Isotretinoin (Accutane) and pregnancy

Assessment of the effectiveness of animal developmental toxicity testing for human safety

Evaluations of studies for four well-known human developmental toxicants clearly suggest that a margin of exposure of 1/100th the NOAEL for the most sensitive animal species tested provides adequate safety for the human conceptus. The lowest reported human teratogenic exposures occurred at doses at least one log above the estimated "safe" or acceptable daily exposure based on the most sensitive animal species, that is, 1/100th animal NOAEL. (The MOE ranged from < 1 to 10.). The data and analyses are consistent with the conclusion that, regardless of the type of in utero effect produced in animals, the margin of safety of 100 is likely to protect the human conceptus in utero from developmental perturbation, and it is a scientifically reasonable and conservative default number.

Induced and genetic mouse middle ear ossicular malformations: a model for human malformative ossicular diseases and a tool for clarifying their normal ontogenesis

Oral administration of 13-cis retinoic acid (RA) to pregnant mice on the 9th gestation day provokes important malformations of the middle ear ossicles, associated with a general kind of craniofacial dysmorphogenesis evoking the human mandibulofacial dysostosis. The malleus, incus and stapes are affected. The malleus exhibits a handle separated from its head and keeping a persistant relationship with the tubotympanic recess. The stapes makes no contact with the otic capsule. The malformation pattern is visible early as shown by the appearance of an abnormally curved Meckel's cartilage at day 12, followed by the development of atypically shaped ossicular anlagen. The mouse "far" (first arch malformation) mutation is responsible for minor ossicular abnormalities which disrupts the normal relationships between the stapes, Reichert's cartilage and stapedial muscle. The administration of RA to pregnant mice and the comparison with a genetically induced malformation (the mutation far) provides some interesting information about the postulated mechanisms of human middle ear dysmorphogenesis, as well as precious data about the features of normal ossicular primordia formation. The comparison of these features with human middle ear abnormalities as revealed by medical imaging sheds light on human malformation patterns and provides a better understanding of normal and abnormal radiologic ossicular aspects.

Morphogenesis of isotretinoin-induced microcephaly and micrognathia studied by scanning electron microscopy

Isotretinoin ingestion during the first trimester of human pregnancy can induce malformations of the skull, ears, face, central nervous system, eyes, palate, lungs, circulatory system, limbs, and digits. A single oral dose of isotretinoin on day 8 of gestation in hamsters induces a similar syndrome of congenital malformation. The present study concerned scanning electron microscopic (SEM) observation of embryonic and fetal hamster craniofacial structures at 4, 8, 12, 24, 48, and 72 hr after administration of an oral dose of 50 mg/kg isotretinoin or an equivalent volume of the vehicle. The variability in development among control embryos recovered 4 hr after treatment precluded objective assessment of pathologic change by SEM at very early time points. Craniofacial damage was obvious within 8-12 hr of isotretinoin treatment, and it included hypoplasia of the maxillary and mandibular processes of the first branchial arch, a rudimentary second arch, and apparent collapse of the forebrain. Equivalent fusion between the lateral nasal process and the maxillary process and between the medial nasal process and the maxillary process in treated and control embryos accounts for the very low incidence of cleft lip observed in fetuses. The terminal microstomia was not associated with excessive merging or overgrowth of the first arch components. Hypoplasia of the first arch can account for retinoid-induced macrostomia and microstomia.

Teratogen update: vitamin A congeners

Despite animal vitamin A congener teratogenicity in animal studies since 1954, striking human findings only arose in 1983 following isotretinoin (ITR) marketing for oral treatment of severe acne. By November 1985, 44 outcomes with central nervous system (CNS), cardioaortic (CV), microtia, facial palsy, micrognathia, cleft palate, and/or thymic aplasia defects, and 33 spontaneous abortions have been reported. The critical period for exposure appears to be two to five weeks postconception, although this is clinically inexact. ITR half life is less than a day, although a teratogenic metabolite, 4-oxo-isoretinoin, has a half life of several days. Seven defect outcomes and one stillbirth have been reported with another congener, etretinate (ETR), used for psoriasis. Three of these had meningomyeloceles. Half life of several months makes levels cumulative. Only one additional defect, which may have occurred by chance, is reported with use stopped before conception (4 months). Other discontinuations 1 to 6 months before conception had 11 normal outcomes and two spontaneous abortions. ITR and ETR dose ranged from 0.5 to 1.5 mg/kg. Normal outcomes are reported both with ITR and ETR, but some of these appear not to have been exposed during the critical period. Less striking defects, abortions, and normal outcomes are less well reported. Because vitamin A analogs are therapeutically important and unplanned outcome not always avoidable, further animal research is needed for better risk/benefits. Megadose vitamin A (retinol) use is widespread, but experience poorly observed. Eighteen suspicious birth defect outcomes have been reported from pregnancies with high dose exposure. Twelve had findings similar to those seen in animals and in human retinoid syndromes, e.g., CNS, CV, microtia, and clefts. Epidemiological controls are lacking to establish human teratogenicity, but based on animal studies and experience with ITR and ETR, avoiding long term megadose Vitamin A use in fertile women is warranted.