Skin sensitization risk assessment model using artificial neural network analysis of data from multiple in vitro assays

The sensitizing potential of chemicals is usually identified and characterized using in vivo methods such as the murine local lymph node assay (LLNA). Due to regulatory constraints and ethical concerns, alternatives to animal testing are needed to predict skin sensitization potential of chemicals. For this purpose, combined evaluation using multiple in vitro and in silico parameters that reflect different aspects of the sensitization process seems promising.We previously reported that LLNA thresholds could be well predicted by using an artificial neural network (ANN) model, designated iSENS ver.1 (integrating in vitro sensitization tests version 1), to analyze data obtained from two in vitro tests: the human Cell Line Activation Test (h-CLAT) and the SH test. Here, we present a more advanced ANN model, iSENS ver.2, which additionally utilizes the results of antioxidant response element (ARE) assay and the octanol–water partition coefficient (Log P, reflecting lipid solubility and skin absorption). We found a good correlation between predicted LLNA thresholds calculated by iSENS ver.2 and reported values. The predictive performance of iSENS ver.2 was superior to that of iSENS ver.1. We conclude that ANN analysis of data from multiple in vitro assays is a useful approach for risk assessment of chemicals for skin sensitization.     

Proposal for a risk assessment methodology for skin sensitization based on sensitization potency data

In this paper, we propose a quantitative risk assessment methodology for skin sensitization aiming at the derivation of 'safe' exposure levels for sensitizing chemicals, used e.g., as ingredients in consumer products. Given the limited number of sensitizers tested in human sensitization tests, such as the human repeat-insult patch test (HRIPT) or the human maximization test (HMT), we used EC3 values from the local lymph node assay (LLNA) in mice because they provide the best quantitative measure of the skin sensitizing potency of a chemical. A comparison of LLNA EC3 values with HRIPT and HMT LOEL, and NOEL values was carried out and revealed that the EC3, expressed as area dose, can be used as a surrogate value for the human NOEL in risk assessment. The uncertainty/extrapolation factor approach was used to derive (a) an 'acceptable non-sensitizing area dose' (ANSAD) to protect non-allergic individuals against skin sensitization and (b) an 'acceptable non-eliciting area dose' (ANEAD) to protect allergic individuals against elicitation of allergic contact dermatitis. For ANSAD derivation, interspecies, intraspecies and time extrapolation factors are applied to the LLNA EC3. For ANEAD derivation, additional application of a variable sensitization-elicitation extrapolation factor is proposed. Values for extrapolation factors are derived and discussed, the proposed methodology is applied to the sensitizers methylchloroisothiazolinone/methylisothiazolinone, cinnamic aldehyde and nickel and results are compared to published risk assessments.     

A dataset on 145 chemicals tested in alternative assays for skin sensitization undergoing prevalidation

Skin sensitization is a key endpoint for cosmetic ingredients, with a forthcoming ban for animal testing in Europe. Four alternative tests have so far been submitted to ECVAM prevalidation: (i) MUSST and (ii) h‐Clat assess surface markers on dendritic cell lines, (iii) the direct peptide reactivity assay (DPRA) measures reactivity with model peptides and (iv) the KeratinoSens^TM assay which is based on detection of Nrf2‐induced luciferase. It is anticipated that only an integrated testing strategy (ITS) based on a battery of tests might give a full replacement providing also a sensitization potency assessment, but this concept should be tested with a data‐driven analysis. Here we report a database on 145 chemicals reporting the quantitative endpoints measured in a U937‐ test, the DPRA and KeratinoSens^TM . It can serve to develop data‐driven ITS approaches as we show in a parallel paper and provides a view as to the current ability to predict with in vitro tests as we are entering 2013. It may also serve as reference database when benchmarking new molecules with in vitro based read‐across and find use as a reference database when evaluating new tests. The tests and combinations thereof were evaluated for predictivity, and overall a similar predictivity was found as before on three‐fold smaller datasets. Analysis of the dose–response parameters of the individual tests indicates a correlation to sensitization potency. Detailed analysis of chemicals false‐negative and false‐positive in two tests helped to define limitations in the tests but also in the database derived from animal studies. Copyright © 2013 John Wiley & Sons, Ltd.     

