Oral thyroxine and IV triiodothyronine in critical care

This chapter is related to Section U2(iv) from the 2017 CICM Primary Syllabus, which expects the exam candidate to "understand the pharmacology of thyroid hormones", even though that has never been included in exam questions. Past papers have only included the following:

• Question 17 from the first paper of 2016 (physiology of thyroid hormones)
• Question 3(p.2) from the second paper of 2008 (hyperthyroidism)

However, thyroid material is highly prevalent in the CICM Second Part exam, and so there appears to be a gulf of sorts, where the Intensive Care trainee is expected to become confident with the use of these substances somewhere in the middle of their training. To bridge that gap, some kind of vague non-peer-reviewed FOAM blather seems in order, which is the right way to regard what follows. 

Chemical structure and chemical relatives of T3 and T4

To classify thyroxine and triiodothyronine as "hormones" would be lazy, but a completely legitimate shortcut for the CICM trainee trying to form an opening statement to a hypothetical SAQ that asks "compare the pharmacology of T3 and T4". To call them "iodinated tyrosine derivatives" would be more accurate and specific. The PubChem entry actually contains a whole host of accurate descriptive terms which could be used as alternatives, including "iodothyronine", "halophenol", and "non-proteinogenic alpha-amino acid". Their less-iodinated cousins and their amino acid precursor are listed below mainly so that we can leave them here and never return to discussing them. 

To achieve maximum marks, the savvy candidate would quip about how storebought off-the-shelf thyroxine is actually the levo-stereoisomer, which is the same isomerism as the T4 produced endogenously, making it completely identical to the natural hormone. On the other hand, dextrothyroxine is not produced in the thyroid, and exists in the weird liminal space between patents, where nobody has found a safe pharmacological use for it (though it was at one stage used unsuccessfully to treat hyperlipidaemia).  A list of the chemical relatives of thyroxine and triiodothyronine would probably have to include this dextroisomer, as well as rT3, the reverse metabolically inactive form of T3 which ends up being produced in excess during states of critical illness, giving rise to the sick euthyroid syndrome.

Pharmaceutical presentation of T3 and T4

Levothyroxine, or L-3,5,5'-tetraiodothyronine if you prefer, is the most common form of thyroid hormone replacement, prescribed usually as tablets. There is nothing wrong with triiodothyronine and it could easily be used as oral thyroid replacement instead, but the shorter half-life makes it inconvenient. Conversely, though nothing prevents it from being made into an intravenous formulation, it would make no logical sense to infuse people with T4 intravenously, it being basically a prohormone, and the active form being already readily available as infusion. In this fashion, the two forms of thyroid replacement therapy have become polarised, with T4 made into oral formulations and T3 made into (surprisingly expensive) infusions.   

The only other pharmaceutical thing that needs to be mentioned about thyroxine tablets is a requirement for storage at something below room temperature. Specifically, Australian Prescriber recommends you keep it at 8-15 °C, which is roughly the temperature of the dairy compartment in your first-world bourgeoise fridge. Considering the broadly everything, no Australian home should be expected to have a cool dry place by 2050, which makes this even more important - improper storage has been implicated as the cause of up to 5% of the cases of poor response to chronic treatment. Similarly, triiodothyronine solutions are inherently unstable, and can only last for about 30 days, even when stored at 2–8 °C.

Pharmacokinetics of T3 and T4

What follows is a summary of Colucci et al (2013) and Sypniewski (1993). Whole verdant fields of literature flourish in the land of oral thyroid replacement, as hypothyroidism is rife and oral levothyroxine is ubiquitous. In contrast, there is virtually nothing out there for the IV formulation. This online PI document for the sodium salt of liothyronine from the edgily named "X-GEN Pharmaceuticals" is apparently the only available information for the excipients and composition. 

Administration and absorption

It is probably important to mention that the most common method of administration for these hormones is "secreted from the thyroid". 80% of all T4 and 20% of T3 is produced in the thyroid gland under normal circumstances.

After oral administration both T3 and T4 are well absorbed in the small intestine, with the caveat that it should ideally be empty. These drugs tend to interact with basically everything in your diet (calcium, iron, coffee, milk, fibre, vitamin C, soy, fruit, what have you), which means usually NG feeds need to be paused for some hour or so prior and following the administration of thyroxine. Unhelpfully, they also tend to adsorb onto the surface of nasogastric tubes. 

