TERATOLOGY 42:131-136 (1990)

Developmental Toxicity of Orally Administered 2 ' ,3'- Dideoxycytidine in Mice PIA LINDSTROM, MARTHA HARRIS, ALAN M. HOBERMAN, JUNE K. DUNNICK, AND RICHARD E. MORRISSEY National Toxicology Program, National Institute of Enuironmental Health and Dynamac Corporation (P.L.), Research Sciences (M.H., J.K.D.) Triangle Park, North Carolina 27709; Argus Research Laboratories, Inc., Horsham, Pennsylvania 19044 (A.M.H.); Merck, Sharp & Dohme Research Laboratories, West Point, Pennsyluania 19486 (R.E.M.)

ABSTRACT

2' ,3'-Dideoxycytidine (DDC), a potent inhibitor of human immunodeficiency virus (HIV), is presently undergoing clinical trials a s a promising anti-AIDS drug. Since there are very limited published animal toxicity data available, and nucleoside analogues are being considered for treatment of HIV-infected pregnant women, a study was conducted in mice to investigate the potential adverse developmental effects of this drug. DDC, suspended in 0.5% methyl cellulose, was administered via gavage twice per day during gestation days (gd) 6 through 15 to C57BU6N mice in a total dose of 0, 200,400,1,000, or 2,000 mgtkgtday. Maternal weight gain during the gestation and treatment period, a s well as gravid uterine weight, decreased significantly in the 2,000 mg group, but weight gain, corrected for gravid uterine weight, was not affected by DDC. The percent resorptions per litter increased significantly in the highest dose group, and there were fewer live litters because of complete litter resorption in six dams. Among litters with live fetuses, the mean litter size was significantly reduced in the 2,000 mg group. Average fetal body weight per litter decreased significantly in the 1,000 and 2,000 mg groups. The number of fetuses with any malformation, the number of litters with one or more malformed fetuses and the percent of malformed fetuses per litter increased significantly in the 1,000 and 2,000 mg groups. There was a n increase in malformations at 400 mg/kg/day; however, i t was not statistically significant. In conclusion, DDC produced developmental toxicity (malformations, reduced fetal body weight, and resorptions) in the absence of overt maternal toxicity except for body weight changes due to resorptions and reduced fetal weights.

An increasing number of women of childbearing age infected with the human immunodeficiency virus (HIV-1) are being detected. Approximately 80% of AIDS in children is believed to be acquired through prenatal transmission (Katz and Wilfert, '89; Blanche et al., '89). HIV-1 has been identified in 15-20 week old fetuses (Jovaisas et al., '85; Sprecher et al., '86) and thus showing its ability to be transmitted across the placenta. 2',3'-Dideoxycytidine, one of the most potent of the HIV inhibitors in the 2',3'-dideoxynucleoside class (Mitsuya and Broder, '86; Mitsuya et al., '87) and therefore a promising anti-AIDS drug, is currently being evaluated in clinical tri0 1990 WILEY-LISS. INC.

als as a single drug and in combination with 3'-azido-3'-deoxythymidine(AZT) (Yarchoa n et al., '88; Merigan et al., '89). Limited animal toxicity data are available for DDC and there is a need for information regarding its potential to cause developmental toxicity since the use of nucleoside analogs for treatment of AIDS in pregnant women is now being considered. DDC is believed to inhibit the replication of HJV-1 by selectively terminating the DNA chain. (For a review on the proposed mechanism of action of DDC, see Jeffries, Received December 8, 1989; accepted February 7, 1990.

132

P. LINDSTROM ET AL.

'89). In vitro DDC provides almost complete protection against the infectivity and cytopathic effect of HIV a t concentrations as low as 0.5-1.0 pM, doses which are a t least a 10-fold less than those preventing cell growth (Mitsuya and Broder, '86). In the ongoing clinical trials with DDC as a treatment for AIDS, daily doses up to 10 mg/kg have been administered (NIAID Clinical Trails, '89). Lower doses are, however, more common in current clinical trials. DDC is given every 4 to 8 h r (depending upon the dose) because i t has a short halflife of approximately 1 hr. Toxic effects in humans a t therapeutic doses include cutaneous eruptions, fever, mouth sores, thrombocytopenia, and neutropenia (Yarchoan et al., '88). In addition, some patients have experienced peripheral neuropathy after 6-14 weeks of treatment; however, this is reversible when treatment stops. No data are currently available in the literature regarding toxic side effects of DDC in pregnant women. MATERIALS AND METHODS

