Arch Gynecol Obstet DOI 10.1007/s00404-014-3210-z
Maternal-Fetal Medicine
Pregnancy outcome in women exposed to dopamine agonists during pregnancy: a pharmacoepidemiology study in EFEMERIS database Caroline Hurault‑Delarue · Jean‑Louis Montastruc · Anna‑Belle Beau · Isabelle Lacroix · Christine Damase‑Michel
Received: 9 July 2013 / Accepted: 4 March 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose The objective of this exposed–unexposed study was to evaluate potential effects of dopamine agonists during pregnancy. Methods Data from EFEMERIS, a cohort of 57,408 pregnant women living in South West France, were used to compare exposed and unexposed women. The exposed group included 183 women (0.3 %) who received at least one prescription for one dopamine agonist during pregnancy. These women were individually matched with two unexposed women from the cohort for age and the monthand-year of the start of pregnancy. Pregnancy losses, birth defects, preterm births, low birth weight and psychomotor development were studied. Results Bromocriptine was the most frequently prescribed dopamine agonist, followed by cabergoline and quinagolide. Most (75 %) of the dopamine agonists were prescribed at the beginning of pregnancy (first trimester). There was no difference between the two groups concerning pregnancy history and demographic data. After adjustment for potential confounders, prescription and
Electronic supplementary material The online version of this article (doi:10.1007/s00404-014-3210-z) contains supplementary material, which is available to authorized users. C. Hurault‑Delarue (*) · J.-L. Montastruc · A.-B. Beau · I. Lacroix · C. Damase‑Michel Pharmacologie Médicale et Clinique, Centre Midi‑Pyrénées de PharmacoVigilance, de Pharmacoépidémiologie et d’Informations sur le Médicament, INSERM UMR 1027, Faculté de Médecine, Université Toulouse 3, Centre Hospitalier Universitaire, 37 allées Jules Guesde, 31000 Toulouse, France e-mail: caroline.delarue‑hurault@univ‑tlse3.fr
dispensation of dopamine agonists was associated with an increased risk of pregnancy loss [PORa = 3.7; 95 % confidence interval (CI) 1.8–7.4] and preterm birth (PORa = 3.6; 95 % CI 1.5–8.3). The prevalence of birth defects and low birth weight was not significantly different between the two groups. No difference in psychomotor development at either 9 or 24 months was observed between the two groups. Conclusion This study suggests that prenatal exposure to dopamine agonists may be associated with an increased risk of pregnancy loss and preterm birth. Keywords Dopamine agonists · Pregnancy · Outcome · Psychomotor development
Introduction Pregnant women may suffer from diseases requiring dopamine agonists: hyperprolactinemia (linked to a prolactinoma in most of cases) or more rarely idiopathic Parkinson’s disease. Prevalence of dopamine agonist-treated hyperprolactinemia in women is estimated between 0.1 and 1.2 %, and is most frequent among those between 25 and 34 years old [1]. Incidence of Parkinson’s disease among women of childbearing age is extremely low [2]. As a consequence, very little is known about the effects of dopaminergic agonists on embryo–fetal development. Dopaminergic neurons appear early during development, at the gestational age of 6–8 weeks in humans [3] and several studies suggest that dopamine is important for development, particularly in motor and cognitive programs [4]. The aim of this exposed–unexposed study was to evaluate potential effects of exposure to dopamine agonists during pregnancy.
