Speed of onset of neuromuscular blockade

This chapter addresses Section L2(i) from the 2017 CICM Primary Syllabus, which expects the exam candidates to have an "understanding of the pharmacology of neuromuscular blocking drugs".  There is  reasonably large fraction of the questions on this topic have asked about the factors that influence the speed of onset of NMJ blockers:

• Question 12 from the first paper of 2016
• Question 9 from the first paper of 2014
• Question 3 from the second paper of 2009

This is concerning, as basically none of the "official" textbooks discuss this topic to any great level of detail. Even worse, wherever one finds the information in some official resource, it is never explained - they just leave it there with the expectation that critical care trainees are desperate enough to memorise it unquestioningly, and exhausted enough to have lost all curiosity about the background. One is forced to reconstruct this body of knowledge by exposing fossilised answers in the compressed silt of ancient academic deposits. Fortunately, these CICM questions seem to be a copypasta of an ANZCA primary exam question from 2004, and we are all grateful to ketaminenightmares.com for their extensive catalogue of these, complete with excellent model answers. For all intents and purposes what follows is a series of detailed footnotes to the summary offered by Stuart N Watson et al, to whom my hat is forever tipped, and on whose structure the following page is extensively reliant.

 The "official" source for the 2009 SAQ seems to be Foundations of Anaesthesia: Basic Clinical Science, by Hemmings and Hopkins (page 453 of the 2nd edition, 2005). For those unwilling to pay for this book only to answer a single set of SAQs, Kim et al (2017) is a reasonable overview and contains about 60% of the relevant material for free. Unfortunately, no single article carries everything required, and what follows is an attempt to amalgamate the work presented by a fairly large range of resources into a single whole, while preserving some of the scientific curiosity of original papers. As the author is neither an expert on the neuromuscular junction nor on copyediting, the result is probably less coherent and more difficult to digest than the corresponding page from Part One, and the time-poor exam candidate is redirected there instead.

Rate of NMJ blocker delivery to the muscle

The rate at which the drug arrives at the organ of interest is obviously going to play a role in the speed of onset, and the main factors that influence this are the route of administration and the rate of blood flow, meaning both globally (cardiac output) and regionally (where some muscles may differ).

• Route of administration is an important factor, considering everything we know about the intramuscular absorption of these agents.  It is sufficiently problematic that some authors (eg. Shaw et al, 2015)  argue it is not fast enough to achieve optimal intubating conditions for a patient with marginal respiratory function. 
• Site of IV administration is also important, for stupid cardiovascular reasons (in the sense that the rate of blood flow returning the drug to the heart is an important factor in the onset time). The closer to the left side of the heart your injection site, the faster the onset of block. Taking this to a logical but preposterous extreme, Iwasaki et al (1994) were able to demonstrate that the onset of block was fastest when the vecuronium was injected into the pulmonary artery (58 seconds), followed by right atrium (75 seconds) and peripheral vein (82 seconds). 
• Rate of IV administration, i.e a fast bolus or a slow bolus, will obviously influence the rate of onset. This being intuitively obvious, one might not expected for anybody to actually test this experimentally. However, because many critical care speciality programs around the world have the expectation their trainees will pollute the world literature with pointless research projects, one might have expected this low-hanging fruit to get picked at some stage or another, and indeed in 2018 Kulkarni et al compared the onset of block with a rocuronium bolus being pushed through with a 20ml saline flush, as compared to just letting it drip in via a gravity-fed IV infusion set. Predictably, the bolus produced effects in literally half the time (55 seconds vs 110 seconds), to a publishable level of statistical significance.
• Cardiac output and muscle blood flow need to be mentioned as essential elements, and hyperdynamic circulatory states such as sepsis or pregnancy can be given as examples of where the onset of the drug effect is hastened. By how much? This probably varies from patient to patient. For Baraka et al (1992), the onset of vecuronium block was about 64 seconds sooner in patients undergoing Caesarian section, versus a similar but nongravid cohort.
• Position of the muscle in relation to the cardiovascular system is, weirdly, important - because the blood hits proximal muscles first, as they receive the rocuronium-rich blood before the distal muscles. It might seem ridiculous, but in fact blood really does not arrive everywhere at the same time, and there is quite a bit of difference in the rate of onset of blockade depending on how far from the heart you are. As an example,  Pansard et al (1987) had determined that the adductor pollicis brevis becomes paralysed a full 30 seconds later than the diaphragm after a dose of suxamethonium.