Animals and animal husbandry C57B116N mice were obtained from the Frederick Cancer Research Center (Frederick, MD). Female mice, 6-8 weeks old upon arrival, were allowed a 2-week acclimation period before mating. They were individually identified by ear tags. On gd 0, the body weights were between 18 and 25 g. Standard NIH-31 rodent chow and deionized water were available ad libitum throughout the study. Cages, food, and water were changed twice per week. Animal rooms were equipped with automatic light cycle timers (lights on 8 AM to 8 PM). Relative humidity was 50% (range 40-60) and the temperature was 22°C (range 20.5O-23.5"C). Females were paired with proven breeder males (from the same strain) 2:l from the NIEHS colony for up to 4 days and examined each morning for the presence of copulatory plugs. Plug-positive mice were grouphoused (maximum of 5 per cage) and assigned to dose groups using stratified randomization based on body weights on gd 5. Test chemical and treatment 2',3'-Dideoxycytidine (> 99% pure) was synthesized by Hoffman La Roche, Nutley, NJ, and supplied by the National Cancer Institute, Bethesda, MD. Dosing solutions were prepared by Radian Corporation (Mor-

risville, NC) as a n aqueous suspension in 0.5% methyl cellulose. Aliquots of each preparation of DDC were analyzed by a liquid chromatographic method prior to administration to verify the concentration, and following the dosing period to confirm its stability. Mice were treated with DDC from gd 6 through 15 by oral intubation. Earlier studies conducted by NIEHS and NCI have shown that DDC is not palatable when given in the feed or drinking water. The doses were selected based on observed LD,, levels of greater than 4,000 mg/kg (Dr. Grieshaber, NCI, personal communication), and a recent study by Luster et al. ('89), which showed t h a t 600 mg/kg/day of DDC, given a s a single daily dose by gavage for 22 days over a 30-day period in 6-8 week old female mice, did not alter the survival rate or body weight. In this study each dose was divided in half and, because of a short halflife, administered twice daily at 9 AM and 3 PM at total doses of 0, 200, 400, 1,000, and 2,000 mg/kg/day. The dosage volume was 10 ml/kg of body weight and the administered volume was adjusted daily according to the body weight.

0bservations Mice were weighed on gd 0, 5, 6-15, and 18. Dams were observed twice daily during the treatment period for clinical signs of toxicity. On gd 18 they were weighed and decapitated. Liver weight, gravid uterine weight, and uterine contents (i.e., number of implantation sites, resorptions, dead fetuses, and live fetuses) were recorded. The uteri of dams with no apparent implants were evaluated using a press plate to visualize implantation sites that may have undergone very early resorption. Live fetuses were removed from the uterus, weighed, and examined for gross morphological abnormalities. Half of the live fetuses were preserved in Bouin's fluid, then sexed and examined for visceral malformations using a variation of Wilson's technique (Wilson, '65). The other half was preserved in ethanol, processed with Alizarin Red S stain, and then examined for skeletal malformations (Staples and Schnell, '63). Statistical analysis Maternal and selected fetal toxicity effects were examined using Analysis of Variance (Snedecor and Cochran, '67) to determine a n overall significant dose effect. If a