Prediction of skin sensitization potency of chemicals by human Cell Line Activation Test (h-CLAT) and an attempt at classifying skin sensitization potency

The human Cell Line Activation Test (h-CLAT), an in vitro skin sensitization test, is based on the augmentation of CD86 and CD54 expression in THP-1 cells following exposure to chemicals. The h-CLAT was found to be capable of determining the hazard of skin sensitization. In contrast, the local lymph node assay (LLNA), widely used as a stand-alone method in Europe and US, identifies the same hazard, but also classifies the potency by using the estimated concentration of SI = 3 (EC3). In this study, several values calculated from the h-CLAT data were evaluated for its correlation to the LLNA EC3 determination. A statistically significant correlation was observed between h-CLAT concentration providing a cell viability of 75% (CV75), h-CLAT estimated concentration of RFI = 150 for CD86 (EC150), and for CD54 (EC200) with LLNA’s EC3. From EC150 and EC200, a minimum induction threshold (MIT) was determined as the smaller of either EC150 or EC200. MIT showed a correlation with EC3 (R = 0.638). Also, MIT had an approximate 80% accuracy for sub-categories of the globally harmonized system (GHS) when a tentative threshold of 13 μg/mL was used. From these data, the h-CLAT values may be one of the useful tools to predict the allergic potency of chemicals.     

Validation of in vitro potency assays for tetanus immunoglobulin

The European Pharmacopoeia (Ph. Eur.) monograph Human tetanus immunoglobulin (0398) gives a clear outline of the in vivo assay to be performed to determine the potency of human tetanus immunoglobulins during their development. Furthermore, it states that an in vitro method shall be validated for the potency estimation. Since no further guidance is given on the in vitro assay, every control laboratory concerned is free to design and validate an in-house method. At the moment there is no agreed method available. The aim of this study was to validate and compare 2 alternative in vitro assays, i.e. an enzyme-linked immunoassay (EIA) and a toxoid inhibition assay (TIA). The potency of 2 tetanus immunoglobulin preparations (Product 1, Product 2) was estimated against the WHO International Standard for tetanus immunoglobulin, using the tetanus EIA and TIA. The coefficient of variation (CV) to characterise the assay precision was 3.2% (EIA) and 3.6% (TIA), and the corresponding CV for intra-assay variation was 4.7% (EIA) and 5.5% (TIA). Using a spiking procedure, the 2nd part of the experiment investigated recovery of a known anti-tetanus potency. The recovery of samples spiked with defined amounts of reference preparation ranged from 104 112% (EIA) and 114 125% (TIA) respectively, resulting in a mean bias of 2.2 IU/ml (95% confidence interval (CI): -1.1-5.4 IU/ml, EIA) and 5.8 IU/ml (95% CI: 1.4 10.2 IU/ml, TIA). Good agreement was observed between the in vivo and in vitro assay results: the relative potency results of the EIA and TIA as compared to those of the in vivo assay performed by the manufacturers of the 2 tetanus immunoglobulins were for the EIA in the range of 104+/-10% for Product 1 and 100+/-6% for Product 2, and for the TIA in the range of 107+/-6% for Product 1 and 100+/-7% for Product 2. Tetanus EIA and TIA are suitable quality control methods for polyclonal tetanus immunoglobulin, which can be standardised in a quality control laboratory using a quality assurance system. In a collaborative study it will now be evaluated whether the validated methods can be proposed as common in vitro batch potency assays for replacement of the in vivo mouse assay.     