13
Methods Study design The present study was conducted using data from EFEMERIS cohort (Evaluation chez la Femme Enceinte des MEdicaments et de leur RISque). EFEMERIS cohort EFEMERIS is a French database of all reimbursed drugs that are prescribed and dispensed during pregnancy (data from the French Health Insurance system [Caisse Primaire d’Assurance Maladie (CPAM) of Haute–Garonne]) and the pregnancy outcomes [data from Maternal and Infant Protection Service (PMI), the Antenatal Diagnostic Centre (CDA) and CPAM] [5]. The present study involved the 57,408 mother–outcome pairs included in the EFEMERIS database between July 1, 2004 and December 31, 2010. The study was approved by the French Data Protection Agency. Women were informed of the study and could refuse to participate. Dopamine agonist drug exposure In France, dopamine agonists are approved for several indications. Bromocriptine is approved for hyperprolactinemia, Parkinson’s disease or suppression of lactation. Cabergoline and quinagolide are approved only for hyperprolactinemia,
Arch Gynecol Obstet
lisuride for hyperprolactinemia and lactation suppression and ropinirole for Parkinson’s disease. Piribedil is a piperazine derivative which acts as a D2 and D3 receptor agonist. When the study was conducted, it was approved in France for the treatment of Parkinson’s disease, minor neurological disturbances related to aging, visual disorders of circulatory origin and adjunctive treatment of arteriopathy of the lower limbs (Table 1). To exclude women receiving a prescription during pregnancy for post-delivery lactation inhibition, women receiving a dopamine agonist only approved for lactation suppression and only during the third trimester of pregnancy, were excluded from the analysis. Women who only received a prescription, during the ninth month of pregnancy, for a dopamine agonist for which at least one indication was lactation suppression were also excluded. The French Health Insurance Service (CPAM) reimburses costs of care and medications. All expecting mothers declare their pregnancy to CPAM which systematically records individual information on all reimbursed drugs prescribed and dispensed in pharmacies to patients registered under the general state coverage [date of dispensation, drug name, Anatomic Therapeutic Chemical (ATC) code]. Collection of data by CPAM is prospective. In France, all dopamine agonists are reimbursed and thus require a medical prescription (licensed medicine). Consequently, all dispensations of these drugs are registered by CPAM. In this study, women who received at least one dispensation of a dopamine agonist during pregnancy were
Table 1 Dopamine agonist exposure during pregnancy in the cohort of 183 women prescribed dopamine agonists between 2004 and 2010 as recorded in the EFEMERIS database Indication for use (according to the French SPC)
Number (%) During pregnancy First trimester Second trimester Third trimester
Bromocriptine (5 and 2.5 mg) Hyperprolactinemia, Parkinson’s disease, lactation suppression
118 (64.5)
83 (60.6)
8 (66.7)
34 (77.3)
Cabergoline Hyperprolactinemia
37 (20.2)
32 (23.4)
2 (16.7)
4 (9.1)
Quinagolide Hyperprolactinemia
18 (9.8)
17 (12.4)
1 (8.3)
0
7 (3.8)
2 (1.5)
0
5 (11.4)
4 (2.2)
4 (2.9)
0
0
1 (0.5)
1 (0.7)
1 (8.3)
1 (2.3)
183 (100)
137 (100)
12 (100)
44 (100)
Lisuride Lactation suppression, hyperprolactinemia Piribedil Parkinson disease, minor neurological disturbances related to aging, visual disorders of circulatory origin, adjunctive treatment of arteriopathy of the lower limbs Ropinirole Parkinson’s disease Total (all drugs) Italic values correspond to percent values SPC summary of product characteristics
13
Arch Gynecol Obstet
considered as having been exposed to dopamine agonists and were classified in the exposed group. Unexposed group Women who did not receive a dispensation of a dopamine agonist during pregnancy constituted the unexposed group. Women exposed to a dopamine agonist were individually matched to two unexposed women, according to their age and the month-and-year of the beginning of their pregnancy. Adverse fetal outcomes Pregnancy losses were identified from data provided by CPAM and CDA. CPAM reimburses care associated with pregnancy loss (of any cause: voluntary and medical abortion, spontaneous abortion, stillbirth and ectopic pregnancy). However, CPAM does not record the details relevant to pregnancy losses. The CDA centralizes all requests for medical abortions in the region, and records data about birth defects, date of pregnancy and medical history. Birth defects were defined according to the International Statistical Classification of Diseases and Related Health Problems (ICD-10 Chapter XVII Congenital malformations, deformations and chromosomal abnormalities) as used by European network for surveillance of congenital anomalies (EUROCAT). Birth defects for live births were assessed by analysis of medical certificates established at ages 8 days, 9 and 24 months and collected by PMI Service. For cases of medical abortion, data about defects were obtained from the CDA. All birth defects, both major and minor, were recorded and analyzed. A birth before 37 gestational weeks was considered to be a preterm birth [6]. Analysis of preterm birth was restricted to exposure to dopamine agonists before 37 gestational weeks. Low birth weight was defined as a weight below 2,500 g [7]. Psychomotor development Data for psychomotor development was extracted from the medical certificates completed during medical examinations at ages 9 and 24 months [8]. Some items of these certificates concern motor development: “sit without support”, “move”, “catch an object”, “point a finger”, and “move the four limbs” at 9 months; and “superimpose objects”, “move limbs symmetrically”, and “walk without support” at 24 months. Other items relate to social and language development: “respond to his/her name”, “repeat a syllable”, and “play peek-a-boo” at 9 months; and “follow an instruction”, “give a name to a picture”, and “associate two words” at 24 months.