DEVELOPMENTAL TOXICITY OF DIDEOXYCYTIDINE

133

TABLE 1. Maternal toxicity in C57B116N mice treated with 2',3'-dideoxycytidtne by Gavage on days 6 through 15 of gestation 2',3'-dideoxycytidine mgikgiday 0 200 400 1,000 2,000 Subjects (Dams) 44 43 Total treated 43 43 43 2l 21.2 31.2 31 7l Removed Non-pregnant 8 13 15 14 16 No. (%) pregnant at sacrifice 26 (63) 33 (80) 28 (68) 26 (65) 20 (56) Maternal weight gain (g) Gestation (gd 0-18) 14.9 t 1.0 16.1 t 0.7 15.1 ? 0.4 16.5 ? 0.6 7.4 ? 0B7.* Treatment (gd 6-15) 8.7 f 0.44 7.9 f 0.65 7.5 2 0.2 8.9 i 0.3 3.9 t 0.4* Corrected3 4.2 t 0.3 3.9 2 0.2 4.1 t 0.36 4.3 ? 0.2 4.7 t 0.67 Gravid uterine wt. (g) 11.9 2 0.6 12.6 L 0.4 11.5 2 0.76 10.8 L 0.3 3.0 2 0.5* Maternal liver weight 1.83 t 0.03 1.88 ? 0.04 1.83 ir 0.04 1.94 ? 0.03 1.77 2 0.06 Absolute (g) 6.88 t 0.09 7.17 L 0.07 6.67 t 0.21 7.10 -c 0.13 6.78 5 0.11 Relative (% body wt) 'Died during treatment. 'One animal removed due to littering. %orrected maternal weight gain = gd 18 wt - gd 0 wt - gravid uterine wt 4N = 34,animal that littered early on gd 18 is included. 5N = 27, animal that littered early on gd 18 is included. 6N = 25, one animal littered early. 7N = 19, missing value for gd 18 hw. *Pairwise comparisons made using Dunnett's test (two-sided, alpha = 0.05).

significant ANOVA result was noted, Dun- weight) maternal liver weight were not afnett's test (Dunnett, '55) was employed to fected by treatment (Table 1). No differences were observed among determine which dose groups differed significantly from the control group. The Coch- groups with respect to the number of imran-Armitage test for linear trend (Armit- plantation sites per dam (Table 2). Percent age, '55) was employed to assess any dose- resorptions per litter increased significantly related effects of DDC on resorptions and in the 2,000 mg group while percent litters malformations. The incidence of individual with resorptions increased significantly in malformations in treatment groups was the 1,000 and 2,000 mg groups. In the 2,000 compared to the control group using Fish- mg group there were also significantly fewer live litters because of complete litter er's exact test (Fleiss, '81) (one sided). resorption in six dams, and the number of RESULTS live fetuses per litter decreased signifiFifteen animals died during the treat- cantly. The average fetal body weight per ment period (Table 1). These animals were litter significantly decreased in the 1,000 found in all dose groups and four were con- and 2,000 mg groups (Table 2). The percent firmed as gavage error as evidenced by a of litters with malformed fetuses and perpunctured esophagus (two in the 1,000 mg cent of malformed fetuses per litter signifigroup and two in the 2,000 mg group). The cantly increased in the two highest dose animals in the two highest dose groups re- groups (Table 2). Among total number of litters with any sisted dosing. Other clinical signs of toxicity, which occurred among very few ani- defect, a significant trend was observed mals, consisted of hair loss and rough coat with increasing dose of DDC (Table 3). Upon and were found in all treatment groups comparison with controls, this parameter increased significantly in the 1,000 and (data not shown). Maternal body weight gain (gestation and 2,000 mg groups. A variety of malformatreatment period) was significantly de- tions (Table 3) were significantly increased creased in the 2,000 mg treatment group at the highest dose level, and micrognathia (Table 1).Weight gain, when corrected for and bent tibia, fibula, andlor femur were uterine weight, did not differ between treat- also significantly increased in the 1,000 and ment groups. Gravid uterine weight in the 2,000 mg groups. No statistically significant highest dose group was significantly less increase in malformations was observed at than in the control group (Table 1).Abso- either 200 or 400 mgikglday. For several lute, as well as relative (percent of body malformations, however, a significant trend

134

P. LINDSTROM ET AL.

TABLE 2 . Developmental Toxicity in C57B116N mice treated with 2',3'-dideoxycytidine by Gavage on days 6 through 15 of gestation 2',3'-dideoxycytidine mgkglday All litters (No.) No. implantation sitesidam % Resorptionsilitter 7c Litters with resorptions % (No.) litters totally resorbed No. live litters No. live fetusesilitter Average fetal body weightilitter (g) 5% Litters with malformed fetuses % Fetuses malformed per litter

0

200

400

1,000

2,000

33 8.9 ? 0.4' 7.3 ? 1.6' 48 0 (0) 33 8.3 f- 0.4' 1.07 2 0.02l 0

28 9.5 t 0.3 10.8 -t 3.6 57 4 (1) 27 9.0 5 0.2 1.10 2 0.01

26 8.6 f 0.5 10.3 ? 4.0

26 9.6 t 0.3 9.9 t 1.6 69* 0 (0)