Updating the skin sensitization in vitro data assessment paradigm in 2009

Approaches to the interpretation of guinea pig skin sensitization data for both hazard identification and potency assessment have been understood for many years. More recently, the local lymph node assay has to a large extent replaced the earlier guinea pig assays, not least because it provides a more clearly defined and transparent means of identifying hazard, and the ability to measure relative skin sensitization potency. However, beginning in 2009 there will be considerable pressure replace all in vivo assays for skin sensitization with alternative approaches that do not require the use of animals (in vitro and/or in silico methods). As there is a common view that multiple assays will be needed to achieve complete replacement of the in vivo tests, a strategy for the integration of the available data will be required. There has been at least one previous attempt to develop a framework that would provide for integration of relevant information from different sources to reach informed decisions about skin sensitization potential and potency. It is timely now, in the light of recent developments and initiatives, to revisit this paradigm with a view to developing recommendations for modification and refinement. In addition to this, the need for performance standards and an agreed ‘gold standard’ dataset against which to validate both alternatives and new prediction models is discussed. Copyright © 2009 John Wiley & Sons, Ltd.     

Multivariate models for prediction of human skin sensitization hazard

One of the Interagency Coordinating Committee on the Validation of Alternative Method's (ICCVAM) top priorities is the development and evaluation of non‐animal approaches to identify potential skin sensitizers. The complexity of biological events necessary to produce skin sensitization suggests that no single alternative method will replace the currently accepted animal tests. ICCVAM is evaluating an integrated approach to testing and assessment based on the adverse outcome pathway for skin sensitization that uses machine learning approaches to predict human skin sensitization hazard. We combined data from three in chemico or in vitro assays – the direct peptide reactivity assay (DPRA), human cell line activation test (h‐CLAT) and KeratinoSens™ assay – six physicochemical properties and an in silico read‐across prediction of skin sensitization hazard into 12 variable groups. The variable groups were evaluated using two machine learning approaches, logistic regression and support vector machine, to predict human skin sensitization hazard. Models were trained on 72 substances and tested on an external set of 24 substances. The six models (three logistic regression and three support vector machine) with the highest accuracy (92%) used: (1) DPRA, h‐CLAT and read‐across; (2) DPRA, h‐CLAT, read‐across and KeratinoSens; or (3) DPRA, h‐CLAT, read‐across, KeratinoSens and log P. The models performed better at predicting human skin sensitization hazard than the murine local lymph node assay (accuracy 88%), any of the alternative methods alone (accuracy 63–79%) or test batteries combining data from the individual methods (accuracy 75%). These results suggest that computational methods are promising tools to identify effectively the potential human skin sensitizers without animal testing. Published 2016. This article has been contributed to by US Government employees and their work is in the public domain in the USA.     

Intralaboratory validation of four in vitro assays for the prediction of the skin sensitizing potential of chemicals

Allergic contact dermatitis is induced by repeated skin contact with an allergen. Assessment of the skin sensitizing potential of chemicals, agrochemicals, and especially cosmetic ingredients is currently performed with the use of animals. Animal welfare and EU legislation demand animal-free alternatives reflected in a testing and marketing ban for cosmetic ingredients beginning in 2013. The underlying mechanisms of induction and elicitation of skin sensitization are complex and a chemical needs to comply several properties being skin sensitizing. To account for the multitude of events in the induction of skin sensitization an in vitro test system will consist of a battery of various tests.Currently, we performed intralaboratory validations of four assays addressing three different events during induction of skin sensitization. (1) The Direct Peptide Reactivity Assay (DPRA) according to Gerberick and co-workers (Gerberick et al., 2004) using synthetic peptides and HPLC analysis. (2) Two dendritic cell activation assays based on the dendritic cell like cell lines U-937 and THP-1 and flow cytometric detection of the maturation markers CD54 and/or CD86 ( [Ashikaga et al., 2006], [Python et al., 2007] and [Sakaguchi et al., 2006]). (3) Antioxidant response element (ARE)-dependent gene activity in a HaCaT reporter gene cell line (Emter et al., 2010). We present the results of our intralaboratory validation of these assays with 23 substances of known sensitizing potential. The sensitivity, specificity, and accuracy of the individual tests were obtained by comparison to human epidemiological data as well as to data from animal tests such as the local lymph node assay.     

