Will alternative assays for developmental toxicity gain regulatory acceptance?
For the first time an international regulatory organization has released safety guidelines that will lead to fewer animal tests for developmental toxicity studies. This is a huge step forward for the field, but are we ready?
In vitro, ex vivo, and non-mammalian in vivo alternative assays are already an important support system in drug development. The guidance recognizes that these assays have “been used as drug discovery screens for adverse effects” on embryo/fetal developmental (EFD) toxicity studies and have assisted in the understanding of the mechanism of toxicity.
When a human specific target is known, these assays can be useful for translating nonclinical data to human risk. In fact, the ICH committee, which regulates the technical requirements for pharmaceuticals for human use, urges the continued use of alternative assays for these purposes. When properly qualified, alternative assays have the potential to defer or replace (in certain circumstances) conventional in vivo studies.
So it seems we could be entering an era where animal tests are no longer the norm in reproductive toxicology.
Are labs and regulators ready?
Yet establishing an alternative assay for regulatory use is not going to be easy. Not by a long shot. The guidance lists 13 points that must be addressed in order to qualify an assay. These points include the perfectly reasonable requests for providing biological plausibility of the model (endpoints), how one reached the endpoint, test data, training sets used for qualification, justification of the use of the assay from the literature and other regulatory agencies who have evaluated these data sets.
While comparison of the test set of chemicals to in vivo test systems will be made, justification of the predictive model will be difficult to establish for most assays. Scientists will need to define which species (e.g. rat, rabbit and/or human) is being mimicked. They will need to demonstrate the performance of the assay compared to data generated from in vivo studies that induce malformation of embryo-fetal lethality (MEFL) in the absence of confounding maternal toxicity. With any compound, it is also important to determine if a chemical is a positive or negative, regardless of the relevance of the dose that humans might be exposed to; understanding that most compounds will be positive at some dose.
Best for the direst prognoses
What is immediately clear is that it will be easiest to use alternative assays when a pharmaceutical is expected to be embryo-fetal toxic, when the pharmaceutical is intended for a severely debilitating or life-threatening disease, or when the pharmaceutical is intended to treat a later-life onset disease. The evaluation of drugs in one of these scenarios are clearly outlined in the guidance.
Annex 2 of the ICH S5 provides extensive information on what is required in order to use an alternative assay. It defines how to apply results when the mechanism of a drug affects a pathway important in development or when exposure in animal models is below that in human. It does not recommend any particular assay but does use basic scientific principles to assist in assay qualification for determining MEFL.
Finally, the guidance does provide extensive information including the animal and/or human data on a reference list of chemicals that can be used for validation of any assay. These data will be very useful for qualifying an assay, but the amount of work needed to test most if not all of these compounds is significant.
So will this guidance promote the use of alternative assays for regulatory purposes or will these assays only be used to add to a weight of evidence for mechanism of action? We shall see. But the bar for regulatory use has been set high for good reason. Animal studies are still the standard, though inroads continue to delay embryo-fetal studies later and later in the testing phase. Scenarios for using alternative assays for certain classes of drugs (those for life threatening conditions or late-life onset diseases) may be feasible.