Respiratory sensitization remains one of the most difficult endpoints in inhalation toxicology. For industry, the real challenge is not only identifying whether an inhaled substance is hazardous, but distinguishing a true respiratory sensitizer from a simple irritant. That distinction matters: irritants may trigger transient inflammation, while sensitizers can induce an immune-mediated response that may lead to long-term occupational respiratory disease, including asthma. As regulatory and toxicology teams face increasing pressure to generate human-relevant, animal-free data, new approach methodologies are becoming central to respiratory hazard identification.
A recent paper published in Frontiers in Toxicology addresses this exact issue. The article reports on a workshop organized by the Johns Hopkins Center for Alternatives to Animal Testing and the Evidence-Based Toxicology Collaboration, which brought together scientists from academia, industry, and regulatory bodies to examine the state of the science for respiratory sensitization assessment. The paper reviews the biological complexity of the endpoint, the current lack of validated and accepted non-animal methods, and the need for more human-relevant testing strategies capable of supporting decision-making in product development and safety assessment. The discussion is highly relevant for companies working with chemicals, formulations, consumer products, pharmaceuticals, and advanced materials where occupational respiratory toxicity is a concern.
One of the paper’s central messages is that respiratory sensitization cannot be solved with simplistic testing systems. Unlike irritation, sensitization involves a more complex chain of biological events, including epithelial barrier interaction, innate immune activation, and downstream adaptive immune mechanisms. This is why the authors emphasize the importance of advanced in vitro systems that better reflect lung physiology. In particular, the paper highlights the value of air-liquid interface lung models and other alveolar epithelial models capable of recreating realistic inhalation exposure conditions. These systems offer a more human-relevant lung model than conventional submerged cell assays and are better positioned to capture the mechanistic differences between a respiratory sensitizer and an irritant.
A key result from the paper is the clear consensus that there is still no single accepted standalone method for respiratory sensitization, but there is growing momentum around integrated NAM-based strategies. The workshop participants point toward a future built on combinations of mechanistic assays, exposure-relevant test systems, and defined approaches that together improve confidence in respiratory hazard identification. This is important because it frames the problem correctly: the industry does not need another oversimplified assay with weak relevance, but rather robust testing frameworks that combine biology, exposure realism, and mechanistic interpretation. In that context, human-relevant lung models are not peripheral tools; they are becoming foundational components of next-generation inhalation toxicology.
The article also reinforces why air-liquid interface approaches matter so much. In the human lung, inhaled substances contact epithelial surfaces under highly specific exposure conditions that submerged systems fail to reproduce. An air-liquid interface lung model more accurately mimics how aerosols, particles, vapors, or volatile chemicals interact with the respiratory tract. This is particularly valuable when trying to understand early epithelial responses that may diverge between irritation and sensitization pathways. For toxicology teams, this means that test systems designed around realistic lung exposure are more likely to generate data that are both biologically meaningful and useful for internal risk assessment.
For industry and regulatory teams, the implications are direct. First, the paper confirms that respiratory sensitization is moving into the broader transition toward animal-free inhalation testing. Second, it suggests that companies should begin building evidence packages around NAMs that are mechanistically informed and exposure-relevant, rather than waiting for a single regulatory “perfect test” to emerge. Third, it signals that organizations investing early in advanced alveolar epithelial models and human-relevant lung models will be better positioned to support product stewardship, safer-by-design development, and future regulatory expectations.
This matters especially in sectors where inhalation exposure is part of the real-world use scenario. Manufacturers of industrial chemicals, cleaning agents, fragrances, aerosols, coatings, advanced materials, and inhaled therapeutics increasingly need data that can discriminate a respiratory sensitizer vs irritant with greater precision. Overcalling irritation as sensitization can unnecessarily restrict innovation. Missing a true sensitizer can create worker safety, legal, and reputational risk. Better inhalation toxicology models reduce both errors.
The broader takeaway from the paper is that respiratory sensitization assessment is entering a transition phase. The science is not yet fully standardized, but the direction is clear: respiratory hazard identification will increasingly rely on new approach methodologies that combine mechanistic depth with human relevance. Advanced air-liquid interface lung models, alveolar epithelial models, and integrated NAM strategies are likely to play a central role in that shift. For toxicology and regulatory teams, the strategic move is not to wait, but to start incorporating these tools into screening, prioritization, and weight-of-evidence frameworks now.
Citation: Haber LT, Bradley MA, Buerger AN, Behrsing H, Burla S, Clapp PW, Dotson S, Fisher C, Genco KR, Kruszewski FH, McCullough SD, Page KE, Patel V, Pechacek N, Roper C, Sharma M and Jarabek AM (2024) New approach methodologies (NAMs) for the in vitro assessment of cleaning products for respiratory irritation: workshop report. Front. Toxicol. 6:1431790. doi: 10.3389/ftox.2024.1431790