Statistical analysis Continuous variables are expressed as means (±SD) and categorical variables are expressed as percentages. The prevalence of dopamine agonist prescriptions and dispensations during pregnancy and during each trimester of pregnancy was calculated. Maternal characteristics and adverse fetal outcomes in the two groups (“exposed” and “unexposed”) were described and compared by simple conditional logistic regression. Then, we used conditional logistic regression to analyze risks for each outcome associated with dispensation of dopamine agonists. To identify confounders, we used simple conditional logistic regression to evaluate the association between each potential confounder (dispensation of folic acid and progesterone during organogenesis, multiple births, preterm birth, gender, mother’s occupation, number of ultrasound scans during pregnancy, pathologies during pregnancy) and each outcome. The variables which were significant at 20 % were introduced into the multivariate model. We used conditional forward stepwise multivariate logistic regression to calculate adjusted prevalence odds ratios, and 95 % confidence intervals (CIs). A value of p
Maternal-Fetal Medicine
Pregnancy outcome in women exposed to dopamine agonists during pregnancy: a pharmacoepidemiology study in EFEMERIS database Caroline Hurault‑Delarue · Jean‑Louis Montastruc · Anna‑Belle Beau · Isabelle Lacroix · Christine Damase‑Michel
Received: 9 July 2013 / Accepted: 4 March 2014 © Springer-Verlag Berlin Heidelberg 2014
Abstract Purpose The objective of this exposed–unexposed study was to evaluate potential effects of dopamine agonists during pregnancy. Methods Data from EFEMERIS, a cohort of 57,408 pregnant women living in South West France, were used to compare exposed and unexposed women. The exposed group included 183 women (0.3 %) who received at least one prescription for one dopamine agonist during pregnancy. These women were individually matched with two unexposed women from the cohort for age and the monthand-year of the start of pregnancy. Pregnancy losses, birth defects, preterm births, low birth weight and psychomotor development were studied. Results Bromocriptine was the most frequently prescribed dopamine agonist, followed by cabergoline and quinagolide. Most (75 %) of the dopamine agonists were prescribed at the beginning of pregnancy (first trimester). There was no difference between the two groups concerning pregnancy history and demographic data. After adjustment for potential confounders, prescription and
Electronic supplementary material The online version of this article (doi:10.1007/s00404-014-3210-z) contains supplementary material, which is available to authorized users. C. Hurault‑Delarue (*) · J.-L. Montastruc · A.-B. Beau · I. Lacroix · C. Damase‑Michel Pharmacologie Médicale et Clinique, Centre Midi‑Pyrénées de PharmacoVigilance, de Pharmacoépidémiologie et d’Informations sur le Médicament, INSERM UMR 1027, Faculté de Médecine, Université Toulouse 3, Centre Hospitalier Universitaire, 37 allées Jules Guesde, 31000 Toulouse, France e-mail: caroline.delarue‑hurault@univ‑tlse3.fr
dispensation of dopamine agonists was associated with an increased risk of pregnancy loss [PORa = 3.7; 95 % confidence interval (CI) 1.8–7.4] and preterm birth (PORa = 3.6; 95 % CI 1.5–8.3). The prevalence of birth defects and low birth weight was not significantly different between the two groups. No difference in psychomotor development at either 9 or 24 months was observed between the two groups. Conclusion This study suggests that prenatal exposure to dopamine agonists may be associated with an increased risk of pregnancy loss and preterm birth. Keywords Dopamine agonists · Pregnancy · Outcome · Psychomotor development
