ABSTRACT
The U-SENS™ is a test method based on the human myeloid U937 cell line to assess the skin sensitisation potential of substances. To demonstrate its robustness, a multicentre validation study with four laboratories testing 24 coded substances has been conducted according to internationally agreed principles. The primary objective of the study was to enlarge the U-SENS™'s reproducibility database. Secondary objectives were to provide additional evidence on its transferability and its predictive capability. Reproducibility within laboratories was approximately 92%, while the reproducibility between laboratories was 87.5%. Predictivity for the 24 validation substances was high, with sensitivity, specificity and accuracy being on average at least 93.8%. Similar performances are obtained for 38 substances when combining the study results with those of an earlier multicentre study, as well as with an automated version of the U-SENS™. With reliability and relevance similar to comparable non-animal skin sensitisation test methods, which have achieved regulatory acceptance, it is concluded that the U-SENS™ is a well reproducible and predictive test method. This profiles the U-SENS™ as a valuable addition to the suite of non-animal testing methods for skin sensitisation with the potential to significantly contribute to the development of integrated testing strategies.