Changes in drug response in the infant and neonate

This chapter answers parts from Section D(iii) of the 2017 CICM Primary Syllabus, which expects the exam candidate to "describe alterations to drug response due to physiological change, with particular reference to neonates/infants". It also refers to the very similar Section B(vi), where the trainees' objective is to understand "the fate of drugs in the body, including ...  how it is affected by extremes of age".  Even though this matter comes up twice in the syllabus, the primary examiners have never created any SAQs to test our understanding here. Weirdly, it has sort-of appeared as one of the sub-elements in the Fellowship Exam, where  Question 28 from the first paper of 2016 asked the candidates to compare assessment and management of poisoning in a two-year-old to the same issues in an adult. 

In summary:

In terms of references and peer-reviewed resources, this field is somewhat barren. In fact, looking for supporting articles on this topic one is often confronted with the realisation that virtually all the original research was done during the early 1970s. This may mean that these early pioneers were able to nail down all the answers and no further investigation was required; or, that ethics permissions for drug experiments on babies were easier to come by during the era of disco. Given how little we still know about neonatal pharmacology, one might assume the latter. 

  Alcorn & McNamara (2003) is probably the single best resource for pharmacokinetics of the newborn, though it is trapped in the dungeons of Elsevier and requires a ransom of around €30.00. The author of these notes has made some effort to ensure its content is well-summarised and represented here for the cash-strapped ICU trainee. Kearns et al (2003) is a good free alternative, though not specifically neonate-oriented. Some well-structured pharmacokinetics material can also be scraped from Geert ‘t Jong's chapter (Ch. 2, p.9) for Bar-Shalom & Rose's Pediatric Formulations: A Roadmap (2014) which is somehow (accidentally?) made gratis by Springer. The best article for pharmacodynamics in the neonate is the rare piece by Ohning (1995), available only through institutional access to the now-defunct Neonatal Network journal. 

For a purely toxicological perspective, one could also explore James Tibballs chapter for the seventh edition of Oh's manual (pp. 1148, "Paediatric  poisoning" )- it seems relevant purely by virtue of its having appeared in exams previously, but the time-poor primary candidate may safely leave it until later, as all toxicology has officially been shunted out of the Part I syllabus since the 2017 revision. For the purposes of quickly revising these issues later, a summary chapter is available to the Part Two candidates in the required reading section for Pharmacology and Toxicology. It contains virtually the same information as this chapter, but focuses on toxicology and detoxification, and has a less meandering narrative.

Drug absorption in infants and neonates

Oral absorption in the neonate is a fairly hit-and-miss affair, as in general is oral administration of medications. Its usual features of convenience and tolerability are usually modified by the fact that an unwell infant is frequently intolerant of everything, and nothing is convenient in their management. The forced administration of some foul-tasting oral formulation by a well-meaning stranger may result in the aforementioned stranger being covered in vomit. 

If one is victorious in the battle of wills over the syrup, the following pharmacokinetic changes in absorption are to be expected:

