This chapter is related to one of the aims of Section D(iii) from the 2017 CICM Primary Syllabus, which expects the exam candidate to "Describe alterations to drug response due to physiological change, with particular reference to ... pregnancy".
In the CICM Part Two exam, this matter has appeared only as a minor footnote in the answer to Question 28 from the first paper of 2016, where the trainees were asked about the toxicological differences between management of the poisoned child, pregnant woman and renally impaired elder. The chapter on "Pharmacology and toxicology in pregnancy" in the Part Two required reading section revisits this topic as a brief overview. In the Part One exam, Question 9 from the first paper of 2020 was the first appearance of this topic, and it clearly took everybody by surprise (the pass rate was 7%).
- Some factors decrease intestinal absorption:
- Gastric emptying is delayed
- Gastic pH is higher
- Some factors increase intestinal absorption:
- More complete absorption because of slower gut transit
- Increased gastrointestinal blood flow
- Net effect is stable oral bioavailability for most drugs over the course of pregnancy.
- Increased cardiac output affects changes in skin and muscle blood flow, supposedly increasing the absorption of drugs via subcutaneous and intramuscular routes
- Increased blood flow speeds up the rate of onset of IV drugs, eg. muscle relaxants and anaesthetic agents
- Increased pulmonary blood flow and increased respiratory rate allow an increased rate of volatile anaesthetic uptake and therefore decreased time until onset of effect
- Decreased peridural space due to venous engorgement decreases the required dose of local anaesthetics
- Increased volume of distribution: at the third trimester, blood volume is increased by 50%
- Decreased protein binding due to diluted serum proteins
- Delayed release of toxins from fat stores: xenobiotics are laid down together with increasing fat stores in early pregnancy, and then mobilised during later stages and the postpartum period
- Increased fatty acid levels: they will compete with drugs for binding sites on albumin
- Increased sensitivity to local anaesthetics (due to decreased α1-glycoprotein levels)
- Altered hepatic clearance:
- circulating hormones can induce or inhibit metabolic enzymes.
- Progesterone induces enzymes
- Oestrogen competes for metabolic enzymes (eg. with vecuronium and rocuronium)
- Decreased plasma cholinesterase activity (though this does not result in increased duration of action for suxamethonium)
- The placenta has liver-like biotransformative enzymes, although these are not equivalent to the liver in their metabolic capacity, and probably not an effective protection for the foetus.
- Renal clearance of drugs increases mainly due to increased glomerular filtration rate. Drugs cleared solely by glomerular filtration are most affected (eg. cephazolin, clindamycin)
- Tubular resorption of substances also increases, counteracting the increased GFR
- Thus, renal clearance for any specific drug is difficult to predict - eg. clearance doubles for lithium, increases 30% for digoxin, and only increases 12% for atenolol (Feghali et al, 2015)
- Hepatobiliary clearance of drugs is reduced by the cholestatic effects of oestrogen
- Increased sensitivity to volatile anaesthetics (decreased MAC)
- Increased sensitivity to IV anaesthetics
- Increased sensitivity to local anaesthetics
- Changed therapeutic indices due to concerns regardin foetal damage and teratogenicity
The best single article on this topic actually happens to be Chapter 60: Obstetric and Fetal Pharmacology (Giacoia & Yaffe), which is inexplicably available free and in full online from www.glowm.com, where the entire contents of a 6-volume O&G textbook (Gynaecology and Obstetrics, Sciarra, 2004) appear to have been uploaded as a HTML file. Kudos to them for doing that, but it may be subject to a takedown notice from Lippincott & Williams at any moment, and is in any case 15 years out of date at the time of writing. Legit peer-reviewed alternatives include Gail Anderson's article for Clinical Pharmacokinetics (2005), which covers the PK component very well, but it is kept under lock and key by Springer, and they're charging 40 EUR for it. Ansari et al (2016) is a satisfactory free option, which can be acquired via ResearchGate and which covers the same territory.