20 9.3 t 0.5 73.5 c 6.7* 95* 30 (6)

4 0.4

12

0

42

4 (1) 25 8.0 ? 0.5 1.10 ? 0.02 2.1

26

8.7 t 0.3 0.89? 0.02*

loo*

53.3*

14

3.1 c 0.6" 0.66 ? 0.04* 93" 81.6*

'Mean 2 S.E.M. *Pairwise comparisons made using Fisher's Exact Test ( P < ,051.

was observed with increasing dose of DDC and the malformations first occurring in the 400 mglkgiday group. In addition, several fetal variations (particularly delayed ossification of the forepaw and hindpaw phalanges) were increased significantly in the higher dose groups (data not shown). At 1,000 and 2,000 mg of DDC, variations such as moderate dilation of the third ventricle of the brain, skeletal variations of the frontal bones, and delayed ossification of the caudal vertebrae and tarsal bones were increased significantly. At 2,000 mgikg of DDC, there were significant increases in incompletely ossified ribs and delayed ossification of the cervical vertebrae and metacarpal bones. DISCUSSION

In the present study DDC produced significant developmental toxicity in the absence of maternal toxicity when administered via gavage during major organogenesis. A significant decrease in maternal weight gain occurred during the treatment period and during gestation in the 2,000 mgl kg group and was due to the increased number of resorptions and reduced fetal weights. Weight gain adjusted for group differences in gravid uterine weights were comparable to that of the control group. Nonpregnant animals did not show a treatment-related weight loss during the treatment period (data not shown). These results are in agreement with those reported by Luster et al. ('891, who showed that 600 mglkglday of DDC, given by gavage once a day for 22 days to C57B116 female mice, did not alter the mortality rate or body weights. Exposure of the dams to DDC caused dose-dependent increases in the incidences

of a variety of gross external, soft tissue, and skeletal malformations. At 200 and 400 mglkglday there were no statistically significant increases in the incidence of malformations or variations. However a significant trend was observed for micrognathia, open eye lids, cleft palate, and bent tibia, fibula, andlor femur with increased concentration of DDC. These malformations first occurred in the 400 mglkglday group. As the dose of DDC increased to 1,000 mglkglday, micrognathia and bent tibia, fibula, and/or femur were significantly increased above controls. In the 2,000 mglkglgroup additional defects occurred in more than 25% of the litters including open eye lids, edematous body, cleft palate, and tail, limb, and vertebral defects. The wide range of malformations observed in this study is indicative of a n agent that can affect systems developing over most of the organogenesis period in mice. Since antimetabolites such as DDC are structural analogues of naturally occurring substances which are critical components of normal development, it is not surprising that DDC produces developmental toxicity affecting the development of a wide spectrum of organ systems. The usefulness of these data in mice for predicting potential hazard to humans given DDC is, to some extent, a function of the comparability of the disposition of the drug in humans and mice. DDC is rapidly absorbed from the gastrointestinal tract and eliminated by the kidney mainly a s unchanged parent compound in humans (Yarchoan et al., '88) and mice (Kelley et al. '87). DDC has been measured in plasma, urine, and cerebrospinal fluid in mice (Kelley et al., '87). The drug is rapidly cleared from plasma in a biphasic manner

