Dimethadione embryotoxicity in the rat is neither correlated with maternal systemic drug concentrations nor embryonic tissue levels.

Pregnant rats treated with dimethadione (DMO), the N-demethylated metabolite of the anticonvulsant trimethadione, produce offspring having a 74% incidence of congenital heart defects (CHD); however, the incidence of CHD has high inter-litter variability (40-100%) that presents a challenge when studying the initiating events prior to the presentation of an abnormal phenotype. We hypothesized that the variability in CHD incidence was the result of differences in maternal systemic concentrations or embryonic tissue concentrations of DMO. To test this hypothesis, dams were administered 300 mg/kg DMO every 12h from the evening of gestational day (GD) 8 until the morning of GD 11 (six total doses). Maternal serum levels of DMO were assessed on GD 11, 12, 13, 14, 15, 18 and 21. Embryonic tissue concentrations of DMO were assessed on GD 11, 12, 13 and 14. In a separate cohort of GD 12 embryos, DMO concentrations and parameters of growth and development were assessed to determine if tissue levels of DMO were correlated with these endpoints. Embryos were exposed directly to different concentrations of DMO with whole embryo culture (WEC) and their growth and development assessed. Key findings were that neither maternal systemic concentrations nor tissue concentrations of DMO identified embryos that were sensitive or resistant to DMO in vivo. Direct exposure of embryos to DMO via WEC also failed to show correlations between embryonic concentrations of DMO with developmental outcomes in vitro. We conclude that neither maternal serum nor embryonic tissue concentrations of DMO predict embryonic outcome.

In Utero Exposure to a Cardiac Teratogen Causes Reversible Deficits in Postnatal Cardiovascular Function, But Altered Adaptation to the Burden of Pregnancy.

Congenital heart defects (CHD) are the most common birth anomaly and while many resolve spontaneously by 1 year of age, the lifelong burden on survivors is poorly understood. Using a rat model of chemically induced CHD that resolve postnatally, we sought to characterize the postnatal changes in cardiac function, and to investigate whether resolved CHD affects the ability to adapt to the increased the cardiovascular (CV) burden of pregnancy. To generate rats with resolved CHD, pregnant rats were administered distilled water or dimethadione (DMO) [300 mg/kg b.i.d. on gestation day (gd) 9 and 10] and pups delivered naturally. To characterize structural and functional changes in the heart, treated and control offspring were scanned by echocardiography on postnatal day 4, 21, and 10-12 weeks. Radiotelemeters were implanted for continuous monitoring of hemodynamics. Females were mated and scanned by echocardiography on gd12 and gd18 during pregnancy. On gd18, maternal hearts were collected for structural and molecular assessment. Postnatal echocardiography revealed numerous structural and functional differences in treated offspring compared with control; however, these resolved by 10-12 weeks of age. The CV demand of pregnancy revealed differences between treated and control offspring with respect to mean arterial pressure, CV function, cardiac strain, and left ventricular gene expression. In utero exposure to DMO also affected the subsequent generation. Gd18 fetal and placental weights were increased in treated F2 offspring. This study demonstrates that in utero chemical exposure may permanently alter the capacity of the postnatal heart to adapt to pregnancy and this may have transgenerational effects.

Maternal rat serum concentrations of dimethadione do not explain intra-litter differences in the incidence of dimethadione-induced birth defects, including novel findings in foetal lung.

To investigate mechanisms of chemical-induced congenital heart defects (CHD) we have developed a rat model using dimethadione (DMO), the N-demethylated metabolite of the anticonvulsant, trimethadione (TMD). Dosing pregnant rats with 300mg/kg DMO every 12h from the evening of gestational day (GD) 8 until the morning of GD 11 (six total doses) produces a mean 74% incidence of CHD with inter litter variability ranging from 40 to 100%. The goal of this study was to determine if the variability in maternal serum concentrations of DMO on GD 14, a surrogate marker for total exposure, was related to the inter-litter differences in teratogenic outcomes. To test this hypothesis, pregnant rats were dosed as described above and serum levels of DMO assessed on GD 14. On GD 21, foetuses were collected by caesarean section, assessed for a number endpoints and the outcomes were correlated with the GD 14 serum concentrations of DMO. DMO exposure was associated with decreased foetal body weight, increased incidence of sternal defects and CHD, but these endpoints were not meaningfully correlated with maternal concentrations of DMO. Novel findings were decreased viability as measured one-hour following caesarean section, and delayed alveolar maturation. The major conclusions from these studies were first, that serum DMO concentrations on GD 14 did not predict teratogenicity, and second, delayed lung development may contribute to the decreased survival of foetuses at the time of caesarean section.