The role of non-covalent protein binding in skin sensitisation potency of chemicals

Skin sensitisation is a delayed hypersensitivity reaction caused by repeated exposure to common natural and synthetic chemical allergens. It is thought that small chemical sensitisers (haptens) are required to form a strong irreversible bond with a self protein/peptide and generate an immunogenic hapten-protein complex in order to be recognised by the immune system and stimulate T cell proliferation. The sensitisers are usually electrophilic chemicals that are directly reactive with proteins or reactive intermediates (metabolites) of chemically inert compounds (prohaptens). Sensitising chemicals are also capable of weak, non-covalent association with proteins and there is an ongoing debate about the role of weak interactions of chemicals and proteins in the chemistry of allergy. The non-covalent interactions are reversible and thus have a major impact on skin/epidermal bioavailability of chemical/reactive metabolites. We investigated the relationship between the relative level of non-covalent association to a model protein and their relative potencies as determined by the EC3 values in the murine local lymph node assay (LLNA) for a number of chemicals. Using human serum albumin as a model protein, we determined that no observable relationship exists between the two parameters for the chemicals tested. Therefore, at least for this model protein, non-covalent interactions appear not to be a key determinant of allergen potency.     

Methods for predicting in vivo pharmacokinetics using data from in vitro assays

Strategies for optimising in vivo predictions from in vitro data on metabolic stability and CYP inhibition are discussed. Potential pitfalls and areas of inaccuracy are highlighted together with recommendations for best practice. The use of both hepatic microsomes and isolated hepatocytes for the assessment of metabolic stability is discussed in terms of scaling from the in vitro system up to whole liver. The importance of integrating metabolic stability data together with other drug pharmacokinetic characteristics (e.g., protein binding and red blood cell uptake) as well as blood flow are presented within the context of different liver models. The assessment of CYP inhibition potential requires in vitro data on the inhibitor potency either in the form of Ki (for reversible inhibition) or KI and kinact (for time-dependent inhibition). The integration of these in vitro parameters together with other pharmacokinetic information is essential for the in vivo prediction. While a qualitative assessment may be made from the I/Ki ratio, a number of additional victim drug and enzyme-related parameters are required for quantitative prediction. Of particular importance is the parameter fmCYP (the fraction of the metabolic clearance of the victim drug that is catalyzed by the enzyme subject to the inhibition). Impact of other victim drug properties (e.g., fractional importance of the intestine) and enzyme properties (e.g., kdeg for time-dependent inhibition) on the drug-drug interaction prediction is discussed. In addition, mechanisms by which false negatives and false positives may result from in vitro strategies are summarized. Finally perspectives for future application and improvements in these predictions strategies are outlined.     

Accounting for data variability,a key factor in in vivo/in vitro relationships: application to the skin sensitization potency (in vivo LLNA versus in vitro DPRA) example

When searching for alternative methods to animal testing, confidently rescaling an in vitro result to the corresponding in vivo classification is still a challenging problem. Although one of the most important factors affecting good correlation is sample characteristics, they are very rarely integrated into correlation studies. Usually, in these studies, it is implicitly assumed that both compared values are error‐free numbers, which they are not. In this work, we propose a general methodology to analyze and integrate data variability and thus confidence estimation when rescaling from one test to another. The methodology is demonstrated through the case study of rescaling the in vitro Direct Peptide Reactivity Assay (DPRA) reactivity to the in vivo Local Lymph Node Assay (LLNA) skin sensitization potency classifications. In a first step, a comprehensive statistical analysis evaluating the reliability and variability of LLNA and DPRA as such was done. These results allowed us to link the concept of gray zones and confidence probability, which in turn represents a new perspective for a more precise knowledge of the classification of chemicals within their in vivo OR in vitro test. Next, the novelty and practical value of our methodology introducing variability into the threshold optimization between the in vitro AND in vivo test resides in the fact that it attributes a confidence probability to the predicted classification. The methodology, classification and screening approach presented in this study are not restricted to skin sensitization only. They could be helpful also for fate, toxicity and health hazard assessment where plenty of in vitro and in chemico assays and/or QSARs models are available. Copyright © 2016 John Wiley & Sons, Ltd.     