TABLE 3. Morphologic defects in C57Bl16N mouse fetuses following maternal exposure to 2',3'-dideoxycytidine on gestational days 6 through 15' 2 ' ,3' -dideoxycytidine (mglkglday) 0 200 400 1,000 2,000 274 24 1 199 44 Total No. fetuses examined2 225 14 33 27 25 26 Total No. litters examined3 External malformations 32 0 2 4 19 Total No. fetuses with defects 12*** 0 1 3 8** Total No. litters with defects* No. litters (fetuses) with 1 Exencephaly 1 4 (5)** 5 (8)** Micrognathia* 1 1 Agnathia 1 1 7 (18)*** Open eye lids* 1 1 Ears, low set Short snout 2 (3) 1 Pointed snout 4 (7)** Edematous body* 3 (3)** Short body* 9 (18)*** Kinked tail* 6 (11)** Short and/or thickened tail* 4 (9)** Front limb, short* 4 (4)** Front limb, 1 digit missing* 4 (9)** Front limb, clubbed paws* Front limb, 4 digits missing 1(2) Rear limb, short 2 (2) I Rear limb, rotated 1 Rear limb, 1 digit missing 3 (4)** Rear limb, clubbed paws* Visceral malformations 133 122 104 110 23 Total No. fetuses examined 14 26 Total No. litters examined 5 10 Total No. fetuses with defects 7*** 4** Total No. litters with defects* No. litters (fetuses) with 1 Exencephaly 1 3rd ventricle, marked dilation Microthalmia 2 (2) 2 (2) 1 Anopthalmia 1 Agenesis of tongue Small tongue 2 (3) 1 One nasal passage 5 (6)** Cleft palate* 2 (2) 1 Dilation of renal pelvis 1 Close set of kidneys Skeletal malformations 95 21 141 119 115 Total No. fetuses examined 10 33 27 25 Total No. litters examined 21 1 2 0 Total No. fetuses with defects lo*** 0 1 2 Total No. litters with defects* No. litters (fetuses) with 1 Small nasals 1 Short nasals 1 Incompletely ossified nasals 1 Unossified nasals 1 Fused tympanic rings 2 Tympanic ring not ossified 1 Short andlor fused mandibles l(2) 2 (2) 1 Unossified mandibles Incompletely ossified palate 1(2) 1 Unossified palate 1 Vertebrae, irregularly shaped cervical arch 1 Vertebrae, small cervical arch 1 5 (8)** Vertebrae, fused cervical arch* Bent humerus, radius and/or ulna* 9 (16)*** 2 (5) 1 Bent and small radius and ulna 1 Unossified radius and ulna 1 26 (114)*** Bent tibia, fibula and/or femur* 10 (20)*** 1 Small fibula 'Fetuses may have more than one malformation. 'Only live fetuses were examined for malformations. 31ncludes only litters with live fetuses. *Significant Trend. **Fisher's Exact, P < .05 vs. controls. ***Fisher's Exact. P < .05 vs. controls

136

P. LINDSTROM ET AL.

(terminal t112 in BDF, mice of 67 minutes) following an iv bolus dose of 325 mg/m2. Yarchoan et al. ('88) reported a n average half-life of 1.2 h r in humans and the cerebrospinal fluid samples contained 20% of the plasma concentration of DDC measured in patients who were given therapeutic doses of DDC up to 0.36 mglkglday. Additional data on distribution and elimination of DDC in mice and humans are needed before the significance of these findings can be extrapolated to pregnant women. In this study the conservative NOAEL for developmental toxicity in mice was 200 mglkglday. ACKNOWLEDGMENTS

The present studies were conducted a t the National Toxicology Program (NTP), National Institute of Environmental Health Sciences, Research Triangle Park, NC. The authors express their appreciation to the Comparative Medicine Branch and its personnel who contributed to the completion of these studies. The fetal skeletal and visceral evaluations were performed by Argus Research Laboratories, Inc., Horsham, PA, under a contract with NTP. The authors thank the personnel at Argus for their contributions. The authors are greatful to Dr. Joseph Haseman and Ms. Ann-Marie Clark for performing the statistical analysis of the data and to Drs. Barbara D. Abbott, Michael Luster, and Bernard A. Schwetz for their careful reading of the manuscript. LITERATURE CITED Armitage, P. (1955) Tests for linear trends in proportions and frequences. Biometrics, 11:375-385. Blanche, S., C. Rouzioux, M.L. Guihard Moscato, F. Veber, M.J. Mayaux, C. Jacomet, J. Tricoire, A. De Ville, M. Vial, G. Firtion, A. De Crepy, D. Douard, M. Robin, C. Courpotin, N. Ciraru-Vigneron, F. LeDeist, C. Griscelli, and the HIV Infection in Newborns French Collaborative Study Group (1989) A prospective study of infants to women seropositive for human immunodeficiency virus type 1. N. Engl. J. Med., 320: 1643-1648. Dunnett, C.W. (1955) A multiple comparison procedure for comparing several treatments with a control. J. Am. Stat. Assoc., 50:1096-1129. Fleiss, J.L. (1981) Statistical Methods for Rates and Proportions, 2nd edition. John Wiley and Sons, New York, pp. 25-29. Jeffries, D.J. (1989) The antiviral activity of dideoxycytidine. J. Antimicrob. Chemother., 23(Suppl. A): 2934. Jovaisas, E., M.A. Kock, A. Schfer, M. Stauber, and D.