This area is strangely somewhat under-researched, possibly because the typical method of researching bioavailability kinetics is to give oral and IV doses of a radiolabelled drug and then to measure the radioactivity of the subjects' bloodstream to compare their concentration/time curves (Dost's law, in case you care). That's probably difficult to get through ethics committees, who would be reluctant to approve the exposure of a developing foetus to ionising radion for scientific purposes.
For oral drug administration, pregnancy is often described as a generalised malabsorption state. There are multiple contributing factors for this:
The mechanisms for this are also multiple:
It is not clear when in pregnancy these effects start to take place, or when they peak. In general with pregnancy-related physiological changes it is generally assumed that the changes get more prominent the more pregnanter one is. Following from this, one might conclude that this absorption problem is maximal during labour.
Also following from the entire discussion, one might conclude that oral dose of drugs in pregnancy would need to be adjusted because of reduced oral availability. That turns out to be incorrect, or at least bioavailability is probably not the main problem. In 1977, Agneta compared oral and IV dosing of ampicillin in women before and after delivery, concluding that bioavailability of ampicillin was not affected (but larger doses were still required because of changes to the volume of distribution and rapid renal clearance). Similarly, Lander et al (1984) found that the bioavailability of phenytoin was not affected by pregnancy, and that increased metabolic clearance was responsible for the generally lower serum levels. Sotalol bioavailability is also unaffected (O'Hare, 1983).
The decreased gastric emptying and other pharmacokinetically counterproductive aspects of pregnancy are therefore probably offset by some absorption-promoting aspects. Again, there is little data to support anything said about the subject, which of course does not stop experts from saying things anyway. Most papers offer something like "increased cardiac output and intestinal blood flow may allow for increased drug absorption overall" (Feghali et al, 2015) without giving any references for this assertion.
Because of these changes in cardiac output and regional blood flow, absorption from non-oral routes is also supposed to be altered. There is increased cutaneous and muscular blood flow and the cardiac output is generally increased, which suggests that the subcutaneous and IM administration of drugs should be successful in achieving rapid plasma peak levels. Turns out, that's also not an accurate assumption. For example, the plasma levels of subcutaneous enoxaparin are not changed in pregnancy (Casele et al, 1999), and that's probably one of the most common subcutaneously administered substance in pregnancy.
Cardiac output changes in pregnancy still do cause significant changes in the response to drugs. For instance, the increased cardiac output causes an apparent sensitivity to muscle relaxants. Baraka et al (1992) found that the onset of vecuronium NMJ blockade was more rapid in the pregnant group. This is not a trivial change - the onset of 50% electromyographic block was 80 seconds in the pregnant group, vs. 140 seconds in the controls. The authors determined that the increased blood flow promoted the delivery of the IV drug to the sites of action.
Main factors affecting the distribution of drugs in pregnancy are:
This gives rise to three main pharmacokinetic differences:
This should have implications for the dosing of drugs. The total body water increases by about 8L on average, of which probably about 60% is sequestered in the amnion placenta and foetus; the body fat increases by about 4kg on average (Krauer & Krauer, 1977). This is generally said to change the volume of distribution for both hydrophilic and lipophilic drugs. The same pre-pregnancy dose, administered to a severely pregnant woman, should result in greatly diminished blood concentrations of the drug, whether it is water or fat-soluble.
As with bioavailability, minimal data exists to support these assertions. Gerdin et al (1990) found that the volume of distribution for the (highly lipophilic) morphine was either essentially the same, or actually decreased in pregnancy (from around 3.5L/kg to about 2.5L/kg) which is completely the opposite of what one might expect. Some increased Vd measurements have been reported for lithium digoxin and lamotrigine, but these references are paywalled, making it hard to determine their value. In summary, if pushed for examples, the trainee could confidently say that the effects of pregnancy on the volume of distribution of drugs are largely theoretical.
Protein-binding effects are also something which sounds potentially serious, but which probably plays no role in clinical decisionmaking. For instance, Tsen et al (1999) found that pregnancy increased the free fraction of bupivacaine (due to decreased α1-glycoprotein levels), which - according to the authors - is a serious safety issue as it may produce toxicity with conventional doses. In reality, this does actually seem to happen. Morishima et al (1989), while trying to kill pregnant ewes with bupivacaine, found that circulatory collapse occurred at 5mg/kg in the pregnant ewe and at 9mg/kg in the non-pregnant version. However, as the maximum recommended dose of bupivacaine is usually quoted as 2mg/kg, no sane person should ever approach this cardiovascular toxicity threshold with a routine anaesthetic technique.