In utero dimethadione exposure causes postnatal disruption in cardiac structure and function in the rat.

In utero exposure of rat embryos to dimethadione (DMO), the N-demethylated teratogenic metabolite of the anticonvulsant trimethadione, induces a high incidence of cardiac heart defects including ventricular septal defects (VSDs). The same exposure regimen also leads to in utero cardiac functional deficits, including bradycardia, dysrhythmia, and a reduction in cardiac output (CO) and ejection fraction that persist until parturition (10 days after the final dose). Despite a high rate of spontaneous postnatal VSD closure, we hypothesize that functional sequelae will persist into adulthood. Pregnant Sprague Dawley rats were administered six 300 mg/kg doses of DMO, one every 12 h in mid-pregnancy beginning on the evening of gestation day 8. Postnatal cardiac function was assessed in control (CTL) and DMO-exposed offspring using radiotelemetry and ultrasound at 3 and 11 months of age, respectively. Adult rats exposed to DMO in utero had an increased incidence of arrhythmia, elevated blood pressure and CO, greater left ventricular volume and elevated locomotor activity versus CTL. The mean arterial pressure of DMO-exposed rats was more sensitive to changes in dietary salt load compared with CTL. Importantly, most treated rats had functional deficits in the absence of a persistent structural defect. It was concluded that in utero DMO exposure causes cardiovascular deficits that persist into postnatal life in the rat, despite absence of visible structural anomalies. We speculate this is not unique to DMO, suggesting possible health implications for infants with unrecognized gestational chemical exposures.

Co-variation in frequency and severity of cardiovascular and skeletal defects in Sprague-Dawley rats after maternal administration of dimethadione, the N-demethylated metabolite of trimethadione.

BACKGROUND: The anticonvulsant trimethadione is a potent inducer of ventricular septation defects, both clinically and in rodents. Teratogenicity requires its N-demethylation to dimethadione, the proximate teratogen. It was previously demonstrated trimethadione only induced membranous ventricular septation defects in rat (Fleeman et al., 2004), and our present goal is to determine whether direct administration of dimethadione increases the incidence and severity of septation defects. METHODS: Pregnant Sprague-Dawley rats were divided into five groups and administered either distilled water (control) or four different regimens of dimethadione. The core treatment was 300 mg/kg dimethadione b.i.d. on gestation day 9, 10 with additional groups given one additional dose of dimethadione 12 hr earlier, 12 hr later or two additional doses 12 hr earlier and later. Caesarian sections occurred on gestation day 21 and fetuses were examined for standard developmental toxicity endpoints. RESULTS: The broadest dosing regimen yielded the highest incidence and the most severe heart and axioskeletal findings with a decrease in mean fetal body weight. The overall incidence of ventricular septation defects was 74%, of which 68% were membranous and 9% muscular. Outflow tract anomalies (17%) were also observed, as were malformations of the axioskeleton (97%), but not of the long bones, and of particular interest was the high incidence of sternoschesis. CONCLUSIONS: Unlike trimethadione, dimethadione induces more serious muscular septation defects that are believed to be more clinically relevant. This, when taken together with the high incidence of total septation anomalies suggests dimethadione is useful for the study of chemically induced ventricular septation defects.

Polytherapy with hERG-blocking antiepileptic drugs: increased risk for embryonic cardiac arrhythmia and teratogenicity.

