This chapter answers parts from Section B(vi) of the 2017 CICM Primary Syllabus, which expects the exam candidate to understand "the fate of drugs in the body, including ... how it is affected by extremes of age, obesity, pregnancy (including foetal) and disease (particularly critical illness)". This learning objective is also duplicated in Section D(iii), "Describe alterations to drug response due to physiological change, with particular reference to ...obesity". The revision chapter ultimately found its way into the section on pathological variability in drug response, as obesity is a disease and the effects it has are both pharmacokinetic and pharmacodynamic.
Like the influence of critical illness on pharmacology, this topic appears in both Part I and Part II exams. Question 20 from the first Fellowship paper of 2009 asks about the influence of morbid obesity on pharmacokinetics. This seems to have extended into Part I material - Question 13 from the second Part I paper of 2012 also asked about pharmacology and obesity, but this time omitting the descriptor morbid. Both college answers focused on pharmacokinetic changes, as the if the physiological changes of obesity have no pharmacodynamic effects.
In the Required Reading section for the CICM Part II, there is a brief point-form entry about the influence of morbid obesity on pharmacokinetics, but overall this topic seems to be very Part I material, and the bulk of the discussion is carried out here.
There is little of use in the official textbook, but thankfully we have plenty of published literature. Cheymol's "Effects of obesity on pharmacokinetics" (2000) has just enough depth for a time-poor candidate in the first three or four pages, and then dives deep into the pharmacokinetic effects of obesity on specific drug classes (which can become tedious if you are pressed for time). Another great review article is by Hanley et al (2010), which pays particular attention to the effects of different body weight calculations on drug dosing in obesity. Probably the best overview of pharmacokinetic changes which does not focus on any specific drug groups is by Blouin and Warren (1999); as a bonus, it is available for free from jpharmsci.org.
The college answer to Question 13 from the second paper of 2012 is actually an example of a well-written model answer, offering useful information about the college's specific expectations in this topic while remaining thankfully devoid of unhelpful criticism or examiner whinge. The answer even offers something of a structure for good responses, hinting that "candidates that took this approach generally did better". The college did point out that many of the candidates did not make any attempt to give examples, and the summary below attempts to address this problem. As with critical illness, major effects of obesity, their mechanisms and direction of change with examples are offered wherever possible in the mechanism = effect (example) template.
- Increased gastric emptying rate = higher oral peak dose (variable and drug-dependent)
- Decreased gastric emptying rate due to bariatric surgery = lower oral peak dose (cyclosporine, thyroxine, phenytoin and rifampicin)
- Poor subcutaneous fat circulation = decreased subcutaneous absorption (hCG)
- Difficult intramuscular access = inadvertant subcutaneous injection
- Increased absolute and proportional amount of body fat = increased Vd for lipophilic drugs (benzodiazepines, lignocaine, thiopentone, verapamil)
- Increased total body water = increased Vd for hydrophilic drugs (amikacin, gentamicin and tobramicin)
- Increased α1-acid glycoprotein levels = decreased free fraction of some drugs (eg. propanolol)
- Increased hepatic blood flow due to increased cardiac output = increased clearance of high extraction ratio drugs (propofol)
- Decreased hepatic blood flow due to fatty liver = decreased hepatic clearance (clozapine, haloperidol)
- Increased Phase II enzyme activity = increased hepatic clearance (lorazepam, oxazepam, paracetamol)
- Increased soluble enzyme activity = increased substrate drug dose requirements (suxamethonium)
- Increased half-life due to increased volume of distribution for extremely lipophilic drugs (desmethyldiazepam, midazolam)
- Increased cardiac output = increased GFR, increased renal clearance of hydrophilic drugs (vancomycin, aminoglycosides)
- Increased tubular secretion = increased clearance out of proportion to increased GFR (ciprofloxacin, cimetidine and procainamide)
- Decreased GFR due to diabetic nephropathy = reversal of the normal obesity-associated increase in clearance (vancomycin, aminoglycosides, ciprofoloxacin, etc)
- Decreased dose response
- Resistance to haemodynamic effects of verapamil
- Resistance to the effects of oral contraceptives
- Resistance to the effects of atracurium
- Resistance to the effects of insulin
Exaggerated dose response
- Increased sensitivity to triazolam
- Increased sensitivity to the respiratory depressant effects of opiates in morbid obesity with sleep apnoea
Oral absorption is variably affected by obesity. Usually, for the better. In fact, good oral absorption is how they got to be obese in the first place, and this is reflected by the well-documented finding that non-diabetic obese individuals have a higher gastric emptying rate than non-obese individuals (Verdich et al, 1999 and Wright et al, 1983). Those were studies of food, however. For instance, Wright's team fed their subjects Technetium-labeled chicken liver. That's not exactly analogous to a couple of tablets. And it is not a major stretch of the imagination to consider that obese patients are frequently diabetic, and diabetic patients frequently have enough autonomic dysfunction to have delayed gastric emptying. Ergo, the effects of obesity on gastric emptying are variable.
