LEVODOPA

Fetal Risk Summary

Levodopa, a metabolic precursor to dopamine, is primarily used for the treatment and prevention of symptoms related to Parkinson's disease. The active agent for this purpose is thought to be dopamine, which is formed in the brain after metabolism of levodopa. Levodopa crosses the blood-brain barrier, but dopamine does not. Combination with carbidopa (see also Carbidopa), an agent that inhibits the decarboxylation of extracerebral levodopa, allows for lower doses of levodopa, fewer adverse drug effects related to peripheral dopamine, and higher amounts of levodopa available for passage to the brain and eventual conversion to dopamine.

Levodopa, either alone or in combination with carbidopa, has caused visceral and skeletal malformations in rabbits (1). Doses of 125 or 250 mg/kg/day in pregnant rabbits produced malformations of the fetal circulatory system (2). This teratogenicity was not observed at 75 mg/kg/day, but all three doses produced fetal toxicity manifested by decreased litter weight and an increased incidence of stunted and resorbed fetuses. In mice, no teratogenicity was observed with doses of 125, 250, and 500 mg/kg/day, but at the highest dose, fetuses were significantly smaller than controls (2). A similar, significant decrease in the weights of newborn mice was observed in a study in which pregnant mice were fed levodopa 40 mg/g of food, but not at lower doses (3). The number of pregnancies and the number of newborns were also significantly decreased in those treated at 40 mg/g compared with those treated with 0, 10, or 20 mg/g of food.

Levodopa, in oral doses of 1–1000 mg/kg/day administered to pregnant rats during the 1st week of gestation, produced a dose-related occurrence in brown fat (interscapular brown adipose tissue) hemorrhage and vasodilation in the newborns (4). A similar response was observed with dopamine. The addition of carbidopa (MK-486) to levodopa resulted in a significant decrease in this toxicity, implying that the causative agent was dopamine.

A study published in 1978 examined the effect of carbidopa (20 mg/kg SC every 12 hours × 7 days) and levodopa plus carbidopa (200/20 mg/kg SC every 12 hours × 7 days) on the length of gestation in pregnant rats (5). Only the combination had a statistically significant effect on pregnancy duration, causing a delay in parturition of 12 hours. The results were thought to be consistent with dopamine inhibition of oxytocin release.

Because Parkinson's disease is relatively uncommon in women of childbearing age, only a few reports, some involving uses other than for parkinsonism, have been located that describe the use of levodopa, with or without carbidopa, in human pregnancy (6,7,8,9,10,11,12,13,14,15,16,17,18,19,20 and 21).

Placental transfer of levodopa at term was documented in a 1989 report (6). A 34-year-old multiparous woman, with a 6-year history of Parkinson's disease, was treated with a proprietary combination of levodopa and benserazide (Madopar, not available in the United States), three 250-mg tablets/day. When pregnancy was diagnosed (15th day postconception), the combination therapy was discontinued and treatment with levodopa (up to 5000 mg/day) alone was started. After 5 months, because of worsening parkinsonism, she was again treated with levodopa/benserazide (six 250-mg tablets/day). She eventually gave birth to a normal, 3070-g female infant who had no signs or symptoms of toxicity from the drug therapy during the first 8 days of life. Her 1-minute Apgar score was 2; then improved to 7 and 10 at 3 and 5 minutes, respectively. The cord plasma level of levodopa was 0.38 µg/mL compared with 2.7 µg/mL in the mother's plasma, a ratio of 0.14.

A 1995 Reference described the placental transfer of levodopa and the possible fetal metabolism of the drug to dopamine, its active metabolite (7). A 34-year-old woman with juvenile Parkinson's disease was treated with carbidopa/levodopa (200/800 mg/day) during two pregnancies (see also Carbidopa). Both pregnancies were electively terminated, one at 8 weeks' gestation and the other at 10 weeks' gestation. Mean concentrations of levodopa (expressed as ng/mg protein) in the maternal serum, placental tissue (including umbilical cord), fetal peripheral organs (heart, kidney, muscle), and fetal neural tissue (brain and spinal cord) were 8.5, 33.6, 7.4, and 7.7, respectively. Corresponding mean concentrations of dopamine at these sites were 0.10, <0.03, 0.29, and 1.01, respectively. The levels in the placentas and fetuses for both levodopa and dopamine were much higher than those measured in control tissue. Moreover, the relatively high concentrations of dopamine in fetal peripheral organs and neural tissue implied that the fetuses had metabolized levodopa. The investigators cautioned that, because neurotransmitters were known to alter early neural development in animals and in cultured cells, the increased amounts of dopamine found in their study suggested that chronic use of levodopa during gestation could induce long-term damage (7).