Evaluation of in vitro assays for the assessment of the skin sensitization hazard of functional polysiloxanes and silanes

The skin sensitization potential of chemicals has traditionally been evaluated in vivo according to OECD testing guidelines in guinea pigs or the mouse local lymph node assay. There has lately been a great emphasis on establishing in vitro test methods reflecting the key biological events in the adverse outcome pathway (AOP) for skin sensitization as published by the OECD. Against this background, a group of 8 polysiloxanes and silanes, seven of them aminofunctionalised, for which in vivo data were already available, has been tested in vitro in the direct peptide reactivity assay (DPRA), the KeratinoSens™ and the human cell line activation test (h-CLAT) and in the modified myeloid U937 skin sensitization test (mMUSST) as far as technically feasible. The main objective of the programme was to determine the utility of these systems for this heterogeneous group of silicone-based substances, recognizing that some substances are outside the assays applicability domains. The presented data provided some interesting mechanistical insights into the performance of these assays for functionalised siloxanes and silanes. The data also allow for a preliminary evaluation of proposed integrated testing strategies (ITS) to determine the skin sensitization potential of chemicals which were not considered in the training sets of the respective ITS.     

An in vitro human skin test for assessing sensitization potential

Sensitization to chemicals resulting in an allergy is an important health issue. The current gold‐standard method for identification and characterization of skin‐sensitizing chemicals was the mouse local lymph node assay (LLNA). However, for a number of reasons there has been an increasing imperative to develop alternative approaches to hazard identification that do not require the use of animals. Here we describe a human in‐vitro skin explant test for identification of sensitization hazards and the assessment of relative skin sensitizing potency. This method measures histological damage in human skin as a readout of the immune response induced by the test material. Using this approach we have measured responses to 44 chemicals including skin sensitizers, pre/pro‐haptens, respiratory sensitizers, non‐sensitizing chemicals (including skin‐irritants) and previously misclassified compounds. Based on comparisons with the LLNA, the skin explant test gave 95% specificity, 95% sensitivity, 95% concordance with a correlation coefficient of 0.9. The same specificity and sensitivity were achieved for comparison of results with published human sensitization data with a correlation coefficient of 0.91. The test also successfully identified nickel sulphate as a human skin sensitizer, which was misclassified as negative in the LLNA. In addition, sensitizers and non‐sensitizers identified as positive or negative by the skin explant test have induced high/low T cell proliferation and IFNγ production, respectively. Collectively, the data suggests the human in‐vitro skin explant test could provide the basis for a novel approach for characterization of the sensitizing activity as a first step in the risk assessment process. Copyright © 2015 John Wiley & Sons, Ltd.     