Distribution and protein binding

Both thyroxine and triiodothyronine have very poor water solubility. Commercially available IV preparations of triiodothyronine get around this by using plenty of alcohol and ammonium hydroxide to solubilise the active ingredient, making it necessary to infuse it via central lines (as it would have an irritant effect otherwise). Fortunately, the human circulation contains thyroxine-binding globulin, a dedicated chaperone protein, and thyroid hormones circulate almost exclusively in the company of this huge 54-kDa molecule. That makes the volume of distribution of T4 only slightly higher than the circulating volume, 11-14L  or about 0.15L/kg. T4 and T3 have very different affinity for this globulin (T3 has 30 times lower affinity) which means the VOD of T3 is much larger, around 1.8L (with most of it ending up in cells). There is also some binding to albumin, and some to transthyretin (which also transports retinol).

Loading dose

On one hand, the small volume of distribution and the delayed mainly genetic-transcription-related effect of thyroid hormones would seem to suggest that a loading dose is not necessary. On the other, the literature on thyroid hormone supplementation, and particularly the literature of the use of emergency thyroid hormone replacement in critically ill patients, seems to uniformly recommend loading doses for bit T4 and T3. For example, the American Thyroid Association
Task Force on Thyroid Hormone Replacement (Jonklaas et al, 2014) recommended loading with 200–400 mcg of levothyroxine or 5–20 mcg of triiodothyronine for myxoedema coma.

But why, one might ask? The rationale for this practice is somewhat obscure, and comes from the distant past. It can be ultimately traced back to Holvey et al (1964), who arrived at their practice by the following series of logical steps:

• The normal extrathyroid iodide content of an euthyorid person is 490 mcg/m² BSA.
• Holvey et al had a bunch of myxoedema patients whose iodide stores were only 245 mcg/m² BSA (exactly half)
• Thus, there was a total iodide deficit of 245 mcg/m²
• Assuming that iodide comprises 65.4% of levothyroxine by weight, and assuming that thyroid hormones are the main form of organic iodide in the circulation, this means these people had a thyroid hormone deficit of about 360 mcg.
• The rationale of administering this deficit as one huge dose was to immediately restore the circulating thyroid hormone pool, which also has the effect of replacing iodine, so that it may continue to act as the substrate for more thyroid hormone synthesis.

Metabolism and elimination

Thyroxine is deiodinated in the circulation by the unimaginatively named deiodinase enzymes, which are widely distributed throughout the body tissues (liver, kidneys, thyroid, pituitary, CNS, brown adipose tissue, skeletal muscle, and heart). These selenoproteins are only really able to deiodinate the free dissolved fraction of the hormones, which means only about 10% of the circulating T4 ends up metabolised per day, giving it a half life of something like 7.5 days in hypothyroid subjects. This activity produces T3 which is pharmacologically active as well as a series of "reverse" pseudohormones that can still bind the receptor but which have no beneficial activity (eg. rT3). Fortunately under normal circumstances T3 is the dominant product. The deiodination (and other metabolic breakdown pathways) of T3 are much faster, mainly because it is less protein bound, and therefore has a half-life of about 6-10 hours. For both hormones, a small fraction undergoes hepatic metabolism by glucuronidation,  and this fraction appears to increase in overdose (as protein binding sites are saturated), resulting in some significant enterohepatic recirculation (Galton & Nisula, 1972).

Physiological effect of thyroid hormones

This pharmacodynamics section is a summary of the chapter dealing with thyroid hormone physiology.

