Blogs

In 2007, the National Academy of Sciences, Engineering, and Medicine created a report titled “Toxicity Testing in the 21st Century: A Vision and a Strategy.” The report advocates for developing in vitro, ex vivo, or in silico technologies to reduce animal testing while still providing adequate toxicity evaluation. Since then, there has been a push to develop new technologies to achieve this reality. New approach methodologies (NAMs) have become an exciting newer area of toxicology that has really taken off. Their popularity isn’t surprising—NAMs take less time to generate data and are often less expensive than traditional animal testing. While NAMs provide great promise, they need to be evaluated fully before they can be incorporated into any toxicology testing pipeline.  

Proper evaluation is especially important for developmental and reproductive toxicity (DART). DART is the study of how toxins affect reproduction, fertility, or developing offspring. As one can imagine, this is a sensitive area of toxicology research and one where NAMs will need careful consideration before they can replace other animal models.  

The 2025 Symposium Session “Qualifying NAMs for Developmental and Reproductive Toxicity: Advancements and Pitfalls” explored the types of questions that are being asked by regulatory agencies to approve NAMs for full-time use in DART pipelines. These approvals will help make NAMs successful in replacing other methods in DART assessments.  

The session began with the work of Jessica Palmer, a researcher in the Stemina Biomarker Discovery group. She spoke about the need to design roadmaps for DART for NAM development, as has been done successfully for developmental neurotoxicity (DNT). There are guidelines called the ICH S5(R3) on how to assess DART toxicity and potential suggestions for where in vivo methods can be applied, but there needs to be a roadmap for moving beyond in vivo methods. Ms. Palmer then introduced the US Food and Drug Administration’s (US FDA’s) biomarker qualification program and walked through the detailed process by which a substance or process gains approval. She used the example of Stemina's product devTOX quickpredict (devTOXqp), which is an assay that uses stem cells to predict toxicity based on metabolism. She outlined the studies and tests required to gain approval by the US FDA. She also cited some difficulties in the review process, including long wait times and lack of clear communication between agencies. Even with those hurdles, having clear guidelines and a roadmap for developing NAM technology is something that is moving in the right direction. Ms. Palmer’s talk plotted out the particular way that one might go about getting approval and was a clear example of how other researchers might go about doing the same thing for their own NAMs.  

Next, Ricardo Tieghi from the University of North Carolina Eshelman School of Pharmacy spoke about a database he has been working on called DeTox. This tool is an example of an in silico NAM since it uses previously gathered data from the US FDA, the Teratogen Information System (TERIS) database, and some other datasets to create a publicly available tool to help design future experiments. With DeTox, anyone can input the chemical structure of a chemical they want to study, and DeTox will then provide a probability of whether that chemical could cause developmental toxicity. The tool can also predict toxicity if there is exposure in the first, second, or third trimester of pregnancy. To do this, several quantitative structure-activity relationship (QSAR) models were built and validated to estimate the developmental toxicity of new compounds. While DeTox has tremendous potential, there are still some things to be worked out. For example, Mr. Tieghi showed that there are some cases where two chemical structures are very similar to each other, but that one is toxic and one is not. He called these cases “activity cliffs.” Since DeTox mainly works by toxicity predictions based on structures, it would not be effective in those cases. Additionally, right now the model cannot give mechanistic insight, but that’s something Mr. Tieghi hopes to implement in future versions of DeTox.  

The next talk was given by Michelle Kossack of Brown University. She spoke about her work with zebrafish as a model of female reproductive toxicity. While in vitro, ex vivo, and in silico models are helpful, there are some aspects that these systems don’t capture. Zebrafish are a NAM that straddle the space between since they are an in vivo system, but their larval stages are not considered invertebrates. The germline cells can be labelled and identified relatively early on in zebrafish, even 2–3 hours post fertilization. Dr. Kossack spoke about a few topics related to DART monitoring in zebrafish. She first spoke about endocrine disruptions in zebrafish. Zebrafish have an organ that will ultimately determine their final sex which differentiates around 20 days post-fertilization. If there is enough signal from oocytes, the fish will be female. If there is not enough signal, they will be male. As Dr. Kossack put it, there is a malleable relationship between sex and environmental exposure. This is an easy readout for whether there is endocrine disruption as a result of a toxic exposure. Next, Dr. Kossack demonstrated how modern experimentation can challenge previously held dogmas. There is a generally held tennant that older fish have poorer gametes. Dr. Kossack challenged this by doing studies that showed that the frequency with which one spawns the fish can have a big impact on fertilization rate. From there, she did literature searches to try and figure out what the conventions are in the field for pairing fish, as well as their spawning rates. She found that the zebrafish literature lacked consistent reporting and that some studies lacked enough information to make comparisons. Dr. Kossack ended her talk by stating that zebrafish are a powerful addition to the NAM systems for female fertility studies, but there is an imperative for zebrafish researchers to make sure they are reporting their methods properly to increase reproducibility in the field.  