BACKGROUND: The antiepileptic drugs (AEDs) phenytoin, phenobarbital, dimethadione, and carbamazepine cause a similar pattern of malformations in humans, with an increased risk after polytherapy. The teratogenicity has been linked to cardiac rhythm disturbances and hypoxic damage as a consequence of their common potential to inhibit a specific potassium ion current (IKr). The IKr is of major importance for embryonic cardiac repolarization and rhythm regulation. This study investigated whether these AEDs cause irregular rhythm and if various combinations of AEDs result in higher arrhythmia risk than exposure to a single AED. METHODS: The effects on heart rhythm of a single AED (monotherapy), and of various combinations (polytherapy) of AEDs, in gestational day 10 C57BL mouse embryos in culture were analyzed and graphically illustrated during a 25 s recording with a digitalization technique. RESULTS: All of the studied AEDs caused increased intervals between heartbeats (resulting in bradycardia) and large variations in the interval between heartbeats (resulting in irregular rhythm) in a concentration-dependent manner in cultured mouse embryos. Dimethadione caused irregular rhythm at concentrations within and phenytoin slightly above the therapeutic ranges. Polytherapy resulted in more substantial prolongation of the mean interval between heartbeats (>60 ms) than monotherapy at clinically relevant concentrations. CONCLUSIONS: The results suggest that polytherapy more than monotherapy causes substantial prolongation of the cardiac repolarization, a marker associated with high risk of developing irregular rhythm during longer exposure periods (days to months). This supports the idea that the increased risk for malformations following polytherapy is linked to an increased risk for cardiac rhythm disturbances.

Embryonic cardiac arrhythmia and generation of reactive oxygen species: common teratogenic mechanism for IKr blocking drugs.

In the adult organism, it is well established that hypoxia followed by reperfusion may be fatal and result in generation of reactive oxygen species (ROS) and subsequent tissue damage. There is also considerable evidence that temporary decrease or interruption in oxygen supply to the embryo and ROS generation during reperfusion result in tissue damage in embryonic tissues. A wide spectrum of different malformations by transient embryonic hypoxia could be produced, depending on the duration, extent, and timing of the hypoxic event. It is the contention of this paper that drugs that block the potassium channel IKr, either as an intended pharmacologic effect or as an unwanted side-effect, are potentially teratogenic by a common ROS related mechanism. Drugs blocking the IKr channel, such as almokalant, dofetilide, phenytoin, cisapride and astemizole, do all produce a similar pattern of hypoxia-related malformations. Mechanistic studies show that the malformations are preceded by embryonic cardiac arrhythmia and periods of hypoxia/reoxygenation in embryonic tissues. Pretreatment or simultaneous treatment with radical scavengers with capacity to capture ROS, markedly decrease the teratogenicity of different IKr blocking drugs. A second aim of this review is to demonstrate that the conventional design of teratology studies is not optimal to detect malformations caused by IKr blocking drugs. Repeated high doses result in high incidences of embryonic death due embryonic cardiac arrhythmia, thus masking their teratogenic potential. Instead, single dosing on specific days is proposed to be a better way to characterize the teratogenic potential of Ikr blocking drugs.

Embryonic arrhythmia by inhibition of HERG channels: a common hypoxia-related teratogenic mechanism for antiepileptic drugs?

PURPOSE: There is evidence that drug-induced embryonic arrhythmia initiates phenytoin (PHT) teratogenicity. The arrhythmia, which links to the potential of PHT to inhibit a specific potassium channel (Ikr), may result in episodes of embryonic ischemia and generation of reactive oxygen species (ROS) at reperfusion. This study sought to determine whether the proposed mechanism might be relevant for the teratogenic antiepileptic drug trimethadione (TMO). METHODS: Effects on embryonic heart rhythm during various stages of organogenesis were examined in CD-1 mice after maternal administration (125-1,000 mg/kg) of dimethadione (DMO), the pharmacologically active metabolite of TMO. Palatal development was examined after administration of a teratogenic dose of DMO and after simultaneous treatment with DMO and a ROS-capturing agent (alpha-phenyl-N-tert-butyl-nitrone; PBN). The Ikr blocking potentials of TMO and DMO were investigated in HERG-transfected cells by using voltage patch-clamping tests. RESULTS: DMO caused stage-specific (gestation days 9-13 only) and dose-dependent embryonic bradycardia and arrhythmia at clinically relevant maternal plasma concentrations (3-11 mM). Hemorrhage in the nasopharyngeal part of the embryonic palate (within 24 h) preceded cleft palate in fetuses at term. Simultaneous treatment with PBN significantly reduced the incidence of DMO-induced cleft palate, from 40 to 13%. Voltage patch-clamping studies showed that particularly DMO (70% inhibition), but also TMO, had Ikr blocking potential at clinically relevant concentrations. CONCLUSIONS: TMO teratogenicity, in the same way as previously shown for PHT, was associated with Ikr-mediated episodes of embryonic cardiac arrhythmia and hypoxia/reoxygenation damage.