Rate of equilibation between plasma and the NMJ

Rate of drug transfer from plasma to the effect site, or rate of effect site equilibration is the broad term you'd give to this group of factors. This is directly related to the rate of block onset, which makes logical sense (as the agents mostly have their effect immediately upon their arrival to the neuromuscular junction, and there are no weird secondary messenger games to delay their effect). The rate of effect site equilibration is affected by multiple factors, which, if you think about it, are largely Fickian in their character. These factors are:

• The molar concentration of the drug, which is affected by:
  • Protein binding
  • Solubility
  • Potency
• The diffusion distance (eg. from the capillary to the motor endplate)
• Surface area of the membrane
• Diffusion coefficient, which is in turn influenced by:
  • Temperature of the solution
  • Viscosity of the fluids involved
  • Drug molecule size (and molecule shape)
  • Solubility of the drug, which itself has many influences, but chiefly:
    • pKa of the drug
    • pH of the solution

These Fickian factors are often mentioned in exam answers to these speed-of-onset questions, but seriously, reader - though these factors do truly influence the rate of effect site diffusion, surely they do not differ over much from subject to subject? The patient will usually be getting the paralysis toxin when they are normothermic, and their capillary membrane surface area, pH, plasma protein content, and the distance between their capillaries and their muscle will not fluctuate wildly even among the biochemically erratic patients of the ICU. In short, the most important Fickian factor here is actually the molar concentration of the drug, which is functionally related to its potency.

Potency and rate of onset of neuromuscular junction blockade

The potency of the agent is often said to be inversely related to the speed of onset. It is well demonstrated by the intentional crippling of the vecuronium molecule, which produced rocuronium - a drug with a fraction of the potency, but a much faster onset. Rocuronium is 5-8 times less potent than vecuronium because of several molecular modifications, and has a 50% faster onset time. The explanation given for this by textbooks is that there are more molecules of rocuronium administered, and so the higher concentration gradient increases the rate of drug delivery to the synapse, where all agents will act fairly similarly (i.e. immediately). 
It is therefore probably not correct to say that the potency of the drug is the main influence, but rather that the dose of the drug plays the greatest role. As an example of this, a drug which a high potency can still be forced to have a fast onset of effect when it is given in a higher dose. Observe, vecuronium:

This is occasionally referred to as "the Bowman principle", or "Bowman's principle", a search term that yields mainly Australasian and British critical care exam resources. It appears to be a concept only known by this name in the Commonwealth, referred to by other terms elsewhere. It is occasionally offered as a broad pharmacokinetic principle that dictates that the molar concentration of a drug is directly related to the speed of onset because of the abovementioned diffusion-related effects. The name appears to originate from a 1988 paper by Bowman et al, where the authors considered the reciprocal relationship between drug potency and speed of onset, concluding that "a nondepolarizing equivalent of suxamethonium, when discovered, may necessarily be a drug of relatively low potency" and thereby predicting the development of rocuronium and rapacuronium.

The concept was tested and supported empirically by  Kopman et al (1999), who compared the speed of onset of a selection of agents to their molar mass, and produced this table:

All this, of course, relies on the premise that, immediately as they arrive at the neuromuscular junction, these molecules will enthusiastically descend upon the nicotinic receptors, and immediately produce their pharmacodynamic effect. That interaction is not necessarily as smooth as presented in this oversimplified model. There are in fact a series of factors that can affect this significantly, delaying or hastening the onset of block by altering the concentration of agent required to produce clinically meaningful muscle relaxation. As follows:

Concentration required to produce NMJ block

Factors that influence the effect site concentration required to produce block can be summarised as "things that interfere with neuromuscular transmission more broadly" that also happen to accidentally influence the activity of neuromuscular junction blockers in one direction or another. These are numerous, and moreover they tend to differ according to which type of agent you plan on using, with some factors having a retardant effect on the speed of onset of nondepolarising agents while potentiating the effects of the depolarising kind, or vice versa. From the perspective of passing exams, the ability to understand them all is not as important as the ability to list them all, which means the brief summary offered in the grey box is entirely sufficient, and what follows is entirely superfluous. It is offered here mainly to satisfy those irrationally idealistic readers that still maintain some curiosity about their specialist training. 

Factors that slow the onset of NMJ block by increasing the necessary dose of neuromuscular junction blocker

Factors that increase the concentration required to produce block are mostly factors that increase either the amount of available acetylcholine or the number of nicotinic receptors, thus making it more difficult for the agent to antagonise and occupy 70-80% of them. There's an excellent article by Jung & An (2018) that describes "resistance" to NMJ blockers in a broader sense, and it includes some of the following:

• Acetylcholinesterase inhibitors, obviously (duh). Increased amounts of acetylcholine will reduce the rate of onset of nondepolarising NMJ blockers, and reduce their potency. Yes, one typically gives these agents to reverse the block, which makes this a desirable property; but sometimes one may encounter patients out in the wild who are already on an acetylcholinesterase inhibitor. These may include patients with myasthenia gravis (who chronically take pyridostigmine) or patients who have been poisoned by organophosphates. 
• Critical illness polyneuromyopathy, burns, tetanus, spinal injury, or stroke: these conditions in one way or another denervate the muscle, and produce an uncontrolled proliferation of nicotinic receptors. A conventional amount of blocker will therefore be too little to create the 70-80% receptor occupancy. This can give rise to a situation where the patient will be profoundly weak, but still appears to be resistant to the effects of the blocker; and the apparent resistance tends to increase over time during prolonged infusions. An excellent example of this can be seen in Yuan et al (2022), who used sustained infusions of cisatracurium in patients with ARDS, and found that the dose needed to increase by a factor of four over thirty days:
  