A comparison of reactivity schemes for the prediction skin sensitization potential

Skin sensitization is an important toxic end point for both regulatory frameworks and safety assessment. There are many hurdles for a chemical to overcome in terms of inducing skin sensitization, although the binding of chemicals to skin protein is thought to be the rate-determining step. Current strategies to predict the skin sensitization potential of chemicals in silico is through the identification of electrophilic characteristics. A number of predictive schemes have been developed in recent years, some based on broad structural rules and some with a reaction chemistry mechanistic basis. This work compares two schemes that are based on reaction chemistry. The first scheme comprises a set of rules that characterize reaction mechanistic domains as proposed by Aptula and Roberts [(2006) Chem. Res. Toxicol. 19, 1097-1105]. The second is a set of structure-toxicity and structure-metabolism pathways that are encoded and embedded into the TIssue MEtabolism Simulator skin sensitization model (TIMES-SS) [(2005) Int. J. Toxicol. 24, 189-204]. Here, a comparison of these schemes has been made using a recently published data set of 210 chemicals that have been tested in the local lymph node assay. The similarities and differences of the schemes are highlighted, together with modifications that could be made to TIMES-SS to harmonize the two approaches.     

Artificial neural network prediction of the patterns of deposition of polydisperse aerosols within human lungs

For optimum therapeutic response from drug administered to the lungs, it is paramount that the aerosolised drug is able to deposit in the lower airways. The filtering characteristics of the respiratory tract, however, make this a particularly challenging task. Computational tools afford a cost-effective means of studying the problem, and here we report on the development of a rapid and reliable method for predicting the pattern of deposition of polydisperse aerosols within human lungs using artificial neural networks (ANNs). Literature (experimental) data on lung deposition of monodisperse aerosols were used to train a single ANN to allow for simultaneous predictions of regional and total aerosol particle deposition patterns in human lungs. When used in modelling the fate of polydisperse aerosols in human lungs, the trained ANN was found to give highly accurate predictions for all lung regions, and all (pharmaceutically relevant) particle sizes and breathing conditions (with errors typically <0.025%). Further testing of the ANN, using 'unseen' in vitro and in vivo data, gave good agreement of lung dosages. It is thus concluded that the ANN produced can be used to provide highly reliable estimates of particle deposition from polydisperse pharmaceutical aerosols generated from breath-actuated dry powder inhalers, nebulizers and metered dose inhalers with spacers.     

Pre-validation of SENS-IS assay for in vitro skin sensitization of medical devices

According to ISO 10993-1:2018, the skin sensitization potential of all medical devices must be evaluated, and for this endpoint ISO 10993-10:2010 recommends the use of in vivo assays. The goal of the present study was to determine if the in vitro SENS-IS assay could be a suitable alternative to the current in vivo assays. The SENS-IS assay uses the Episkin Large and SkinEthic RHE reconstructed human epidermis models to evaluate marker genes. In our study, the SENS-IS assay correctly identified 13 sensitizers spiked in a non-polar solvent. In a subsequent analysis six medical device silicone samples previously impregnated with sensitizers were extracted with polar and non-polar solvents. The SENS-IS assay correctly identified five of these extracts, while a sixth extract, which contained the weak sensitizer phenyl benzoate, was classified as negative. However, when this extract was concentrated, or a longer exposure time was used, the assay was able to detect phenyl benzoate. The SENS-IS assay was transferred to a naïve laboratory which correctly identified sensitizers in six blinded silicone samples, including the one containing phenyl benzoate. In light of these results, we conclude that the SENS-IS assay is able to correctly identify the presence of sensitizers in medical devices extracts.     

The prediction of methylmercury elimination half-life in humans using animal data: a neural network/rough sets analysis

Artificial neural networks and Rough Sets methodology have been utilized to predict human pharmacokinetic elimination half-life data based on animal data training sets. Methylmercury (Hg) pharmacokinetic data was obtained from 37 literature references, which provided data on species, gender, age, weight, route of administration, dose, dose frequency, and elimination half-life based on either whole-body Hg analysis or blood Hg analysis. Data were categorized into various formats for analysis comparisons. Rough Sets methodology was utilized to identify and remove redundant independent variables. Artificial neural networks were used to produce models based on the animal data, which were in turn used to predict and compare to the human elimination half-life values. These neural network predictions were compared to allometric graphical plots of the same data. The best artificial neural network prediction was based on a "thermometer" categorical representation of the data.