Lowenthal (1985) LAViHTLV I11 in 20 week fetus. Lancet, 2t1129. Katz. S.L.. and C.M. Wilfert (1989) Human immunodeficiency virus infection of newborns. N. Engl. J. Med., 320t1687, 1688. Kelley, J.A., C.L. Litterst, J.S. Roth, D.T. Vistica, D.G. Poplack, D.A. Cooney, M. Nadkarni, G.M. Balis, S. Broder, and D.G. Johns (1987) The disposition and metabolism of 2',3'-dideoxycytidine, an in vitro inhibitor of human T-lymphotropic virus type 111 infectivity, in mice and monkeys. Drug Metab. Dispos., 15: 595 -601. Luster, M.I., D.R. Germolec, K.L. White, Jr., B.A. Fuchs, M.M. Fort, J.E. Tomaszewski, M. Thompson, P.C. Blair, J.A. McCay, A.E. Munson, and G.A. Rosenthal(1989) A comparison of three nucleotide analogs with anti-retroviral activity on immune and hematopoietic functions in mice: In vitro toxicity to precursor cells and microstromal environment. Toxicol. Appl. Pharmacol., 101:328-339. Merigan, T.C., G. Skowron, S.S. Bozzotte, D. Richman, R. Uttamchandani, M. Fischl, R. Schooley, M. Hirsch, W. Soo, C. Pettinelli, H. Schaumburg, and the ddC study group of the AIDS Clinical Trials Group (1989) Circulating p24 antigen levels and responses to dideoxycytidine in human immunodeficiency virus (HIV) infections. Ann. Intern. Med., IIUt189-194. Mitsuya, H., and S.Broder (1986) Inhibition of the in vitro infectivity and cytopathic effect of human T-lymphotropic virus type IIIilymphadenopathy-associated virus (HTLV-IIIiLAV) by 2',3'-dideoxynucleosides. Proc. Natl. Acad. Sci. U.S.A., 83t1911-1915. Mitsuya, H., R.F. Jarrett, M. Matsukura, F. Di Marzo Veronese, A.L. Devico, M.G. Sarngadharan, D.G. Johns, M.S. Reitz, and S.Broder (1987) Long-term inhibition of human T-lymphotropic virus type IIIi lymphadenopathy-associated virus (human immunodeficiency virus) DNA synthesis and RNA expression in T cells protected by 2',3'-dideoxynucleosides in vitro. Proc. Natl. Acad. Sci. U.S.A., 84:2033-2037. NIAID Clinical Trials (1989) Quoted from protocol numbers: ACTG 011, ACTG 012, ACTG 047, ACTG 050, ACTG 112 identified in a search of NIAID Clinical Trials Information Service Database, August 1989. Snedecor, G.W., and W.G. Cochran (1967) Analysis of variance. In: Statistical Methods, 6th Edition. Iowa State University Press, Ames, IA, pp. 258-275. Sprecher, S., G. Soumenkoff, F. Puissant, and D. Degueldre (1986) Vertical transmission of HIV in 15 week fetus. Lancet, 2:288. Staples, R.E., and V.L. Schnell (1963) Refinement in rapid clearing technique in the KOH-Alizarin R e d 3 method for fetal bone stain. M. Techno]., 39%-63. Wilson, J.G. (1965) Methods for administering agents and detecting malformations in experimental animals. In: Wilson, J.G. and J. Warkany, ed. Teratology, Principles, and Techniques. University of Chicago Press, Chicago, pp. 262-277. Yarchoan, R., R.V. Thomas, J.P. Allain, N. Mcatee, T. Dubinsky, H. Mitsuya, T.J. Lawley, B. Safai, C.E. Myers, C.F. Perno, R.W. Klecker, R.J. Wills, M.A. Fischl, M.C. McNeely, J.M. Pluda, J. Leuther, J.M. Collins, and S. Broder (1988) Phase I studies of 2',3'-dideoxycytidine in severe human immunodeficiency virus infection as a single agent and alternating with zidovudine iAZT). Lancet, 16t76-80.