Jan Willem van der Laan
Centre for Biological Medicines and Medical Technology
National Institute of Public Health and the Environment
Labelling of medicines is important for their safe and effective use. Information on pregnancy and lactation is summarised on the label. This information should be relevant for the prescriber. What kind of information is relevant for the prescriber should be better defined, such as which data should be communicated, when these data are solid enough and how they should be communicated.
European regulation authorities are currently working on a Note for guidance on risk assessment of medicinal products on human reproductive and development toxicities: from data to labelling. This document intends to advise how to communicate on the risks of exposure to medicinal products during pregnancy.(1) The prescriber should be informed adequately on the risk of possible toxicity of a drug to human reproductive function and development, and also on the management of this risk in clinical practice. The risk assessment is established by combining the results of nonclinical and clinical data.
The nonclinical integration process assesses the potential risk of an adverse effect in humans, based on reproductive and developmental toxicity studies in animals. Therefore, knowledge of the pharmacology, pharmacokinetics and toxicology of the substance is important. Which pharmacodynamic effects may be expected in the embryo/fetus? Does the drug reach the embryo/fetus/ breastfed neonate? What are the findings in the reprotoxicological studies? Are there findings in repeated dose studies, genotoxicity and carcinogenicity studies that should be taken into account? Additionally, it is possible that nonclinical data are absent or insufficient. What decisions will have to be made in that case?
The clinical integration process assesses the potential risk of an adverse reproductive/developmental effect in humans concerning fertility, pregnancy, health of the fetus and child, based on human clinical data. The guidance will describe how the risk analysis should be performed. The guidance will also express how to take into account the nature of the risk (eg, malformative, fetotoxic), the importance of the fetotoxic effect (eg, reversibility), the maternal indication and the risk to the fetus/newborn of the maternal disease itself, if not treated.
NSAIDs during pregnancy
The process of integrating nonclinical and clinical data for labelling in pregnancy may be difficult. We will discuss it on the basis of an example. The product information on pregnancy for NSAIDs (nonsteroidal anti-inflammatory drugs) has been under discussion for the last few years, as it was felt that the labelling should warn prescribers about cardiac malformations if these drugs were to be used during the first trimester of pregnancy. However, different opinions are possible on the interpretation of putative adverse findings in animal and human studies.
Although there is nonclinical evidence that use during the embryonic stage (comparable to the first trimester in humans) may lead to cardiac malformation, the interpretation of these data is difficult due to inadequacy or inconsistency of the data. Some studies may point to a developmental risk, whereas others do not. Cook and colleagues extensively reviewed studies on nonclinical developmental toxicity after administration of NSAIDs.(2) They presented incidences of diaphragmatic hernia (DH), midline defects (MDs) and ventricular septum defects (VSDs). In most of the studies, no such malformation occurred. In one rat study with meloxicam, VSDs were observed in the treated groups, but also in the concurrent controls.(2) Furthermore, effects were not dose-related. DH, MD and VSDs occurred incidentally in other studies, although no relation with treatment was observed. Incidences of MDs and VSDs were increased in some studies if aspirin was administered at high doses (200mg/kg/day).(2)
A marginally positive association was present between NSAIDs/aspirin treatment and the occurrence of MDs in rats (p=0.043). Obviously, the marginal difference was due to aspirin; excluding aspirin data resulted in a lack of correlation between NSAID administration and MDs. Furthermore, there was no indication of an association between NSAID/aspirin treatment and the occurrence of DH and VSDs in rats. In rabbits, no relation was observed between NSAID treatment and the three developmental defects discussed. In our opinion, this review does not show that it is correct to classify cardiovascular defects as a class effect of NSAIDs.
Human data should also be assessed critically. Some authors suggest an association between NSAID use and (cardiac) malformations, whereas other authors found no association between these parameters. Ericson and colleagues suggested that there is a higher incidence of cardiac defects in infants exposed to NSAIDs prenatally.(3) However, several issues should be taken into account. For example, in this study the number of affected infants was small, and often exposure time and/or dosage were not known. Exposure may have taken place beyond the period of cardiac development. Furthermore, the classification of cardiac defects is a matter of discussion. For example, some forms of VSDs may disappear spontaneously and will not be classified as a congenital malformation.
In a more recent study from the same group, in which the same database as in the above study is used, earlier associations disappeared and a relation was found only with naproxen (OR: 1.14–2.54).(4) The authors themselves admitted the weakness of the database used and stated that the results of the study should be regarded merely as a signal. They concluded that further studies would be necessary.
Some studies concluded on the absence of a relationship between maternal NSAID use and (cardiac) defects in the child. In a case-control study with 168 muscular VSD-carrying infants and 692 controls, it was concluded that: “[…] significant associations were not detected between the occurrence of VSD and maternal use of NSAIDs or acetominophen adjusting for maternal fever […]”.(5) From the study of Nielsen and colleagues, it was concluded that use of NSAIDs does not increase the risk of adverse birth outcome.(6)
In our opinion, data are not convincing enough to conclude that NSAIDs can induce cardiovascular defects after use during the first trimester of pregnancy. Further confirmative studies are needed before considering inclusion of a warning for these effects in the medicinal product information. The debate on NSAIDs could have gained from a specific expert meeting under the auspices of the European Medicines Agency, working along the lines of the guidance document, which is currently under development.
Clear guidance is needed to integrate adequately nonclinical and clinical data on the use of medicinal products during pregnancy. European regulatory agencies are now developing a document that may facilitate the integration of available data in order to obtain a clear and informative instruction for the pharmaceutical industry, regulatory agencies and prescribers.
- Concept paper on the development of a CPMP note for guidance on risk assessment of medicinal products on human reproductive and development toxicities: from data to labelling. 2001. Available from: http://www.emea.eu.int/pdfs/human/swp/037301en.pdf
- Cook JC, Jacobson CF, Gao F, et al. Analysis of the nonsteroidal anti-inflammatory drug literature for potential developmental toxicity in rats and rabbits. Birth Defects Res B Dev Reprod Toxicol 2003;68:5-26.
- Ericson A, Kallen BA. Nonsteroidal anti-inflammatory drugs in early pregnancy. Reprod Toxicol 2001;15:371-5.
- Kallen BA, Otterblad Olausson P. Maternal drug use in early pregnancy and infant cardiovascular defect. Reprod Toxicol 2003;17:255-61.
- Cleves MA, Savell VH, Raj S, et al. Maternal use of acetaminophen and nonsteroidal anti-inflammatory drugs (NSAIDs), and muscular ventricular septal defects. Birth Defects Res A Clin Mol Teratol 2004;70:107-13
- Nielsen GL, Sorensen HT, Larsen H, et al. Risk of adverse birth outcome and miscarriage in pregnant users of non-steroidal anti-inflammatory drugs: population based observational study and case-control study. BMJ 2001;322:266-70.