering highly toxic drugs directly to cancer cells while sparing healthy tissue. However, this precision comes at a cost: ADCs are among the most complex medicines ever developed, combining an antibody, a chemical linker, and a potent payload into a single product. Traditional laboratory tests typically measure only one feature at a time, making it difficult to fully assess safety, consistency, and quality. This review explains how Multi-Attribute Methods (MAM) use advanced mass spectrometry to examine multiple critical features of an antibody–drug conjugate (ADC) simultaneously, including drug loading, conjugation sites, structural stability, and the presence of unexpected changes or impurities. By replacing fragmented, single-attribute testing with a unified analytical approach, MAM has the potential to make cancer medicines safer, more reliable, and faster to develop. The review further discusses emerging regulatory and industry adoption of MAM, as well as how automation and artificial intelligence may enable its routine implementation, ultimately strengthening patient confidence in complex biologic therapies. Within the framework of ICH Q14 and ICH Q12, Multi-Attribute Methods facilitate continuous knowledge management, risk-based change evaluation, and sustained control of ADC quality from early development through post-approval lifecycle management. #MultiAttributeMethod #MassSpectrometry #ADC #Biologics #AnalyticalScience #ProteinCharacterization #LCMS #CQAs #PharmaAnalytics https://lnkd.in/eYmkPAG5

🚨 I'm excited to share our latest review, “New approach methodologies for drug discovery,” published in Cell by Cell Press and selected as a Featured Article. For decades, drug discovery has relied heavily on animal models. Yet, with persistently high clinical failure rates, a fundamental question remains: How predictive are animal models of human biology, and are there better alternatives? In this review, we highlight a paradigm shift toward human-centric New Approach Methodologies (NAMs), driven by rapid advances in both policy and technology. On the regulatory front, we discuss major transitions led by agencies such as the FDA (FDA Modernization Acts 1.0 → 2.0 → 3.0) and The National Institutes of Health (stem cell guidelines and the establishment of national organoid initiatives). 🔬 On the technology side, we frame NAMs evolution across three domains: ·        “New” - foundational 2D stem cell–based systems ·        “Newer” - advanced 3D organoid-based models ·        “Newest” - future-facing in silico and AI-driven platforms Across these domains, we highlight emerging therapeutic candidates, cutting-edge models, and translational and clinical applications. We also examine key biological, technical, and regulatory bottlenecks that need to be addressed to enable robust translational adoption, and discuss ongoing clinical efforts and societal considerations for responsible implementation. 🌍 Looking forward. If the past 30 years of drug discovery were shaped by animal models, the next 30 years, animal models will likely transition from a central to a supporting role, following the 3Rs principle: refinement, reduction, and ultimately replacement. Instead, the field will likely be defined by human-centric NAMs, powered by multiscale platforms, multi-omics data, and AI-enabled pipelines. This transformation is not only scientific, but also societal, aligning drug development more closely with human biology while reducing cost, inefficiency, and ethical burden. 👏 Congratulations to an outstanding team: Wenqiang (Eric) Liu, Paul Pang, Catherine Wu and Danilo Tagle from Stanford University School of Medicine, Stanford Cardiovascular Institute, Stanford Department of Medicine, Greenstone Biosciences, National Center for Advancing Translational Sciences (NCATS), The National Institutes of Health. 📄 Please check the full paper here: https://lnkd.in/estaR2cq #NAMs #DrugDiscovery #StemCell #Organoids #AI #PrecisionMedicine #TranslationalScience

More than 90% of drugs that were found to be effective on animals fail in human trials. The #FDA is now charting a new course. Within the next 3-5 years, animal testing will become the exception, not the rule. This shift is driven by the rise of New Approach Methodologies (NAMs) — tools designed to better predict human outcomes without relying on animal models. These include: • Organ-on-chip systems (in vitro, human-derived) • AI/ML models and computational simulations • High-throughput, human-based cell assays It’s a much-needed evolution; not just because animal testing is ethically fraught, but because animal testing often fails to reflect human biology.

Interesting development from the FDA, they’re planning to phase out animal testing requirements for monoclonal antibody drugs. A move that could reshape how early-stage drug safety and efficacy are evaluated. What’s exciting is that this shift not only reduces reliance on animal models, but also accelerates the adoption of next-generation, and often more predictive, technologies, from AI-driven toxicity screening to organ-on-chip platforms and human cell-based assays. Companies like Emulate, Inc. (organ-on-chip), Recursion and Insilico Medicine (AI-first drug discovery), and Benchling (molecular design infrastructure) have been building toward this moment for years. Now the regulatory environment is finally catching up. The FDA’s pilot program will be one to watch. If successful, it could accelerate timelines, improve predictability, and raise the ethical standard for how therapies are brought to market. link to article: https://lnkd.in/e-5WgbhP