Introduction Pregnant women may suffer from diseases requiring dopamine agonists: hyperprolactinemia (linked to a prolactinoma in most of cases) or more rarely idiopathic Parkinson’s disease. Prevalence of dopamine agonist-treated hyperprolactinemia in women is estimated between 0.1 and 1.2 %, and is most frequent among those between 25 and 34 years old [1]. Incidence of Parkinson’s disease among women of childbearing age is extremely low [2]. As a consequence, very little is known about the effects of dopaminergic agonists on embryo–fetal development. Dopaminergic neurons appear early during development, at the gestational age of 6–8 weeks in humans [3] and several studies suggest that dopamine is important for development, particularly in motor and cognitive programs [4]. The aim of this exposed–unexposed study was to evaluate potential effects of exposure to dopamine agonists during pregnancy.
13
Methods Study design The present study was conducted using data from EFEMERIS cohort (Evaluation chez la Femme Enceinte des MEdicaments et de leur RISque). EFEMERIS cohort EFEMERIS is a French database of all reimbursed drugs that are prescribed and dispensed during pregnancy (data from the French Health Insurance system [Caisse Primaire d’Assurance Maladie (CPAM) of Haute–Garonne]) and the pregnancy outcomes [data from Maternal and Infant Protection Service (PMI), the Antenatal Diagnostic Centre (CDA) and CPAM] [5]. The present study involved the 57,408 mother–outcome pairs included in the EFEMERIS database between July 1, 2004 and December 31, 2010. The study was approved by the French Data Protection Agency. Women were informed of the study and could refuse to participate. Dopamine agonist drug exposure In France, dopamine agonists are approved for several indications. Bromocriptine is approved for hyperprolactinemia, Parkinson’s disease or suppression of lactation. Cabergoline and quinagolide are approved only for hyperprolactinemia,
Arch Gynecol Obstet
lisuride for hyperprolactinemia and lactation suppression and ropinirole for Parkinson’s disease. Piribedil is a piperazine derivative which acts as a D2 and D3 receptor agonist. When the study was conducted, it was approved in France for the treatment of Parkinson’s disease, minor neurological disturbances related to aging, visual disorders of circulatory origin and adjunctive treatment of arteriopathy of the lower limbs (Table 1). To exclude women receiving a prescription during pregnancy for post-delivery lactation inhibition, women receiving a dopamine agonist only approved for lactation suppression and only during the third trimester of pregnancy, were excluded from the analysis. Women who only received a prescription, during the ninth month of pregnancy, for a dopamine agonist for which at least one indication was lactation suppression were also excluded. The French Health Insurance Service (CPAM) reimburses costs of care and medications. All expecting mothers declare their pregnancy to CPAM which systematically records individual information on all reimbursed drugs prescribed and dispensed in pharmacies to patients registered under the general state coverage [date of dispensation, drug name, Anatomic Therapeutic Chemical (ATC) code]. Collection of data by CPAM is prospective. In France, all dopamine agonists are reimbursed and thus require a medical prescription (licensed medicine). Consequently, all dispensations of these drugs are registered by CPAM. In this study, women who received at least one dispensation of a dopamine agonist during pregnancy were
Table 1 Dopamine agonist exposure during pregnancy in the cohort of 183 women prescribed dopamine agonists between 2004 and 2010 as recorded in the EFEMERIS database Indication for use (according to the French SPC)
Number (%) During pregnancy First trimester Second trimester Third trimester
Bromocriptine (5 and 2.5 mg) Hyperprolactinemia, Parkinson’s disease, lactation suppression
118 (64.5)
83 (60.6)
8 (66.7)
34 (77.3)
Cabergoline Hyperprolactinemia
37 (20.2)
32 (23.4)
2 (16.7)
4 (9.1)
Quinagolide Hyperprolactinemia
18 (9.8)
17 (12.4)
1 (8.3)
0
7 (3.8)
2 (1.5)
0
5 (11.4)
4 (2.2)
4 (2.9)
0
0
1 (0.5)
1 (0.7)
