Alternative Integrated Testing for Skin Sensitization: Assuring Consumer Safety

Abstract

Cosmetics legislation in Europe has driven the validation and acceptance of non-animal alternatives, most recently in the area of skin sensitization. Despite use of these methods to meet regulatory needs, it is also essential that they allow evaluation regarding human safety. For cosmetic product safety, it is necessary to understand how they can be used and with what limitations, and thereby reveal what remains to be addressed. A dataset of 165 ingredients (137 cosmetic ingredients +28 reference substances) has been identified, curated, and subjected to testing using accepted in vitro methods, with additional information, including physicochemical data and in silico results. The inputs from multiple determinants of skin sensitizing activity have been used in five individual supervised classification models (or machine learning approaches), which were then collated in a robust statistical manner, a stacking meta-model, to deliver a prediction with an optimized level of confidence. For the training set, with the probability cutoffs at 70% and 30%, predictive sensitivity was 97%, specificity was 88%, the overall accuracy was 93%, and kappa was 85%. A further 52 substances were used to test the effectiveness of the model: the predictive sensitivity was 89%, specificity was 95%, overall accuracy was 91%, and kappa was 82%. In conclusion, this stacking meta-model delivers improved performance and therefore enhanced confidence in the discrimination of skin sensitizers from nonsensitizers. The key remaining gap, prediction of skin sensitization potency, may benefit from a similar approach, maximizing use of evidence from individual strands of prediction, while minimizing the impact of the limitations from any particular one.

Prediction of the skin sensitization potential of polyhexamethylene guanidine phosphate, oligo(2-(2-ethoxy)ethoxyethyl) guanidinium chloride, triclosan, and mixtures of these compounds with the excipient propylene glycol through the local lymph node assay: BrdU-FCM

“…PG was predicted to be negative for SS via the LLNA: BrdU-FCM, consistent with previous reports that used a traditional LLNA or other alternative approaches, such as KeratinoSens™ or U-SENS™ (Del Bufalo et al, 2018; European Medicines Agency, 2017). Since PG is widely used as a vehicle or excipient for carrying active ingredients in consumer products, more discussion has arisen about how the addition of PG to PHMG, PGH, triclosan, or d -limonene solutions could influence the intensity of their SS effects.…”

“…PG is considered to be a mild skin irritant; however, its SS potential has been questioned based on LLNA results showing that it is a nonsensitizer. In any case, very low SS potential has been reported in human subjects (Del Bufalo et al, 2018; European Medicines Agency, 2017; Lessmann et al, 2005).…”

In silico Prediction of Skin Sensitization: Quo vadis?

“…These data were run in 5 different classification models (Boosting, Naive Bayes, support vector machine (SVM), Sparse Partial Least Squares Discriminant Analysis, and Expert Scoring). The validation results were used in the stacking meta-model to evaluate the skin sensitization potential ( Del Bufalo et al, 2018 ). The predictions achieved the accuracies of 93 and 91% for the training set and test set, respectively when compared with LLNA hazard data.…”

Towards a tiered test strategy for plant protection products to address mixture toxicity by alternative approaches in human health assessment

“…Three of the four adverse outcome pathway key events can be addressed by regulatory accepted in vitro methods. 49 Multiple IATAs have been published for skin sensitisation [50][51][52][53][54] and the OECD has published guidance on the background and reporting of skin sensitisation IATA. 49 A quantitative evaluation of approaches designed for hazard identification has been summarised by Kleinstreuer et al, 55 who report accuracies of 70.1% and 88.3% for the BASF 'two out of three' approach 50 and 88.3% for the Interagency Coordinating Committee for the Validation of Alternative Methods (ICCVAM) support vector machine approach, 54 respectively.…”