Knockout mouse models are predictive of malformations or embryo-fetal death in drug safety evaluations

New approach methodologies to confirm developmental toxicity of pharmaceuticals based on weight of evidence

The aim of embryo-fetal developmental toxicity assessments for pharmaceuticals is to inform potential risk of adverse pregnancy outcome, which has traditionally relied on studies in pregnant animals. Recent updates to international safety guidelines (ICH S5R3) have incorporated information on how to use weight of evidence and alternative assays to reduce animal use while still informing risk of fetal harm. Uptake of these alternative approaches has been slow due to limitations in understanding how alternative assays translate to in vivo effects and then relevance to human exposure. To understand the predictivity of new approach methodologies for developmental toxicity (DevTox NAMs), we used two pharmaceutical examples (glasdegib and lorlatinib) to illustrate the value of DevTox NAMs to complement weight of evidence (WoE) assessments while considering the relationship of concentration-effect levels in NAMs to in vivo studies. The in vitro results generated in a battery of assays (mEST, rWEC, zebrafish, and human based stem cells) confirmed the WoE based on literature and further confirmed by preliminary embryo-fetal development data. The data generated for these two compounds supports integrating DevTox NAMs into the developmental toxicity assessment for advanced cancer indications.

The postnatal resolution of developmental toxicity induced by pharmacological diacylglycerol acyltransferase 2 (DGAT2) inhibition during gestation in rats

Ervogastat (PF-06865571) is a small molecule diacylglycerol acyltransferase 2 (DGAT2) inhibitor being developed for the oral treatment of non-alcoholic steatohepatitis (NASH) with liver fibrosis. DGAT2 is a key enzyme in triglyceride synthesis in tissues and in regulating energy metabolism. Fertility and developmental toxicity studies with ervogastat were conducted in female rats and rabbits. There were no effects on female rat fertility or rabbit embryo-fetal development. Administration of ervogastat to pregnant rats during organogenesis reduced fetal weight and caused higher incidences of bent bones in fetuses that were shown to resolve by postnatal day 28 and were therefore considered to be transient variations secondary to developmental delay. Extended dosing in rats through the end of gestation and lactation (pre- and post-natal development study) caused impaired skin development, reduced offspring viability and growth retardation. The spectrum of developmental effects in rats is consistent with the intended pharmacology (altered triglyceride metabolism) and the transient nature of the skeletal findings, along with the late gestational window of sensitivity for the effects on skin barrier development, reduce the concern for potential adverse developmental effects following unintended early gestational exposure to ervogastat in humans where treatment can be discontinued once pregnancy is determined.

Inhibition of ACC causes malformations in rats and rabbits: comparison of mammalian findings and alternative assays

Acetyl-CoA carboxylase (ACC) is an enzyme within the de novo lipogenesis (DNL) pathway and plays a role in regulating lipid metabolism. Pharmacologic ACC inhibition has been an area of interest for multiple potential indications including oncology, acne vulgaris, metabolic diseases such as type 2 diabetes mellitus, and non-alcoholic fatty liver disease/non-alcoholic steatohepatitis. A critical role for ACC in de novo synthesis of long-chain fatty acids during fetal development has been demonstrated in studies in mice lacking Acc1, where the absence of Acc1 results in early embryonic lethality. Following positive predictions of developmental toxicity in alternative in vitro assays (positive in murine embryonic stem cell [mESC] assay and rat whole embryo culture, but negative in zebrafish), developmental toxicity (growth retardation and dysmorphogenesis associated with disrupted midline fusion) was observed with the oral administration of the dual ACC1 and 2 inhibitor, PF-05175157, in Sprague Dawley rats and New Zealand White rabbits. The results of these studies are presented here to make comparisons across the assays, as well as mechanistic insights from the mESC assay demonstrating high ACC expression in the mESC and that ACC induced developmental toxicity can be rescued with palmitic acid providing supportive evidence for DNL pathway inhibition as the underlying mechanism. Ultimately, while the battery of alternative approaches and weight-of-evidence case were useful for hazard identification, the embryo-fetal development studies were necessary to inform the risk assessment on the adverse fetal response, as malformations and/or embryo fetal lethality were limited to doses that caused near complete inhibition of DNL.