Molecular effects of thyroid hormones

Thyroid hormones exert the majority of their physiological effects by binding to the thyroid hormone receptor. Zhang & Lazar (2000) explain the cellular mechanism their action over thirty pages, but the main points can be distilled as follows:

• T3 is the main actor here, as the affinity of T4 for this receptor is very poor
• Thyroid hormone receptors are a nuclear receptor superfamily, similar to the steroid receptors (the family also includes receptors for retinoids, Vitamin D, prostaglandins, androgens, and so on).
• There are multiple receptor isoforms (TRα1, TRα2, TRβ1, etc)
• They are all transcription factors, i.e. when the hormone binds to them, the receptor undergoes a conformational change which leads to the transcription of genes, and a change in the protein synthesis
• Thus, T3 needs to penetrate the cell and enter the cytoplasm, which would not work through passive diffusion as the distance is too great (and the concentration of T3 is too small). Ergo, active transport is required, and Henneman et al (2001) identify a whole range of cell membrane carrier proteins that may be responsible for this, which should give the reader the correct impression that nobody really knows what exactly happens here.
• Apart from genomic effects, thyroid hormones have various other pathways by which they exert their effects, involving the binding of non-nuclear cytoplasmic receptors. According to Sinha & Yen (2018), these can produce an increase in cGMP and nitric oxide, modify mitochondrial activity, and stimulate the availability of intracellular calcium. A lot of the clinically important effects of thyroid hormones are attributed to this "non-canonical" pathway which does not require gene transcription.

Physiological and clinical effects of thyroid hormones

The effect of thyroid hormones on visible body functions can be summarised as "more everything". An excellent (because short) table in  Sypniewski (1993) can be reproduced here with basically zero modification, as it distils this subject very efficiently:

• Airway effects
  • Increased airway reactivity resulting in bronchospasm
  • Skeletal muscle myopathy leading to reduced upper airway muscle tone
  • Goitre-related effects (lower airway displacement or compression)
• Respiratory effects
  • Increased respiratory demand (increased CO₂ production because of increased metabolic rate)
  • Increased airway resistance
• Cardiovascular effects
  • Increased contractility
  • Decreased peripheral vascular resistance
  • Thus, increased cardiac output
  • Increased myocardial oxygen consumption
• Neurological effects
  • Mainly excitatory effects (insomnia, anxiety, seizures)
  • Neurodevelopmental effects
  • Increased 
• Renal effects
  • Increased renal blood flow
  • Increased clearance rate of renally cleared substances
  • Increased synthesis of vasoactive mediators
• Hepatic effects
  • Increased rate of protein synthesis
  • Increased hepatic blood flow
  • Increased clearance rate of substances metabolised by the liver
• Gastrointestinal effects
  • Increased gastrointestinal motility
  • Increased appetite
• Metabolic effects
  • Increased carbohydrate metabolism (increased hepatic glucose output, increased gluconeogenesis, increased use of carbohydrate as fuel source)
  • Increased fat metabolism (increased peripheral lipolysis, increased hepatic lipid release, increase in free fatty acids and increased rate of lipid use as metabolic fuel)
  • Increased protein turnover (both hepatic synthesis and peripheral catabolism)
• Thermoregulatory effects
  • Increased shivering and nonshivering thermogenesis
  • Mainly by decreasing the efficiency of mitochondrial fuel metabolism via the uncoupling of electron transport (the major reason for increased carbohydrate and fat metabolism)
• Musculoskeletal effects
  • Increased blood flow to skeletal muscle
  • Increased contractility and oxygen consumption by skeletal muscle
  • Increased muscle protein synthesis (in great excess, increased protein catabolism and myopathy)
  • Increased muscle tone and tremor

Cardiovascular effects of thyroid hormones

The intensivists' attention is turned hard to thyroid hormone supplementation in several specific scenarios, which mainly concern patients at the ragged edge of survival. We are talking about:

• Severe hypothyroidism and myxoedema coma (where IV thyroid hormone replacement becomes necessary because of complete cardiovascular collapse), 
• Situations where the thyroid hormone is replaced intravenously because of lost pituitary regulation (eg. brain death), again often related to terrible haemodynamic problems
• Situations where the patient has cardiogenic shock and every other class of inotrope has already been tried.

The reader will now realise that the cardiovascular effects of thyroid hormones are therefore of the greatest interest to the intensivist, in the same sense that the neurodevelopmental effects probably are not. For this reason, we will go into a little extra depth here. The specific subject is not "what do thyroid hormones do to maintain the function of the cardiovascular system in normal physiological circumstances", but "how does a physiologically unrealistic dose of exogenous thyroid hormone do anything helpful when the patient's circulatory system is disassembling itself in front of me". 

Well, in short, triiodothyronine is a potent inodilator. 