The pregnancy outcomes of two women who were treated with levodopa or carbidopa/levodopa during three pregnancies were described in a 1985 paper (8). The first woman, with at least a 7-year history of parkinsonism, conceived while being treated with carbidopa/levodopa (five 25/250-mg tablets/day) and amantadine (100 mg twice daily). She had delivered a normal male infant approximately 6 years earlier, but no medical treatment had been given during that pregnancy. Amantadine was immediately discontinued when the current pregnancy was diagnosed. Other than slight vaginal bleeding in the 1st trimester, there were no maternal or fetal complications. She gave birth to a normal term infant (sex and weight not specified) who was doing well at 1.5 years of age. The second patient, a 32-year-old woman with parkinsonism first diagnosed at age 23, was being treated with levodopa (4 g/day) when a pregnancy of about 6 months' duration was diagnosed. Attempts to lower her levodopa dose were unsuccessful. She delivered a term, male, 7-lb 2-oz (about 3235-g) infant. Two years later, while still undergoing treatment with levodopa, she delivered a term, female, 6-lb 8-oz (about 2951-g) infant. Both children were alive and well at 7 and 5 years, respectively.

A 1987 retrospective report described the use of carbidopa/levodopa, starting before conception, in five women during seven pregnancies, one of which was electively terminated during the 1st trimester (9). A sixth woman, taking levodopa plus amantadine, had a miscarriage at 4 months. All of the other pregnancies went to term (newborn weights and sexes not specified). Maternal complications in three pregnancies included slight 1st trimester vaginal bleeding, nausea and vomiting during the 8th and 9th months (the only patient who reported nausea and vomiting after the 1st trimester) and depression that resolved postpartum, and preeclampsia. One infant, whose mother took amantadine and carbidopa/levodopa and whose pregnancy was complicated by preeclampsia, had an inguinal hernia. No adverse effects or congenital anomalies were noted in the other five newborns and all remained healthy at follow-up (approximately 1–5 years of age).

A 27-year-old woman, with a history of chemotherapy and radiotherapy for non-Hodgkin's lymphoma occurring approximately 4 years earlier had developed a progressive parkinsonism syndrome that was treated with a proprietary preparation of carbidopa/levodopa (co-careldopa, Sinemet Plus; 375 mg/day) (10). She conceived 5 months after treatment began and eventually delivered a healthy, 3540-g male infant at term. Apgar scores were both 9 at 1 and 10 minutes. Co-careldopa had been continued throughout her pregnancy.

A brief case report, published in 1997, described a normal outcome in the 3rd pregnancy of a woman with levodopa-responsive dystonia (Segawa's type) who was treated throughout gestation with 500 mg/day of levodopa alone (11). The male infant weighed 2350 g at birth and was developing normally at the time of the report. Two previous pregnancies had occurred while the woman was being treated with daily doses of levodopa 100 mg and carbidopa 10 mg. Spontaneous abortions had occurred in both pregnancies; one at 6 weeks and the other at 12 weeks. An investigation failed to find any cause for the miscarriages.

Two 1998 case reports described the use of levodopa, in combination with selegiline and benserazide, during human pregnancy (see Selegiline for details) (12,13). In another 1998 case report, a woman took levodopa/carbidopa (Sinemet CR 50/200) four times daily throughout gestation (14). Healthy male infants were delivered in all three pregnancies.

A brief 1996 communication included data obtained from a manufacturer (the Roche Drug Safety database) on the effects of levodopa on the fetus during human pregnancy (15). This database, current up to March 31, 1995, had six reports involving the use of levodopa/benserazide during gestation. The outcomes of these pregnancies included two elective terminations, one spontaneous abortion, two normal outcomes, and one lost to follow-up. No information was available on the condition of the fetuses in the three abortions. Another normal pregnancy outcome, without further details, from a mother who used levodopa/benserazide during gestation, was also cited (16). To the knowledge of these authors, two of whom were representatives of the manufacturer, no reports of human birth defects resulting from use of levodopa had been discovered.