The next speaker was Anna Kreutz from Inotiv. She presented her work on models she is working on for in vitro to in vivo extrapolation (IVIVE). IVIVE is a way of asking the question, “What do the in vitro models that we have potentially tell us about in vivo systems?” Dr. Kreutz presented her work with physiology-based kinetic (PBK) models. These are models that simulate chemical distribution in blood and fluids into the fetus and brain tissue. There are several of these models, so Dr. Kreutz compared three of them: GastroPlus, the R package ‘httk,’ and PK-Sim. Each model had their advantages and disadvantages, such as cost, ease of use, and how the models handle pregnancy. Dr. Kreutz used the models to derive human-administered equivalent doses (AEDs) of chemicals and compared those against human exposure predictions. She found that the predictions made by the PBK models were relatively consistent and supported the predictivity of the approach. Dr. Kreutz looked at any points of departure against the in vivo data that she had and found that they were within range of the AEDs, meaning that the AEDs from the in vitro modeling managed to predict what the risk assessment ranges would be for an in vivo study. Dr. Kreutz pointed out that this work was done for DNT, but there is great potential for expanding into DART as well.  

The last speaker was Maria Baltazar from Unilever. Her talk was about using frameworks to have NAMs replace traditional animal study replication for DART safety. She spoke about using a tiered next generation risk assessment (NGRA) framework. NGRA is a set of risk tests that incorporates in silico, in vitro, in vivo, and human experiments and creates a set of guidelines for risk assessment. Specifically, Dr. Baltazar spoke about the selection of NAMs as the first tier of the NGRA for DART assessments. Their group assessed 37 compounds that already have known outcomes and their benchmarks and used in vitro and in silico assays to determine if these chemicals would have a risk for DART. They found that by using their NGRA-tiered guidelines, they correctly identified 16/17 compounds that had a high DART risk. Dr. Baltazar concluded by saying repeating animal studies may not be necessary at all and that this approach can be used as a protective measure to flag compounds that may pose DART risks. Dr. Baltazar acknowledged that this approach does not determine potential mechanisms of DART, but that at this level, this tool is meant for protection, not pathology prediction.  

While the speakers are all working on different projects at different institutions, their talks share themes. They are working to make NAMs top of mind when toxicologists consider new chemicals or substances that may affect developmental or reproductive endpoints. They don’t want NAMs to be an interesting side project but rather the starting point of assessing toxins. Their work proves that NAMs can provide concrete evidence that could altogether replace the first rounds of animal testing. Many of these talks presented tools and technologies that can assist the whole community in using NAMs in DART research. Each of the presenters also mentioned some of the drawbacks of their current work, but they each expressed how they want to use these pitfalls to make their work stronger. With collaborations and current computer advancements, there’s no doubt that each of these researchers will find ways to reinforce their work and make NAMs a powerful force in DART research.  

This blog reports on the Symposium Session titled “Qualifying NAMs for Developmental and Reproductive Toxicity: Advancements and Pitfalls” that was held during the 2025 SOT Annual Meeting and ToxExpo. An on-demand recording of this session is available for meeting registrants on the SOT Online Planner and SOT Event App. 

This blog was prepared by an SOT Reporter and represents the views of the author. SOT Reporters are SOT members who volunteer to write about sessions and events in which they participate during the SOT Annual Meeting and ToxExpo. SOT does not propose or endorse any position by posting this article. If you are interested in participating in the SOT Reporter program in the future, please email SOT Headquarters.