Padwal et al (2009) reviewed this topic and found that there is quite a lot of variability in the way drug absorption is affected by bariatric surgery, depending on what kind of drug and what kind of surgery. For instance, drugs which depend on gastric pH for their disintegration will typically have their availability decreased by various bypass procedures and increased by various restrictive procedures (i.e. ones which limit gastric emptying rate). As most bariatric surgery today is of the restrictive variety and gastric emptying being the most important determinant of absorption rate, one can generally expect post-op bariatric surgery patients to have slowed onset of orally administered medications, and likely a diminished peak effect. Examples of drugs which consistently demonstrate diminished absorption following bariatric surgery include cyclosporin, thyroxine, phenytoin and rifampicin.
The subcutaneous fat of obese individuals expands in volume without significantly increased vascularity. Drug molecules injected into this forgotten space will be lost among the adipocytes. This is probably even more relevant among ICU patients whose circulatory function is compromised. Even in the relatively well obese population of women who were trying to get their oocytes to mature for IVF, the absorption of subcutaneously injected hCG was much slower (Chan et al, 2003).
Generally speaking, there is nothing physically different about the muscles of obese patients which might impair absorption from these sites. The problem is more pragmatic than that. The subcutaneous fat layer being so thick, it is often impossible to guarantee a good intramuscular injection. The injection ends up being subcutaneous.
Blood flow in fat is poor in people of normal weight: it is only about 5% of the total cardiac output (Cheymol, 2000). In obese individuals, blood flow to fat is even poorer. Obese individuals are also likely to have a degree of heart failure which further decreases blood flow. This makes their fat a large compartment of potential distribution for lipophilic drugs which fills gradually, and then becomes a slowly emptying reservoir. Abernethy and Greenblatt (1986) documented several substances for which this is a serious concern. Of note were substantial increases in the apparent volume of distribution for "most benzodiazepines, thiopentone, phenytoin, verapamil and lignocaine".
The absolute volume of body fluid may not change very much in the obese individual, but the long term upshot of having a large difficult body is probably going to be a state of cardiovascular unhealth which will be interpreted as a relative hypovolaemic state by the renin-angiotensin-aldosterone system, with retention of salt and water being the ultimate consequence. Which is a verbose way of saying that obese people are often oedematous. This hidden pool of water is demonstrated in the distribution characterstics of drugs which are usually confined to the aqueous partition; for example, in 1983 Bauer et al were able to demonstrate an increased volume of distribution associated for amikacin, gentamicin and tobramicin.
This is a somewhat controversial point, as some authors have found increased protein binding due to increased levels α1-acid glycoprotein, whereas others did not reproduce this finding. Hypothetically, if it were true- what of it? Well. If anybody were able to replicate the experimental results of Benedek et al (1983), their patients would have lower free levels of propanolol.
Most drugs are metabolised by the liver, and in most obese patients the liver is fatty and dysfunctional. Ergo, drug metabolism is affected adversely. The specific dysfunction appears to be related to poorer hepatic blood flow. According to Ijaz et al (2003), hepatic sinusoid space can be reduced by up to 50%, or occluded entirely. The liver becomes literally clogged with fat. Merrell et al (2011) reports increased toxicity due to clozapine and haloperidol, attributed to this phenomenon.
As the obese individual is likely obese due to unregulated nutritional substrate intake, so the liver becomes the unwilling recipient of vast amounts of foods and non-toxic food-like substances. In response to this, some adaptive remodeling of enzymatic activity surely must take place. Indeed, Abernethy et al (1983) found an increase in the rate of glucouronidation in obese subjects, which affected the clearance of lorazepam oxazepam and paracetamol. Specifically, the clearance rate doubled or tripled. The volume of distribution also increased, however; which means that the total half life of each drug remained roughly unchanged. This is not the case for all benzodiazepines - some are so lipophilic that the increase in their hepatic metabolism cannot adjust enough, and half life is increased (see below).