The FDA just made a groundbreaking move, and what does this mean for the future of drug testing? Over the next 3 years, the FDA plans to shift preclinical toxicity testing away from animals and toward validated New Approach Methodologies (NAMs), including human cell models and in silico tools. 🔹 Accept international human data in lieu of new animal studies 🔹 Encourage organoid, MPS, and AI model data in IND/BLA submissions 🔹 Reduce or waive animal testing where NAMs are strong 🔹 Build a global toxicity database with the National Toxicology Program and the Tox21 initiative 🔹 Shorten or eliminate primate testing for mAbs 🔹 Track cost, time, and safety impact biannually 📍Goal: Make animal studies the exception, not the rule, by using a NAM-first framework. A future with fewer animals and better science is within reach. FDA reference published Apr 10, 2025: https://lnkd.in/eVGKjt_Z

𝗡𝗲𝘄 𝗙𝗗𝗔 𝗱𝗿𝗮𝗳𝘁 𝗴𝘂𝗶𝗱𝗮𝗻𝗰𝗲 𝗼𝗻 𝗡𝗔𝗠𝘀 On March 18th, FDA released draft guidance on New Approach Methodologies (NAMs) in drug development. This does 3 important things:  • gives companies a clearer framework for using NAMs  • shifts the discussion from “animal vs non-animal” to “fit-for-purpose”  • signals that human-relevant models can support regulatory decisions 𝗪𝗵𝗮𝘁 𝗰𝗵𝗮𝗻𝗴𝗲𝗱:  • NAMs are framed as decision tools, not replacements  • acceptance is tied to context of use, not full validation  • weight-of-evidence approaches are explicitly supported 𝗜𝗻 𝗽𝗿𝗮𝗰𝘁𝗶𝗰𝗲:  • companies can integrate NAMs earlier  • data packages can combine animal and non-animal evidence  • organoids, organ-on-chip and in silico models gain traction ✅ 𝗙𝗗𝗔 𝗱𝗲𝗳𝗶𝗻𝗲𝘀 𝟰 𝗰𝗼𝗿𝗲 𝗽𝗶𝗹𝗹𝗮𝗿𝘀 𝗳𝗼𝗿 𝗡𝗔𝗠 𝘂𝘀𝗲 1️⃣ 𝗖𝗼𝗻𝘁𝗲𝘅𝘁 𝗼𝗳 𝘂𝘀𝗲: clearly define the regulatory question and decision the NAM is intended to support, ensuring it addresses a specific development need or data gap 2️⃣ 𝗛𝘂𝗺𝗮𝗻 𝗯𝗶𝗼𝗹𝗼𝗴𝗶𝗰𝗮𝗹 𝗿𝗲𝗹𝗲𝘃𝗮𝗻𝗰𝗲: demonstrate that the model captures key human biology, including relevant cell types, architecture, and mechanisms of toxicity or drug response 3️⃣ 𝗧𝗲𝗰𝗵𝗻𝗶𝗰𝗮𝗹 𝗰𝗵𝗮𝗿𝗮𝗰𝘁𝗲𝗿𝗶𝘇𝗮𝘁𝗶𝗼𝗻: provide evidence of reliability and robustness, including system properties, reproducibility, and absence of artefacts (e.g. compound absorption or platform interference) 4️⃣ 𝗙𝗶𝘁-𝗳𝗼𝗿-𝗽𝘂𝗿𝗽𝗼𝘀𝗲: show that the method is sufficiently validated to support the intended regulatory decision, even within a weight-of-evidence framework 𝗪𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀:  • less regulatory ambiguity and clearer standards for credible NAM evidence  • improved human relevance, reducing preclinical uncertainty and late-stage risk  • competitive advantage for teams integrating NAMs early into development strategy This is not just about replacing animals. It is about improving prediction and decision-making for humans. This is a directional shift. #FDA #NAMs #Preclinical #Toxicology #Organoids #OrganOnChip

The Warranty Is Now Required: FDA's NAM Guidance Changes Everything On March 18, 2026, FDA stopped asking whether New Approach Methodologies belong in drug development. They published the blueprint for how they must be used. The draft guidance, titled 'General Considerations for the Use of New Approach Methodologies in Drug Development,' is not regulatory theater. It is a decision architecture that will determine which sponsors accelerate and which ones stall. Acting CDER Director Tracy Beth Hoeg stated directly: 'It is time for the FDA to shift the drug development paradigm away from the current default of using animals.' Here is what matters: The guidance organizes validation around four pillars: context of use, human biological relevance, technical characterization, and fit-for-purpose assessment. This is not a checklist. It is a framework for building regulatory confidence. But here is the operational insight most sponsors will miss: full validation is not always required. A fit-for-purpose NAM, even if not formally validated, may adequately address specific toxicological concerns when integrated into a weight-of-evidence approach. Next Steps 1.    Submit comments by May 18, 2026 via regulations.gov (Docket FDA-2025-D-6131) https://lnkd.in/e5Z_X75P 2.    Audit your NAM portfolio against the four pillars 3.    Engage with your review division early The warranty is not optional anymore. The question is whether you are building systems that can deliver it.

