Abstract for RACB98001

Developmental Toxicity Evaluation of 3'-Azido-3'-deoxythymidine (AZT) and Dapsone Administered by Gavage to Swiss Albino (CD-1®) Mice on Gestational Days 6 Through 15


  • 3'-Azido-3'-deoxythymidine (AZT) (CASRN 30516-87-1)
  • Dapsone (CASRN 80-08-0)

Report Date: September 2004


The following abstract presents results of a study conducted by a contract laboratory for the National Toxicology Program. The findings may not have been peer reviewed and were not evaluated in accordance with the levels of evidence criteria established by NTP in March 2009. For more information, see the Explanation of Levels of Evidence for Developmental Toxicity. The findings and conclusions for this study should not be construed to represent the views of NTP or the U.S. Government.

In the present study, timed-mated Swiss albino (CD-1®) mice were dosed by gavage with AZT (0, 100, 200, or 400 mg/kg/day) and dapsone (0, 50, 100 or 200 mg/kg/day) alone or in combination on gestational days (gd) 6 through 15. The vehicle was 0.5% aqueous methylcellulose. Total daily doses were administered in two divided doses (morning and afternoon), each with a volume of 10 ml/kg body weight. The oral route corresponds to one of the commonly used routes in human patients. Dose selection was based on previous studies in which CD-1 mice were treated by gavage with AZT alone or dapsone alone or in combination from gd 6 through 15 (NTP, 1994; NTP, 2002).

Timed-mated mice (20-21 per group) were monitored at regular intervals for clinical signs of toxicity, feed consumption, and body weight. At necropsy (gd 17), the following observations were made: clinical condition; maternal body, and gravid uterine weights; pregnancy status; and number of corpora lutea. In the gravid uterus, the numbers of prenatal deaths (resorptions and/or dead fetuses) and live fetuses were recorded. Live fetuses were weighed, sexed, and examined for morphological anomalies (external, visceral, and skeletal). The study was performed in two replicates. Thus, the experimental design was a three-way matrix: AZT (4 dose levels) x DAP (4 dose levels) x REP (2 replicates).

For AZT alone, there were no maternal deaths or remarkable clinical signs. At scheduled necropsy (gd 17), gross findings included enlarged maternal spleens in 1, 3, 8, and 15 dams in the control through high dose, respectively. No significant differences or significant linear trends for maternal body weight, weight change or gravid uterine weight were noted. Average daily feed consumption was also unaffected. Prenatal mortality (resorptions and/or late fetal deaths) was equivalent among groups. However, fetal body weight (males, females, or both sexes) showed dose-related reductions of 8-17% in all AZT dose groups. Dose-related fetal morphological variations were observed, most notably a dose-related increase in the percent fetuses with skeletal variations at 200 and 400 mg/kg/day. The higher incidence of extra ribs on Lumbar I contributed to this outcome.

For dapsone alone, maternal mortality (10%) was noted at the high dose, but otherwise clinical observations were unremarkable. At scheduled necropsy (gd 17), gross findings included enlarged spleens in 1, 12, 13 or 17 dams in the control through high dose groups, respectively. Maternal body weight, weight change, and gravid uterine weight were each reduced at the highest dose of dapsone (200 mg/kg/day). As noted during the pre-treatment period, there was a tendency for dapsone groups to consume more feed than the controls. Consistent with this observation, maternal feed consumption was usually at or above controls for subsequent measurement periods. The exception to this pattern was a significant reduction in maternal relative feed consumption at the highest dose of dapsone at the beginning of the treatment period (gd 6 to 9). Indices of prenatal mortality appeared to be elevated at the highest dose of dapsone (e.g., 16% resorbed implantation sites vs. 2% for controls). Some indices of prenatal mortality reached statistical significance and effects tended to be more severe in the second replicate. Fetal body weights (male, female or both) were significantly reduced by 25% at the highest dose of dapsone. Altered incidences of fetal morphological anomalies (malformations or variations) were noted, but patterns of effects varied across replicates.

For endpoints with a significant p value in the whole (16-group) model, each AZT/dapsone combination was compared to its constituent doses. Numerous differences were found between drug combination groups and their respective constituent doses, even in the absence of statistically significant drug interactions.

