A Tiered Approach to Risk Assess Microbiome Perturbations Induced by Application of Beauty and Personal Care Products

In the consumer goods sector, there is a rapid increase in launches of products that affect the human microbiome. Whilst more and more studies and product claims focus on the health benefits of the manipulation of microbiomes, ensuring that perturbations of the microbiome by the application of beauty and personal care products do not have potential unwanted consequences on the health of consumers is less well described in the scientific literature. There is currently no agreement on approaches to assess the possible impacts on consumer safety nor quantitatively defined endpoints of concern. We propose a 3-tier framework to qualitatively assess the potential impact of skin and oral microbiome perturbations on consumer health. The framework is established in accordance with the next generation risk assessment principles used in toxicology and avoids the use of animal testing. It was developed using a collaborative consultation including oral and skin microbiome experts, bioinformaticians and microbiological risk assessors. The first tier is based on a "history of safe use" concept, where the efficacy of a bioactive of interest is benchmarked against formulations generally regarded as safe because of their long history of consumer use. One of the endpoints identified during the development of the approach is that the microbiome's resilience is not compromised, that is its capacity to respond to challenges without going to dysbiosis. Therefore, the second tier is based on the notion of microbiome stability and its resilience to short term perturbations. The third tier aims to utilise next generation sequencing data and relate these to health status. Whilst 16S rRNA data have brought unprecedented resolution in determining the species present in microbiomes, we illustrate the challenges associated with predicting potential consequences for consumer health and disease from this type of data in a case study. With the development of whole genome sequencing technology and progress with integration of -omics data, we propose that the active functions of the microbiome rather than taxonomic classification should be the basis of a safety assessment. We suggest a research strategy to define the potential endpoints of concern quantitatively, based on a concomitant development of in vitro 3D tissue models in which the host response can be assessed, in silico approaches to describe the microbiome and longitudinal human studies to validate learnings in situ.