1 (8.3)
1 (2.3)
183 (100)
137 (100)
12 (100)
44 (100)
Lisuride Lactation suppression, hyperprolactinemia Piribedil Parkinson disease, minor neurological disturbances related to aging, visual disorders of circulatory origin, adjunctive treatment of arteriopathy of the lower limbs Ropinirole Parkinson’s disease Total (all drugs) Italic values correspond to percent values SPC summary of product characteristics
13
Arch Gynecol Obstet
considered as having been exposed to dopamine agonists and were classified in the exposed group. Unexposed group Women who did not receive a dispensation of a dopamine agonist during pregnancy constituted the unexposed group. Women exposed to a dopamine agonist were individually matched to two unexposed women, according to their age and the month-and-year of the beginning of their pregnancy. Adverse fetal outcomes Pregnancy losses were identified from data provided by CPAM and CDA. CPAM reimburses care associated with pregnancy loss (of any cause: voluntary and medical abortion, spontaneous abortion, stillbirth and ectopic pregnancy). However, CPAM does not record the details relevant to pregnancy losses. The CDA centralizes all requests for medical abortions in the region, and records data about birth defects, date of pregnancy and medical history. Birth defects were defined according to the International Statistical Classification of Diseases and Related Health Problems (ICD-10 Chapter XVII Congenital malformations, deformations and chromosomal abnormalities) as used by European network for surveillance of congenital anomalies (EUROCAT). Birth defects for live births were assessed by analysis of medical certificates established at ages 8 days, 9 and 24 months and collected by PMI Service. For cases of medical abortion, data about defects were obtained from the CDA. All birth defects, both major and minor, were recorded and analyzed. A birth before 37 gestational weeks was considered to be a preterm birth [6]. Analysis of preterm birth was restricted to exposure to dopamine agonists before 37 gestational weeks. Low birth weight was defined as a weight below 2,500 g [7]. Psychomotor development Data for psychomotor development was extracted from the medical certificates completed during medical examinations at ages 9 and 24 months [8]. Some items of these certificates concern motor development: “sit without support”, “move”, “catch an object”, “point a finger”, and “move the four limbs” at 9 months; and “superimpose objects”, “move limbs symmetrically”, and “walk without support” at 24 months. Other items relate to social and language development: “respond to his/her name”, “repeat a syllable”, and “play peek-a-boo” at 9 months; and “follow an instruction”, “give a name to a picture”, and “associate two words” at 24 months.
Statistical analysis Continuous variables are expressed as means (±SD) and categorical variables are expressed as percentages. The prevalence of dopamine agonist prescriptions and dispensations during pregnancy and during each trimester of pregnancy was calculated. Maternal characteristics and adverse fetal outcomes in the two groups (“exposed” and “unexposed”) were described and compared by simple conditional logistic regression. Then, we used conditional logistic regression to analyze risks for each outcome associated with dispensation of dopamine agonists. To identify confounders, we used simple conditional logistic regression to evaluate the association between each potential confounder (dispensation of folic acid and progesterone during organogenesis, multiple births, preterm birth, gender, mother’s occupation, number of ultrasound scans during pregnancy, pathologies during pregnancy) and each outcome. The variables which were significant at 20 % were introduced into the multivariate model. We used conditional forward stepwise multivariate logistic regression to calculate adjusted prevalence odds ratios, and 95 % confidence intervals (CIs). A value of p