Large amounts of fatty tissue are heavy, and require a large amount of muscle to mobilise. This mass of fat and muscle therefore calls for a robust circulatory system, at least during the initial honeymoon period (later, congestive cardiac failure and diabetic cardiovascular complications will lay to waste this early advantage). Anyway, this increased cardiac output requires higher doses of hypnotic agents (Lemmens, 2010). The increase is correlated with the estimated lean body weight rather than total body weight, and is most relevant for high extraction ratio drugs like propofol (where cardiac output is inversely related to peak concentration and effect duration).
Not only is hepatic metabolic function altered in obesity, but also the synthetic function. By logical extension, the synthesis of soluble enzymes increases. A clinically relevant example of this phenomenon is increased plasma pseudoesterase activity seen in obesity. Bentley et al (1982) found an increased requirement for suxamethonium in morbidly obese subjects which was out of proportion to their predicted lean body weight, and in proportion to their total body weight. Moreover, tissues other than the liver can metabolise drugs. Human adipocytes contain glutathione transhydrogenase which can cleave insulin, contributing to insulin resistance and offsetting the hyperinsulinaemia associated with obesity (Rafecas et al, 1995).
It stands to reason that drugs which distribute extensively into fat would have their half-lives extended by the addition of extra fat into which to distribute. Indeed, this is seen with particularly lipophilic drugs like benzodiazepines. For some (as mentioned above) the hepatic clearance rate increases as well, which keeps the half life relatively stable over a range of BMIs. For others the clearance rate actually decreases because of extensive distribution into fat. This is seen with desmethyldiazepam (Abernethy et al, 1982) and midazolam (Greenblatt et al, 1984).
Obese individuals have an increased amount of lean body mass, which makes sense considering the mechanical requirements of transporting large amounts of adipose tissue from place to place. This gives rise to an increased cardiac output and therefore increased renal blood flow. GFR increases proportionately to lean body mass (Janmahasatian et al, 2015) and normalises the "apparent" decrease in GFR which would otherwise be seen if the GFR was calculated from total body weight. This has an effect on drugs which are predominantly cleared renally, for example vancomycin (Bauer, 1998) and the aminoglycosides (Bauer, 1983).
Some drugs which are cleared by the kidneys appear to be eliminated more rapidly in obesity than would be explained purely by the increased GFR. Ergo, active tubular secretion must also be increased. This is seen in the case of ciprofloxacin, cimetidine and procainamide (Blouin et al, 1999). Why does this happen? Nobody seems to offer a solid hypothesis. Perhaps there is some merit to increasing the active tubular secretion of organic acids (urate, etc) which are generated in the course of tissue metabolism if there is significantly more tissue around.
Clearly, renally cleared substances will have slowed clearance and increased half-lives if the kidneys are somehow damaged. For instance, by decades of poorly controlled diabetes. This is not exactly a change unique to obesity (plenty of skinny people out there with diseased kidneys) and less time will be spent on it here. It is perhaps worth mentioning in the exam, particularly as it is included in the college's own SAQ answers (eg. "renal clearance may be impaired in renal disease caused by obesity related diseases, e.g. diabetes" and the mention of "co existing disease processes eg diabetes" in the Fellowship paper). LITFL also list it in their pharmacokinetics of obesity chapter.
Sankaralingam (2015) reviewed the literature for drugs which are used to control cardiovascular risk factors, and found that the pharmacodynamic effects of these drugs are changed unpredictably, in any direction. For the majority of drugs, there are no pharmacodynamic efects. Weinstein et al (1988) dosed obese patients with vecuronium and atracurium, and found no differences in drug effect which would not be well explained by differences in clearance. Dunn et al (1991) found no differences in the effects of methylprednisolone among obese and nonobese male volunteers. Wang et al (2002) was not able to find any differences in the haemodynamic response to nitroglycerine among fat and skinny rats. Sanderlink et al (2002) explored subcutaneous enoxaparin in obesity - there were no pharmacodynamic effects. In short, there is little evidence that obesity has major effects on the dose-response curves of most drugs, provided pharmacokinetic changes are controlled for.
So, where are the differences in pharmacodynamics due to obesity? The candidate needs to write something in the exam. Question 13 asked for pharmacology, not just pharmacokinetics. With effort and time the author was only able to find few examples, which were painstakingly mined out of the Pubmed mountainside.
Decreased dose response
Exaggerated dose response