[In utero exposure to benzodiazepine. Is there a risk for anal atresia with lorazepam?]

Drugs of the Benzodiazepine family are among the most frequently prescribed in France. Since anxiety disorders, for which these substances are mostly indicated, affect 10% of pregnant women, it is very likely that such a treatment could expose many foetuses to BZD during the first three Months of pregnancy. We know that the teratologic effect is not necessarily based on dose rate, but that it is associated with fetal drug exposure during the first 12 weeks of gestation, when organ formation occurs. Most epidemiologists concur that the baseline incidence of congenital damage is 2-2,5% in Europe. The results from a large number of stu-dies on associations between the use of BZDs in pregnancy and congenital malformations are conflicting. An in-depth analysis of existing literature shows results that are hardly comparable, if not contradictory, due to extreme differences in methodological approaches. In a recent meta-analysis case-control studies and cohort studies were analyzed separately. Among the case-control studies significant associations were found between BZD exposure and both, major malformations and oral clefts, whereas the cohort studies showed no association between BZD and any kind of malformation. The purpose of our study is to search for a specific teratogenic effect of this class of drugs, using data collected (1976-1997) by the French Central-East (FCE) registry of congenital malformations, member of the International Clearinghouse for Birth Defects Monitoring Systems (ICBDMS) located in Lyon, France. This registry monitor malformations among 100,000 births per Year. We analyzed 13,703 cases where information is available on whether or not the mother took a drug during her first trimester of pregnancy. Among them, 3,603 (28%) actually took a drug, and 262 (6.8%) took some sort of benzodiazepine (BDZ). BZD were divided into 9 categories, 8 being the most frequently present, plus one broad category of "others". Malformations were divided into ten categories: congenital anomalies of heart, cleft lip and/or cleft palate, neural tube defects, other anomalies of central nervous system, hypospadias, urinary malformations, anal atresia, other digestive anomalies, limb reduction defects, and genetic anomalies, including chromosome aberrations and monogenic conditions. Other malformations were grouped in an eleventh category. The interesting aspect of this study is that it takes into account the BZD metabolism. It is worth noting that the hepatic catabolism of benzodiazepine is a very complex one, because it leads to derived molecules which are sometimes active and/or present in the common metabolic route of major commercial drugs. Our hypothesis is that if one BZD is associated specifically to a certain type of congenital defect, we may find this BZD to be overrepresented, as compared with other BZDs, in newborns exhibiting some type of congenital defect. The analysis was run according to a case-control approach. Odd ratios (OR) and their 95% confidence intervals were calculated by logistic regression with adjustment for maternal age and parity. When one category of defects was considered, infants having the corresponding malformation were considered as cases, while infants with other malformations were considered as controls. In a similar way infants having being exposed to a given drug were considered as exposed, while infants exposed to any other drug were considered as unexposed. The analysis then was run in 4 steps. Step 1: full sample. With 13,703 cases. We observed no increased risk for any specific malformation type associated with use of BZD. Step 2: further defining drug exposure as a specific BZD, and all others unexposed, a significant association was seen between lorazepam and anal atresia. OR=6.2 (95% CI2.4-15.7, p=0.01). Step 3: this finding was upheld and no other emerged when exposure was defined as the drug or any of its active metabolites. This step was performed because hepatic catabolism of BDZs leads to derived molecules that are sometimes active and/or present in the common metabolic route of major marketed BZDs. Step 4: similarly, the lorazepam/anal atresia finding was upheld when the analysis was restricted to the 262 malformed infants exposed to BZDs in utero. Six cases of anal atresia were found among all newborns exposed to BZD in utero, and five of them were exposed to lorazepam, representing a hypothesis to be tested in further. We are not aware of other reports of this association, and it should be regarded as preliminary until confirmed in other data sets.