A study designed to investigate the inhibitory effect of levodopa on prolactin levels in late pregnancy was reported in 1973 (17). Four women in the 3rd trimester of pregnancy were given a single oral dose of levodopa, 1000 mg. A statistically significant decrease in serum prolactin occurred, with the lowest levels measured 4 hours after the dose, but 2 hours later the levels had returned to pretreatment values, similar to nontreated controls. No further decreases in prolactin were measured during the next 3 days. All of the treated women eventually delivered normal infants at or near term.

Levodopa has been used in the treatment of coma resulting from fulminant hepatic failure occurring during pregnancy (18,19). A 22-year-old woman, in her 6th month of pregnancy, had severe viral hepatitis with a grade IV coma (18). She was successfully treated with levodopa, 1 g orally every 6 hours, neomycin, vitamins, and parenteral hydrocortisone (1000 mg/day). Recovery from the coma occurred 24–48 hours after the start of levodopa. Approximately 3 months later, she delivered a full-term infant (details of the infant and its condition were not included). In the second case, a 30-year-old woman in her 5th month of gestation was treated for 1 week with levodopa, 1 g orally every 6 hours (19). This patient, who also had viral hepatitis, began to recover consciousness 1 hour after the first dose of levodopa and was fully awake within 24–48 hours. She delivered a term, healthy infant (sex and weight not specified).

Angiotensin pressor responsiveness (vascular sensitivity) was decreased by the administration of levodopa in pregnancy (mean 29–30 weeks, range 24–36 weeks' gestation) (20). Five patients were treated with a single 500 mg oral dose and 16 received 1000 mg. Only the higher dose produced a significant decrease in angiotensin sensitivity. The authors hypothesized that chronic treatment with levodopa might not only decrease angiotensin sensitivity but also decrease pregnancy-induced hypertension. No data were provided on the outcome of these pregnancies following the study. The effect of levodopa on plasma prolactin secretion and inhibition of the renin-aldosterone axis was reported in a subsequent paper by this same group of investigators (21). Using methods and a patient population nearly identical with those of their previous study, the investigators found that a 1000-mg oral dose of levodopa produced a significant decrease in serum prolactin, plasma renin activity, and plasm aldosterone.

In summary, levodopa-induced teratogenicity and dose-related toxicity have been observed in animals, but no adverse outcomes associated with this drug have been observed in a limited number of human pregnancies. Although conditions requiring the use of levodopa during the childbearing years are relatively uncommon, exposure to this agent during gestation does not appear to present a major risk to the fetus. However, evaluation of the effect of chronic in utero exposure to dopamine, the active metabolite of levodopa, on neurodevelopment is warranted.

Breast Feeding Summary

Levodopa is excreted into breast milk. A 1998 case report described a woman who took levodopa/carbidopa (Sinemet CR 50/200) four times daily throughout gestation (14). She continued the therapy after delivery and during breast feeding. At about 5 months postpartum, the peak concentration of levodopa in milk (based on 5 mL milk samples) was 1.6 nmol/mL, 3 hours after a dose. The milk:plasma ratio, based on area under the concentration curves (6 hours for plasma and milk) was 0.28. When the dose was changed to two tablets of Sinemet 25/100 four times daily, the peak milk concentration (3 hours after a dose) was 3.47 nmol/mL and the milk:plasma ratio was 0.32. With both tablet strengths, the milk concentrations returned to baseline in 6 hours. The authors estimated that the nursing infant was ingesting 0.127–0.181 mg/day (0.016–0.023 mg/kg/day) of levodopa. No adverse effects were observed in the infant and he was developing normally at 2 years of age (14).

Levodopa inhibits prolactin release in both animals (22) and humans (17,21). In lactating rats, levodopa injected IP in doses of 1.25, 2.5, 5, and 10 mg/100 g body weight produced a dose-related inhibition of milk ejection (22). Doses of 5 and 10 mg/100 g body weight, but not lower doses, prevented the release of prolactin induced by suckling. Small doses of oxytocin given immediately before suckling produced a normal milk-ejection response, indicating that the mechanism of inhibition by levodopa was not caused by mammary gland response but to an increase in catecholamines at the hypothalamic-hypophysial axis (22).

In one study, four women in the 3rd trimester of pregnancy received a single 1000-mg oral dose of levodopa (17). A significant decrease in serum prolactin occurred with the lowest concentration at 4 hours. Two hours later, the level had returned to pretreatment values. After delivery at or near term, all of the women had completely normal lactation. A second study also found that a single 1000-mg oral dose of levodopa produced a significant decrease in serum prolactin (21).