Abnormalities in ureter and kidney development in mice given acetazolamide-amiloride or dimethadione (DMO) during embryogenesis.

These experiments more accurately define the effects of the combination acetazolamide-amiloride or a single dose of dimethadione (DMO), the active metabolite of trimethadione, on the development of the ureter. When acetazolamide-amiloride was administered in C57BL/6NCrlBR mice on day 9, 9.5, or 10 of gestation (plug = day 0) a second ureter was formed, anterior to the original ureter, inducing a second kidney. The second ureter then fails to make a connection with the developing bladder and remains attached to the mesonephric duct. The mesonephric duct becomes the vas deferens in the male and deteriorates completely in the female leading to either a restricted ureter or a blocked ureter depending on the sex of the fetus. Administration of a single dose of DMO between gestational day 9 and 10.3 produced both renal agenesis and ureters of varying lengths. Some ureters were of normal length with a tuft of one or two nephrons at their tip, while others were one half or one quarter of their normal length. In some instances the ureter was completely absent. The reason for this strong effect on the ureter is unknown.

Effects of the oxazolidinedione anticonvulsants trimethadione and dimethadione and the barbiturate homolog 5,5-dimethylbarbituric acid on N-nitrosodiethylamine-initiated renal and hepatic carcinogenesis in the F344/NCr rat.

The oxazolidinedione anticonvulsant trimethadione (3,5,5-trimethyl-2,4-oxazolidinedione, TMO) as well as its major metabolite, dimethadione (5,5-dimethyl-2,4-oxazolidinedione, DMO), and a structural analog from the barbiturate series, 5,5-dimethylbarbituric acid (DMB), were fed to F344/NCr male rats previously given a single initiating injection of N-nitrosodiethylamine (NDEA). The known promoter, phenobarbital (5-ethyl-5-phenylbarbituric acid, PB), was employed in this study as a positive control. At dosage levels equimolar to 500 ppm PB, none of the three compounds promoted development of hepatocellular adenomas or carcinomas, in contrast to PB. The two oxazolidinedione analogs and DMB caused minimal or no induction of cytochrome P450 isozyme 2B1 (CYP2B1)-mediated alkoxyresorufin O-dealkylase activities following short-term (2 weeks) feeding to separate groups of 6-week-old male F344/NCr rats, in contrast to the dramatic induction caused by PB. Promotion of neither thyroid nor renal neoplasia was observed following prolonged feeding of any of the tested compounds, although a significantly higher frequency of premalignant renal cortical tubular lesions (dysplasias) was seen in rats exposed to TMO following NDEA initiation than in those treated with NDEA alone. These studies provide important additional data on structure/liver tumor promoting activity relationships, and yield further evidence that within this group of structurally related anticonvulsants, it is possible to separate anticonvulsant activity from tumor promoting activity in the rat liver.

The effect of administration time on malformations induced by three anticonvulsant agents in C57BL/6J mice with emphasis on forelimb ectrodactyly.

Exposure of C57BL/6J mice to three anticonvulsant derivatives, namely, dimethadione, sodium valproate, and sodium diphenylhydantoin, each induced postaxial forelimb ectrodactyly. The agents were administered at gestational days 9, 9 1/3, 9 2/3, and 10. It was determined that administration at day 9 2/3 induced the highest percentage of forelimb ectrodactyly for each of the three agents. The forelimb ectrodactyly response in the C57BL/6J strain was compared with the A/J strain (Collins et al., Teratology, 41:61-70, 1990); it was found that the C57BL/6J strain was more sensitive to dimethadione and the A/J strain was more sensitive to diphenylhydantoin and sodium valproate. The position of vertebral defects induced by sodium valproate correlated with the time of drug administration. The overall syndrome of malformations induced by the three anticonvulsant agents was relatively similar in the two mouse strains and differed between each of the anticonvulsant agents.