• Gastric pH is elevated in infants (~ 4.0) which leads to increased bioavailability of acid-labile compounds (eg. phenoxymethylpenicillin) and decreased bioavailability of weak acids (eg. phenobarbital). Lipid-soluble drug absorption will also be altered unpredictably. For the just-born neonate, the gastric pH is essentially neutral: Miclat et al (1978) swooped randomly upon "unselected neonates" in the  3rd and 4th minutes of life and suctioned their gastric juices in front of their horrified mothers; the pH was around 7 on average, particularly where the baby was premature. Heimann et al (1980) claim that gastric pH then drops to around 4.0 within hours of birth, and remains like that for many months.  Rødbro et al (1967) measured the stomach pH of infants up to 18 months of age and found pH values around 4.0, suggesting that gastric acid secretion does not "adultify" until well into the second year of life.
• Gastric emptying is poor during the first week of life, but normalises quickly. The main reason for this sluggishness is the immaturity of neurological regulatory mechanisms responsible for gastric and intestinal motor function. Berseth (1996) discusses this in lavish detail - for this summary, it will suffice to say that delayed gastric emptying is more of a problem the more premature the infant. 
• Active and passive intestinal molecule transport is slow in the first four months of life, but transport mechanisms mature around the time you are supposed to be starting solids. Villous formation completes its maturation process around the 20th week of gestation. This probably has minimal influence on the total bioavailability of drugs, as was demonstrated with xylose and arabinose by Heimann et al (1980). Though absorption was overall slower, the total absorbed drug (using Dost's law) was roughly the same.
• Intestinal microflora changes throughout infancy, influencing the degradation of drugs like digoxin. Linday et al (1986) studied digitalised infants and found that their gut bacteria were not able to reduce the lactone ring on digoxin, unlike adults. Adult patterns of gut microflora are not established until well into the second year of life, presumably in parallel to the introduction of solids and adult-like diet.
• Cutaneous  absorption is more rapid, because babies and small children have a thinner stratum corneum. One ends up with more corticosteroids absorbed systemically than one might expect from the sparse application of creams and ointments. Toxicity may result from increased systemic absorption of usually benign antiinflammatory NSAID creams used for nappy rash. West et al (1981) performed an amazingly detailed review of paediatrric skin pharmacology, listing altered absorption kinetics for drugs such as oestrogens, chlorhexidine, adrenaline, salicylic acid, boric acid, phenols, and so on.
• Intramuscular depot absorption is less rapid in neonates because skeletal muscles in a neonate receive a decreased blood supply, and because skeletal muscle contractions which usually disperse the drug are inefficient. However, there are more capillaries in those skeletal muscles, and the absorption of some drugs is actually increased, particularly if they are poorly lipid-soluble and depend on large surface areas for absorption. Back in the day when you'd be routinely giving IM amikacin to babies, Kafetzis et al (1979) demonstrated that its absorption was better in the younger age group.
• Increased absorption by inhalation increases susceptibility to volatile toxins; children have a higher baseline respiratory rate and minute volume (in terms of ml/kg/min) than adults. A classic example of this is carbon monoxide toxicity: among exposed groups from the same fire, the most severe symptoms are usually in the youngest child. The same is not true of nebulised or aerosolised medications (which are used extensively in the NICU population, where De Luka et al (2011) list numerous examples). Aerosolised drugs are absorbed poorly because of smaller tidal volumes and lower peak inspiratory flow rates; Dolovich (1999) reports bioavailability figures of around 2% for nebulised drugs and suggests that adult dosing may be needed to counteract this phenomenon.

Drug distribution in infants and neonates

The main changes in distribution are due to differences in body composition, protein affinity and changes in cardiac output:

• Neonatal bodies contain more water (70%)
• Neonatal bodies contain less fat

Thus:

• Volume of distribution is greater for water-soluble drugs: neonates and infants have relatively larger extracellular and total-body water spaces. Friis-Hansen (1983) gives a good overview of these matters; the paper also includes an excellent graph which is reproduced here with zero permission:
  
  As one can see, gender differences in body water content do not develop unti puberty.
• Volume of distribution is smaller for fat-soluble drugs: neonates and infants have a higher water-to-fat ratio in their adipose tissue; drugs which rely on lipid redistribution for their offset effect will linger. This is probably not as significant as the raw fat:water proportions would make you think.  Bartelinke et al (2006) reported that lipophilic drugs such as diazepam and lorazepam have a relatively similar Vd in neonates and adults.

Also:

• Albumin and α1-acid glycoprotein levels are lower in neonates and infants, increasing the free fraction of highly protein-bound drugs. To again misappropriate a copyrighted image from the published literature, one may refer to this chart from Alcorn et al, where AAG is α1-acid glycoprotein and ALB is duh.
  
   Also foetal albumin is quite unlike adult albumin, and has a decreased capacity to bind acidic drugs. Ehrnebo et al (1971) found that this affected the free fractions of ampicillin, benzylpenicillin, phenobarbital and phenytoin. The hyperbilirubinaemia of infancy also plays a role in changing the free drug fraction, as bilirubin has the tendency to compete for binding sites with other ligands.
• The blood brain barrier is immature, apparently with larger pore sizes. This results in an enhanced penetration of the drug into the CNS. Good examples of this are phenytoin and phenobarbital: Painter et al (1981) found increased ratios of brain:blood phenobarbital levels in pre-term neonates at 28-39 weeks of gestational age. This is assisted by proportionally larger blood flow (as a fraction of cardiac output) and by the proportionally larger neonatal brain as compared to other organs (by mass).

Drug metabolism in infants and neonates

Apart from protein binding, hepatic blood flow and intrinsic metabolic enzyme activity are the main governing factors of drug metabolism, and all of these are different in the infant and neonate.

