Prescribing in pregnancy and lactation

The topic of drug ingestion related to pregnancy, and the peri-pregnant period including lactation has always been important but under-studied. It is therefore gratifying to see that in this issue there are 4 papers devoted to this topic, covering the period from 3 months prior to conception to early post-partum, including lactation.

The importance of observational data

The paper by Engeland et al.[1], describes prescription drug use in Norway from 3 months before to 3 months after pregnancy. What makes the paper particularly interesting is that it effectively covers the entire Norwegian population, courtesy of 2 major databases – the NorPD (Norwegian Prescription Database) which captures all dispensed prescriptions for the entire Norwegian population (4.7 M) from the start of 2004, and MBRN (the Medical Birth Registry of Norway), which contains information on all births in Norway (>2.3 M) since 1967. The two databases link easily through common identification numbers.

A creditable and unusual aspect was the inclusion of data related to drug ingestion by fathers, in the three month period prior to conception. It is largely unknown how highly teratogenic drugs such as isotretinoin (which was taken by 13 men in this study) affect offspring. The study showed that around 25% of fathers had a prescription medicine, with around 12% taking one drug only. By comparison, around 39% of women had a prescription during this period, which dropped to around 30% for each of the first to third trimesters, and rose to 57% for the three month post-partum period (presumably excluding drugs given at the time of delivery). The changing pattern of drug ingestion in the relevant periods was interesting, and generally reflected what might be regarded as good medical practice in a developed country. The extent of drug ingestion around the time of conception remains a concern, especially related to unplanned pregnancies.

Our knowledge about the safety of drugs in the immediate pre-pregnant period is scanty, yet the ova and sperm about to make contact are subject to possible effects at this time. Methotrexate, a folate antagonist and potential teratogen and abortifacient, may adversely affect developing sperm and ova, and it is recommended that pregnancy should be avoided for at least 3 months after cessation of treatment in males, and for at least 1 ovulatory cycle in females. However, proof of this is lacking [2]. Pharmacokinetic considerations are also important at this time. Drugs with very long half-lives will continue to have influence long after they have been ‘stopped’. Patients are very likely not to declare such drugs at the time of a drug history (if indeed asked!). Fortunately there is a short window of relative safety between conception and implantation, where the conceptus is relatively spared of drug in the maternal circulation [3].

Pregnancy and the post-partum period have historically been largely excluded from drug trials, for obvious reasons. Koren [4] has stated that the best way to achieve better knowledge of effects of drugs during pregnancy is collection and follow-up of observational data. More countries should follow Norway's example in their data collection strategies. Interestingly, I have tried to establish a database related to pregnancy and drugs in our local region in New Zealand, but persistently struck resistance (largely from medical colleagues) related to privacy issues. This is political correctness gone mad.

Transporters and the placenta

Transporters seem to be a flavour of our times, so it is not surprising that they have been looked for, and found, in the human placenta. Transporters often serve a useful Darwinian purpose of protection, which, in the case of the placenta, is of the foetus. Inhibitors of transporters are therefore of great importance, because they potentially increase exposure.

Sudhakaran et al.[5] investigated the effect of P-glycoprotein (P-gp) inhibition on maternal to foetal transfer of indinavir. They had previously shown that indinavir was subject to efflux transport (i.e. foeto-protective) in the isolated human placenta, which was presumed to be via P-gp [6]. In the study in this issue of the journal, a P-gp inhibitor PSC833 enhanced maternal to foetal transfer of indinavir, but not of ritonavir. The differential effect was considered to be due to clinically relevant concentrations in relation to different binding affinities of indinavir and ritonavir.

Since P-gp has a large number of substrates and inhibitors, one can only speculate as to how many other drug interactions may occur at this level.

From pregnancy to lactation

A cooperative study involving researchers from Western Australia and Papua New Guinea, (Law et al.[7]), investigated the transfer of chloroquine and desethylchloroquine (the primary active metabolite) across the placenta, and into milk, in Melanesian mothers. In Papua New Guinea chloroquine is recommended for prophylaxis of malaria during pregnancy, and many women are routinely prescribed this drug at the time of delivery.

In this study foetal cord and maternal serum samples were collected at delivery, and milk samples were collected over 3 weeks after delivery. It was commendable that the active metabolite was measured as well as the parent drug. Results showed that during pregnancy concentrations were similar on the infant and maternal sides, while the relative infant dose of the 2 compounds (as chloroquine equivalents) was around 3.2%, and therefore compatible with breastfeeding. It is generally accepted that for drugs without excessive toxicity, a relative infant dose of <10% is considered ‘safe’ for full term infants.

This study nicely confirms the most common situation related to drugs in pregnancy and lactation – that in pregnancy the infant is usually exposed to similar concentrations as is the mother, while during lactation the relative exposure is much lower. It is amazing how often this comes as a surprise to many people.

A second paper from Western Australia, from Ilett et al.[8], examined the use of a sparce sampling design to assess transfer of tramadol and its active o-desmethyl metabolite, into transitional milk. This paper was interesting for 2 reasons. Firstly it provided the first detailed data on tramadol in lactation, when used as postoperative analgesia on days 2-4 after Caesarian section. The combined relative infant dose of parent and metabolite drugs was 2.9%, indicating that short term use in this setting is compatible with breastfeeding.

The other point of interest in this paper is the use of a sparce sampling study technique. Most studies of drugs in human milk involve few woman/child pairings (usually ≤6), with intensive sampling across a dosing interval. In this study, data were provided on 73 women, with 3 milk samples and one plasma sample taken in each woman, distributed across a dosing interval. The authors believed that this was the first time such an approach has been used in this field. Such an approach could lend itself to population pharmacokinetic modelling, and I know the authors are interested in pursuing this approach further.