Clinical evidence of lactation inhibition was provided in a 1974 study (23). Ten female patients with a history of inappropriate galactorrhea were treated orally with levodopa, 500–3500 mg/day. Partial to complete suppression of lactation was observed in the women, but when treatment was stopped, lactation returned to normal levels.

References

1. Product information. Sinemet. DuPont Pharmaceuticals, 1997.
2. Staples RE, Mattis PA. Teratology of L-dopa (abstract). Teratology 1973;8:238.
3. Cotzia GC, Miller ST, Tang LC, Papavila PS. Levodopa, fertility, and longevity. Science 1977;196:549–50.
4. Kitchin KT, DiStefano V. L-Dopa and brown fat hemorrhage in the rat pup. Toxicol Appl Pharmacol 1976;38:251–63.
5. Seybold VS, Miller JW, Lewis PR. Investigation of a dopaminergic mechanism for regulating oxytocin release. J Pharmacol Exp Ther 1978;207:605–10.
6. Allain H, Bentue-Ferrer D, Milon D, Moran P, Jacquemard F, Defawe G. Pregnancy and parkinsonism. A case report without problem. Clin Neuropharmacol 1989;12:217–9.
7. Merchant CA, Cohen G, Mytilineou C, DiRocco A, Moros D, Molinari S, Yahr MD. Human transplacental transfer of carbidopa/levodopa. J Neural Transm Park Dis Dement Sect 1995;9:239–42.
8. Cook DG, Klawans HL. Levodopa during pregnancy. Clin Neuropharmacol 1985;8:93–5.
9. Golbe LI. Parkinson's disease and pregnancy. Neurology 1987;37:1245–9.
10. Ball MC, Sagar HJ. Levodopa in pregnancy. Mov Disord 1995;10:115.
11. Nomoto M, Kaseda S, Iwata S, Osame M, Fukuda T. Levodopa in pregnancy. Mov Disord 1997;12:261.
12. Hagell P, Odin P, Vinge E. Pregnancy in Parkinson's disease: a review of the literature and a case report. Mov Disord 1998;13:34–8.
13. Kupsch A, Oertel WH. Selegiline, pregnancy, and Parkinson's disease. Mov Disord 1998;13:175–94.
14. Thulin PC, Woodward WR, Carter JH, Nutt JG. Levodopa in human breast milk: clinical implications. Neurology 1998;50:1920–1.
15. von Graeventiz KS, Shulman LM, Revell SP. Levodopa in pregnancy. Mov Disord 1996;11:115–6.
16. Bauherz G. Pregnancy and Parkinson's disease. A case report. N Trends Clin Neuropharmacol 1994;8:142. As cited by von Graeventiz KS, Shulman LM, Revell SP. Levodopa in pregnancy. Mov Disord 1996;11:115–6.
17. Pujol-Amat P, Gamissans O, Calaf J, Benito E, Perez-Lopez FR, L'Hermite M, Robyn C. Influence of L-dopa on serum prolactin, human chorionic somatomammotrophin (HCS) and human chorionic gonadotrophin (HCG) during the last trimester of pregnancy. In: Human Prolactin. Proceedings of the International Symposium on Human Prolactin, Brussels, June 12–14, 1973:316–20.
18. Datta DV, Maheshwari YK, Aggarwal ML. Levodopa in fulminant hepatic failure: preliminary report. Am J Med Sci 1976;272:95–9.
19. Chajek T, Friedman G, Berry EM, Abramsky O. Treatment of acute hepatic encephalopathy with L-dopa. Postgrad Med J 1977;53:262–5.
20. Kaulhausen H, Oney T, Feldmann R, Leyendecker G. Decrease of vascular angiotensin sensitivity by L-dopa during human pregnancy. Am J Obstet Gynecol 1981;140:671–5.
21. Kaulhausen H, Oney T, Leyendecker G. Inhibition of the renin-aldosterone axis and of prolactin secretion during pregnancy by L-dopa. Br J Obstet Gynaecol 1982;89:483–88.
22. Prilusky J, Deis RP. Effect of L-dopa on milk ejection and prolactin release in lactating rats. J Endocrinol 1975;67:397–401.
23. Ayalon D, Peyser MR, Toaff R, Cordova T, Harell A, Franchimont P, Lindner HR. Effect of L-dopa on galactopoiesis and gonadotropin levels in the inappropriate lactation syndrome. Am J Obstet Gynecol 1974;44:159–70.