• Decreased expression of Phase I metabolic enzymes means infants and neonates are slow metabolisers. Generally, an increase in the activity of these systems results in rates of drug clearance which in infants and toddlers exceed the of adults, for example in the case of caffeine and theophylline, or phenytoin. However, in the neonatal period, all of these enzymes are somewhat dormant. In fact, it appears that for many CYP enzymes, parturition triggers postnatal expression, and all the aforementioned drugs (theophyllline, caffeine, etc) are actually metabolised more slowly in the first few months of life. Alcorn & McNamara (2003)  spend a lot of time going through the numerous CYP enzymes and how their activity changes in infancy. For the purposes of sanity, these data can be summarised as "metabolism of everything is sub-normal until several months have passed".
• Variable expression of Phase II enzymes means that during early infancy the expression of some enzymes is depressed (eg. uridine 59-diphosphate-glucuronosyltransferase and N-acetyltransferase) and expression of other enzymes is mature almost adult-like (sulfotransferase and glutathione-S-transferase.). What does this mean, realistically? Probably not a lot. "The literature remains sparse and presents conflicting information", Alcorn & McNamara complain. All that the exam candidate can be expected to say on this issue is that the delayed maturation of some conjugation reactions may theoretically produce the accumulation of toxic intermediate metabolites.

Drug clearance in infants and neonates

Like with drug metabolism, one can summarise the pharmacokinetic changes in the neonatal kidney as "starts out useless and matures rapidly over the first year of life". 

• Glomerular filtration rate is reduced in pre-term neonates, with associated decreases in drug clearance. It reaches adult values by around 12 months of age, but until then it is markedly reduced. Nephrogenesis should be complete around the 36ths week of life, but renal function has not yet achieved adult performance levels at this stage. Most of the increase of GFR over the first year actually seems to occur over the first two weeks of life (Kearns et al, 2003). This is mainly important for drugs which rely on glomerular filtration for their clearance, of which the most clinically relevant example would probably have to be aminoglycosides. Szefler et al (1980) ended up having to adjust their gentamicin dosing intervals to have to avoid toxic effects; of course because this was 1980 everybody gave their gentamicin as an eight-hourly dose, which meant that this "internal attenuation" merely brought this old study back into line with modern dosing concepts. 
• Active tubular secretion is immature at birth, but achieves an adult-like level of function by the end of the first year.  reports that generally tubular absorption matures faster than tubular secretion. Secretion in the tubule is impaired for multiple reasonss, including "poor peritubular blood flow, shorter tubular length, reduced urine concentrating ability, lower urinary pH values and decreased energy for transporters as well as decreased transporters’expression" (Ligi et al, 2013). Though there is little data to support this, the failure of active secretion may delay the clearance of drugs which depend on (for example) organic anion transporters, eg. non-steroidals and beta-lactams. 

Pharmacodynamics in infants and neonates

This is poorly understood. According to Kearney et al (2003), "little information exists about the effect of human ontogeny on interactions between drugs and receptors and the consequence of these interactions". The majority of the difference in drug response in infants and neonates is because of differences in pharmacokinetics, particularly clearance. However, there are probably a few true age-dependent differences in the interaction between a drug and its specific receptor. Thus, there are age-related differences in the relationship between the plasma level and pharmacologic effect. This is seen in warfarin, cyclosporine, midazolam, and valproate. For the rest, a list of known specific differences in drug effects is produced here, as a pragmatic offering to the exam candidate who is  asked for examples.

• Respiratory depression occurs more readily in children, as their susceptibility to respiratory failure is increased (this includes both opiate effects and the effects of aspiration)
• Relative digoxin resistance explored by  Linday et al (1986) and found to be unrelated to absortion kinetics
• Hypoglycaemia occurs more readily because of decreased glycogen stores.
• Cardiovascular collapse occurs more precipitously: cardiac output is more heavily reliant on heart rate, and adrenergic tone is increased. Tachycardia will be the solitary vital sign abnormality until everything falls in a heap.
• QT prolongation occurs more readily in infants
• Paradoxical reaction to benzodiazepines and sedating antihistamines has been observed in the under-2s
• Reduced MAC requirements for volatile agents are seen in infants
• Decreased response to bronchodilators because of a supposed decrease in the expression of beta-2-receptors on the bronchial mucosa (Chavasse et al, 2002)