Induction of postaxial forelimb ectrodactyly with anticonvulsant agents in A/J mice.

Exposure of A/J mice on day 9.5 of gestation to the derivatives of three acidic anticonvulsant agents, namely dimethadione, sodium valproate, and sodium diphenylhydantoin, each induced postaxial forelimb ectrodactyly predominantly of the right side. This specific malformation has previously been associated with the administration of acetazolamide to rodents; however, several agents can induce this same defect including other carbonic anhydrase inhibitors, carbon dioxide, cadmium, ethanol, ammonium chloride, and 13-cis retinoic acid. The relative potency of the three agents indicates no direct relationship to the pKa of the acid. Other than ectrodactyly, each of the anticonvulsant agents induced a compound-specific spectrum of malformations despite the uniform administration time. This finding suggests that these agents are capable of acting via different mechanisms or by the differential spatial and temporal dynamics of a common mechanism.

Inhibition of trimethadione and dimethadione teratogenicity by the cyclooxygenase inhibitor acetylsalicylic acid: a unifying hypothesis for the teratologic effects of hydantoin anticonvulsants and structurally related compounds.

Teratogenicity of the anticonvulsant phenytoin may be due in part to its bioactivation by prostaglandin synthetase, forming a reactive free radical intermediate. We examined whether teratogenicity of the structurally similar oxazolidinedione anticonvulsants, trimethadione and its N-demethylated metabolite dimethadione, could be inhibited by the prostaglandin synthetase inhibitor acetylsalicylic acid (ASA). Trimethadione, 700 or 1000 mg/kg intraperitoneally (ip), was given to pregnant CD-1 mice during (Gestational Days 12 and 13) or before (Days 11 and 12) the critical period of susceptibility to phenytoin-induced fetal cleft palates. Dimethadione was given similarly on Days 11 and 12, or 12 and 13, in a dose (900 mg/kg ip) that was equimolar to 1000 mg/kg of trimethadione. ASA, 10 or 1 mg/kg ip, was given 2 hr before trimethadione or dimethadione on Days 11 and 12, and before trimethadione on Day 11 only. Dams were killed on Day 19 and fetuses were examined for anomalies. Either dose of trimethadione given on Days 12 and 13 was negligibly teratogenic, as evidenced by a non-dose-related, 1.1% mean incidence of fetal cleft palates. However, when given earlier on Days 11 and 12, trimethadione 1000 mg/kg caused an 8.9% incidence of cleft palates (p less than 0.05). Similarly, dimethadione caused a 3.9-fold higher incidence of cleft palates when given earlier on Days 11 and 12 (17.3-34.9%) than on Days 12 and 13 (4.4%) (p less than 0.05). At equimolar doses, dimethadione caused a 1.9- to 3.9-fold higher incidence of cleft palates compared to trimethadione (p less than 0.05), suggesting that dimethadione may be the proximate teratogen. Either dose of ASA given on both days before trimethadione totally prevented cleft palates, and ASA 10 mg/kg given only on Day 11 reduced the incidence of trimethadione-induced cleft palates to 1.1% (p less than 0.05). ASA reduced the incidence of cleft palates caused by dimethadione given on Days 11 and 12 from 34.9 to 20.3% (p less than 0.05). These results suggest that the teratogenic potential of trimethadione may depend at least in part upon its prior N-demethylation to dimethadione, which then can be bioactivated by prostaglandin synthetase to a teratogenic reactive intermediate, possibly involving a free radical located in the oxazolidinedione ring. This would provide a unifying hypothesis for the teratogenicity of hydantoins, as well as structurally related teratogens like trimethadione, which lack the molecular configuration necessary for the formation of a teratogenic arene oxide intermediate.