🧰 Artificial Intelligence and Bayesian Randomization : A new signal from FDA guidance 💠 When AI-powered analytics reshape randomization in a clinical trial, the study itself can begin to learn while it is still running. In January 2026, the U.S. Food and Drug Administration released a new draft guidance that signals an important shift in how clinical trials may be designed: 🧰 Use of Bayesian Methodology in Clinical Trials of Drugs and Biological Products The document explains how Bayesian statistical methods can be used in drug and biologic development to support adaptive trial designs. 🔷 How Bayesian adaptive randomization works In traditional clinical trials 💊, the randomization ratio is fixed from the start. For example: 🔹 50% of participants receive the investigational therapy 🔹 50% receive the control treatment. That ratio usually remains unchanged until the study ends. Bayesian adaptive trials allow something different. 💠 As patient outcomes accumulate, statistical models update the probability that a treatment is effective. If early results suggest that one therapy 💊 is performing better, the trial can adapt its randomization. This means that: 🔹 more participants may be assigned to the treatment showing stronger responses 🔹 fewer participants are assigned to less promising therapies. Instead of continuing in a fixed direction, the trial 💊 can use emerging data to move toward the most promising treatment signals. 🔷 Why this matters today Bayesian statistics is not new. What has changed is the ability to apply these methods at scale using modern computational power, advanced analytics, and increasingly AI-supported 💠 data processing. These tools allow clinical trials to: 🔷 identify effective treatments faster 🔷 reduce exposure to ineffective therapies 🔷 accelerate evidence generation in complex studies. 🔷 A well-known example One of the best-known clinical research protocols applying adaptive statistical approaches is the breast cancer trial: 💠 I-SPY 2 Trial ClinicalTrials.gov: NCT01042379 This ongoing clinical trial evaluates investigational therapies for high-risk breast cancer, using adaptive statistical approaches within a continuously evolving trial structure. 💠 The new FDA guidance signals that Bayesian methodology — increasingly supported by modern computational tools — is moving into the mainstream regulatory toolkit for clinical trials. For clinical research teams, this means that trials are gradually becoming dynamic learning systems, rather than static experiments. 📚 FDA Draft Guidance Use of Bayesian Methodology in Clinical Trials of Drugs and Biological Products January 2026 #ArtificialIntelligence #ClinicalTrials #BayesianStatistics #RegulatoryScience #FDA #DrugDevelopment #ClinicalResearch #GCP #GxP #FloorSoldiersStrategy 🌸🌿

Following up on yesterday’s post about the FDA’s game-changing announcement around New Approach Methodologies (NAMs) for drug safety testing — the agency also released an 11-page document outlining where they’re heading. One section really stood out to me: 3D human-derived in vitro systems — including organoids and microphysiological systems (MPS), or “organs-on-chips.” Here’s the essence: ➡️ Organoids: Miniature, self-organizing structures (e.g., liver or gut) that mimic real tissue architecture and function. ➡️ Organ-on-a-chip: Takes it a step further with microfluidics, mechanical cues, and co-cultures — all built on a chip to simulate a living, breathing human organ (or several organs). ➡️ Applications? Everything from identifying DILI and pro-arrhythmic risks to immune-mediated cytokine storms (yes, even the infamous TGN1412-type responses). These systems don’t just replace animal models — in many cases, they outperform them in predicting human biology. That’s not science fiction anymore. At InSphero , we’ve been in this space for 16 years — long before it was cool or regulatory guidance was even on the horizon. Our primary cell-derived spheroids may not have pumps or perfusion built-in, but they’ve proven to be scalable, automation-friendly, and reproducible — key traits if you want to actually do this at scale in pharma pipelines. It’s exciting to see the FDA explicitly acknowledge these tools in their roadmap. It means the conversation is shifting — from “can we replace animal testing?” to “how fast can we scale what works better?” Let’s keep the momentum going. Here's the link to the FDA doc if anyone’s interested - https://lnkd.in/emiRHBVz #NAMs #3DModels #MPS #Organoids #OrganOnChip #DrugSafety #InVitroToxicology #InSphero #LiverChip #TranslationalScience #MonoclonalAntibodies #FDA #HumanRelevant #NoAnimalTesting