There are additional factors affecting distribution in pregnancy. Specicially, this refers to the difference in doses of spinal anaesthetic required for peripartum woemn. Supposedly, the engorgement of the peridural venous structures results in the displacement of the CSF out of the spinal canal. As th rsult, the pool of lumbar CSF is smaller,and so the local anaesthetic injected there is likely to end up more concentrated and therefore of a higher potency. Wherever this is mentioned in the textbooks, one study always ends up being the reference. It is Assali & Prystowsky (1950), who found a four-fold reduction in dose requirements to achieve the same level of sympathetic blockade.
In short, the effects of pregnancy on drug metabolism are variable and not entirely preductable:
Curare toxins like vecuronium and rocuronium have a slower offset because of pregnancy-related changes in hepatic blood flow. On top of that, the female reproductive program forces the synthesis of steroid reproductive hormones on a truly titanic scale, and these compete with the aforementioned muscle relaxants for transport proteins in the hepatic sinusoids. The upshot of this is delayed metabolism. Khuenl-Brady et al (1991) found that the clinical effects of vecuronium took about 15-17 minutes longer to wear off in pregnant women, on average.
Some hepatic enzymes are less active during pregnancy. For example, CYP1A2, xanthine oxidase and N-acetyltransferase activities are decreased. Bologa et al (1991) demonstrated this using caffeine as a metabolic probe (the enzyme decreased activity by 30%). Other enzymes are more active: CYP3A4 and CYP2D6 activity is increased, producing an increase in the clearance of phenytoin and phenobarbital (Lander et al, 1977).
The increase in cardiac output increases the glomerular filtration rate, such that at the end of the third trimester it is about 50% increased from baseline. Thus, all dissolved drugs end up in the glomerular filtrate. What happens thereafter is completely random. The increased tubular resorption which goes hand in hand with the increased GFR could absorb some unpredictable proportion of the filtered drug, which would decrease the elimination rate. Each drug seems to be relatively
Hepatobiliary excretion of the (few) drugs which rely on this mechanism is supposed to be slowed instead. Oestrogen is said to have some cholestatic properties, and some authors (Dawes et al, 2001) suggest that clearance of drugs which rely on this mechanism "may be attenuated" giving rifampicin as an example. That turns out to be a reasonably good example. Looking at a population of TB/HIV patients in Soweto, Denti et al (2015) found the clearance of rifampicin to be reduced by 14%, which was again completely clinically irrelevant because dose adjustment was not needed.
Pregnant women are more sensitive to volatile anaesthetic agents. Palahniuk et al (1974) demonstrated this in pregnant ewes, and Gin & Chan (1994) in pregnant women undergoing termination of pregnancy (the MAC for isoflurane was reduced by 28% as compared to nonpregnant controls).
Pregnant women are also more sensitive to IV anaesthetic agents, at least in the case of thiopentone. The dose required is reduced by about 18% as compared to non-pregnant controls (Gin et al, 1997). Propofol is apparently not affected because of the changes to its volume of distribution.
Susceptibility to local anaesthetic is also increased, though exactly how and why this happens is not clear to any of the experts. It' clearly increased - for example, Butterworth et al (1990) found pregnant median nerves are blocked faster and more completely with lignocaine. Again progesterone is blamed.
Pregnancy expands the range of pharmacokinetic adverse effects which need to be taken into account, most famously by the presence of the delicate developing foetus. Beyond the period of organogenesis when you can really do some real harm, there is still an opportunity to cause injury to the foetus. This is a vast topic which should probably be left to a chapter all of its own. For the mother, one must consider the needs of the pregnant uterus. For example, volatile anaesthetic agents can cause uterine relaxation, which would be a good thing or a bad thing, depending on what the anaesthetist or obstetrician need for it to do.