Maternal and developmental endpoints were analyzed for three-way drug-by-replicate (AZT*DAP*REP) interactions. When the threeway interaction term was significant, data were analyzed separately within each replicate. Three-way interactions were noted for maternal body weight change and indices of prenatal mortality, such that similar effects were noted in both replicates, but effects were more severe in Replicate II. For morphological endpoints, three-way interactions appeared due to spurious distribution of sparse data across the 32 cells in the design more so than to differences in severity between replicates.

Maternal and developmental endpoints were analyzed for two-way global drug (AZT*DAP) interactions. When a significant global drug interaction was detected, each of the nine groups receiving both AZT and dapsone were analyzed individually for pairwise drug interactions between their constituent doses. Specific dose combinations yielding non-additive responses were identified, and these drug combinations and their constituent doses were compared to the vehicle control group. Based on the outcome of these analyses, each pairwise drug interaction was classified as synergism, potentiation or antagonism. In a limited number of cases, pairwise drug interactions were unclassified because neither the drug combination, nor the constituent doses differed significantly from vehicle control.

The two-way drug interaction (AZT*DAP) was significant for maternal body weight change during early treatment (gd 6 to 9). For this endpoint, antagonism was noted at AZT 200/ DAP 200 and AZT 400/DAP 200 such that each group was reduced below controls, but the magnitude of effect was not as great as predicted by the additive model. The two-way interaction (AZT*DAP) was also significant for maternal body weight change at the end of gestation (gd 15 to 17) and for gravid uterine weight. For these related endpoints, potentiation was noted at AZT 200/DAP 200, AZT 400/DAP 100, and AZT 400/DAP 200 such that each group was reduced below controls and the effect was greater than predicted by the additive model. Two-way interactions were marginally significant for maternal relative feed consumption (gd 12 to 15 and gd 0 to 17), but pairwise drug interactions for specific dose combinations were not significant (gd 12 to 15) or produced antagonism (gd 0 to 17) resulting in no change relative to the control group (AZT 200/DAP 100).

Adverse drug interactions resulted in increased prenatal mortality beyond that predicted by the additive model. Specifically the two-way drug interactions (AZT*DAP) were significant for percent of litters with dead fetuses, percent litters with full litter loss (100% nonlive), percent resorbed implantation sites, percent nonlive implantation sites, percent adversely affected implantation sites, and number of live fetuses per litter. Potentiation was noted at AZT 200/ DAP 200, AZT 400/DAP 100, and AZT 400/ DAP 200 such that prenatal mortality was greater than predicted by the additive model, while the number of live fetuses per litter was lower than predicted by the additive model. To illustrate the magnitude of this effect, the incidence of resorbed implantation sites was 2.5% for controls, 7.8% for AZT 400, and 16% for DAP 200 as compared to 76% for the AZT 400/DAP 200 combination. Similarly, there were no dams with full litter loss in the control, AZT 400 or DAP 200 groups, as contrasted with 41% of dams with full litter loss at AZT 400/200 DAP.

Significant two-way drug interactions (AZT*DAP) were noted for incidences of fetuses or litters with external or skeletal (but not visceral) malformations. Malformations found in this study included anasarca (generalized edema), exencephaly (brain lying outside the skull), cleft palate, unossified skull, unossified nasals and supraoccipital regions, split cartilage of sternum, cleft sternum, hole in cartilage of sternum, hole in ossified portion of sternebrae, agenesis of rib, and fused ribs. For malformation incidences, significant two-way drug interaction terms were not accompanied by clear-cut replicable drug interactions at specific dose combinations.

In summary, co-administration of AZT and dapsone to Swiss albino mice during major organogenesis resulted in significant adverse drug interactions. Drug interactions were highly significant for indicators of fetal viability, including gravid uterine weight and maternal weight change during late gestation. Potentiation of adverse developmental effects was found when AZT (400 mg/kg/day) was combined with dapsone (100 or 200 mg/kg/day), and when 200 mg/kg/day AZT was combined with 200 mg/kg/day of dapsone. Lower dose combinations (AZT 100 mg/kg/day combined with dapsone 50, 100 or 200 mg/kg/day) were not associated with statistically significant nonadditive drug interactions, but did occasionally result in effects that exceeded those of AZT alone and/or dapsone alone. These differences were considered biologically relevant in the evaluation of dose-response relationships for AZT/dapsone combinations, especially when they involved endpoints for which significant drug interactions occurred at higher dose combinations. Thus, almost all AZT/dapsone combinations showed some evidence for enhancement of developmental toxicity in the CD-1 mouse.