• Increased contractility: Thyroid hormones have potent direct inotropic effects which are not merely related to the reversal of some kind of low grade subclinical hypothyroidism. Malik et al (1999) gave 20 mcg/hr of IV thyroxine to a bunch of euthyroid cardiogenic shock patients and found their haemodynamic parameters improved substantially (for example, the CI increased from 1.75 to 2.44 over 36 hours). The effect appears to have some kind of genetic transcription component, in the sense that it takes a while to develop and only becomes clinically significant over the course of about 24 hours. According to Jabbar et al (2017), the molecular effects are numerous, and include the increased expression of Na⁺/K⁺ ATPase,  SERCA2, and whole chunks of the actual contractile apparatus such as myosin heavy chain­ α. On top of that, the expression of the β1 ­adrenergic receptor is increased. In short, apart from their own direct inotropic effects, thyroid hormones increase the sensitivity of the myocardium to the positive inotropic effects of catecholamines and digoxin.
• Left ventricular afterload reduction: Triiodothyronine is an arteriodilator with predominantly systemic effects.  Ojamaa et al (1993) reported on some haemodynamic variables in patients with hyperthyroidism and found their SVRI was roughly half of what one might expect. Moreover, the effect is rapid, not unlike the effect of GTN or nitroprusside, suggesting some direct activity (i.e. the change in vessel properties is faster then what one might expect if it were solely due to some kind of gene transcription effect. The precise molecular mechanism remains to be established. Ozmak-Tizon et al (2014), in their paper on the nongenomic effects of thyroid hormones, theorised that some mechanism involving nitric oxide synthesis must be involved. 
• Right ventricular afterload increase:  Thyroid hormones appear to increase pulmonary vascular resistance, and hyperthyroidism is associated with a rather high prevalence of pulmonary hypertension (Sugiura et al, 2014). Specifically the effect seems to be related to the increase in hypoxic pulmonary vasoconstriction (Herget et al, 1987). From this, it follows that one would want to maintain normoxia in their patient if they wanted to avoid this side effect during the use of thyroxine in the haemodynamically fragile right heart failure patient. 
• No real effect on preload. Right atrial pressures remained unchanged in the case series by Malik et al (1999), suggesting that thyroid hormones do not have any direct venoconstrictor or venodilator effects. 
• No real effect on heart rate. Though tachycardia and rapid AF are often associated with presentations of hyperthyroidism, acute use in cardiogenic shock does not seem to be associated with much of an increase in the heart rate  - though, to be fair, in Malik et al (1999), all of the patients were running at rates of 115-120 anyway.

Overdose and excess

All good pharmacology monographs need to end with the discussion of adverse effects and overdoses. For thyroid hormones, the undesirable effects are largely the extensions of the desirable ones - for example, maintenance of body temperature homeostasis turns to hyperthermia, control of cardiac catecholamine sensitivity turns to tachyarrhythmia and hypertension, and so forth. The metabolic rate increases, generating pointless energy expenditure and heat, and the appetite increases to accommodate the increased metabolic fuel requirements (which means abuse of thyroxine to achieve weight loss could be pointless).

But, these are cases of slightly misinformed people accessing thyroxine illegally to take small doses regularly over weeks or months, eventually presenting to ED with palpitations and diarrhoea. For the critical care practitioner the more interesting scenarios are thyroid hormone overdoses which are several orders of magnitude in excess of their normal dose, and which present with life-threatening clinical features. These are fortunately very rare, and known mainly from case reports. For example, Khan et al (2019) report a case of severe thyrotoxicosis arising from a 1000-fold drug concentration error, where the poor patient continued taking a daily dose of about 15mg of thyroxine, and eventually presented with what looked like encephalitis and sepsis. Another case report (Allen et al, 2015) presented following an intentional self-poisoning with 15,800 mcg of levothyroxine, also with an altered level of consciousness and tachycardia. For those sorts of situations, in addition to the usual therapies for thyroid storm, the reader is offered some toxicological strategies which were described in the literature as being effective:

• Cholestyramine, to prevent enterohepatic recirculation of the thyroxine glucouronides
• Dexamethasone, to prevent the peripheral conversion of T4 into T3
• Plasma exchange, to remove the THBG-bound hormone molecules together with their chaperone protein