  At the same time, the effect of this on depolarising blockers should probably be the opposite. More receptors, means more channel pores to open, and therefore a faster onset - or so one might expect. In practice, however, this is rarely seen, because the patient usually dies of a devastating hyperkalemic crisis, confounding any study looking at the onset of NMJ blockade. 
• Antiepileptic agents also increase the number of receptors, but influence the potency and rate of onset in a number of other ways, some of which are simple and pharmacokinetic (Soriano & Martyn, 2004, discuss things like displacement from plasma protein binding sites and interference with hepatic metabolism).  
•  Factors that hyperpolarise the myocyte membrane will obviously make it more difficult to depolarise it. There are several of these, and they mostly fall into a category of "things which, if you have them, are much bigger problems than a loss of NMJ blocker sensitivity". They include:
  • Hyperkalemia (though this could go both ways, as below; i.e. the effect is different depending on whether you are using a depolarising or a nondepolarising agent)
  • Hypercalcemia: in an excellent study by Okamoto (1992), increasing the concentration of calcium antagonised the activity of neuromuscular junction blockers. How high must the calcium level be? Impossible to say, as the paper is in Japanese.
  • Malignant hyperthermia is occasionally listed as one of the factors that increases the ED₉₅ of neuromuscular junction blockers (presumably nondepolarising ones), slowing the onset of effect and decreasing their potency. This seems logical, as the muscle rigidity produced by malignant hyperthermia is entirely post-junctional in its mechanism, and indeed entirely divorced from the activities of the junction. On this basis we can speculate that neuromuscular junction blockers would have no effect on this downstream rigidity. This seems logical, but hardly bears repeating in polite company, as the only evidence to describe this point of our understanding comes from a measurement in three dead pigs by Hall et al (1976). 

Factors that hasten the onset of NMJ block by decreasing the necessary dose of neuromuscular junction blocker

• Factors that decrease the amount of acetylcholine at the junction:
  • Immature (eg. neonatal or foetal) receptors decrease the sensitivity of the neuromuscular junction, which mainly affects nondepolarising agents. Specifically, human neonatal nicotinic receptors first appear in a "foetal" form, which is more sensitive to acetylcholine and which stays open for longer. The main reason is thought to be a reduced availability of acetylcholine at the sites - Meakin (2007) reports that neonates can only muster something like a third of the adult concentration. The result is a sensitivity to nondepolarising agents (which require about 50% of the normal dose); as there is less acetylcholine around to compete with.  Curiously, neonates and young infants also seem to have a resistance to suxamethonium (where neonates seem to require about 3mg/kg), which is the opposite of what you might expect from sensitised receptors, and which has been attributed to pharmacokinetic shenanigans related to an increased volume of distribution.
  • Factors that decrease acetylcholine release, which may be numerous, and which include, in descending order moving from the central nervous system down:
    • Anaesthetic agents, which act at the level of the motor neuron to reduce firing, and therefore should reduce the availability of acetylcholine. This is ancient anaesthetic lore: Bonta et al looked at this in 1968, comparing some awake paralysed cats to ones anaesthetised with thiopentone or halothane. The effects of a general anaesthetic decreased the required dose of pancuronium by around 30%. This effect appears to continue across all blockers, including depolarising ones. It also raises an interesting point - if anaesthesia empowers NMJ blockade by decreasing acetylcholine release from (both tonic and voluntary) muscle contraction, does this mean that one, through voluntary effort, can send enough signals down their motor neurons to overcome the effects of a nondepolarising blocker? 
    • Regional local anaesthetic techniques, particularly those that produce motor block, can interfere with acetylcholine release by blocking the upstream propagation of action potentials. No action potentials arriving to the junction means no acetylcholine release, and the result, theoretically, should be an increased sensitivity to nondepolarising blockers. How one might detect this clinically, of course, is impossible to say, considering the patient already has motor block (so how much more flaccid paralysis do you really get to observe?) Still, an apparent decrease in the dose requirements for rocuronium appears among observational data such as this  observational work by Santos et al (2017), looking at patients with spinal and epidural anaesthesia.
    • Agents that interfere with presynaptic action potential propagation can inhibit the exocytosis of acetylcholine. Local anaesthetic would definitely do this, but it is also an effect seen with volatile agents such as sevoflurane and isoflurane (Castro et al, 2015). Voltage-gated sodium channels are implicated in the mechanism of this.
    • Agents that interfere with acetylcholine synthesis or storage, such as hemicholinium and vesamicol, would definitely reduce the availability of acetylcholine at the synapse and increase the speed of onset of NMJ blockers, except nobody anywhere uses these drugs for anything, and for most normal people they will only ever appear on the pages of textbooks
    • Agents that interfere with calcium flux and secondary intracellular messaging decrease acetylcholine release. There are actually quite a number of these:
      • Frusemide potentiates the effects of nondepolarising agents by a mixture of different mechanisms including calcium-related effects as well as displacement of cAMP from its binding sites (Scappaticci et al, 1982).
      • Calcium channel blockers, antihypertensive agents with supposedly high selectivity for cardiac and smooth muscle, can also interfere with NMJ transmission by interfering with calcium-mediated vesicle exocytosis, and patients chronically using amlodipine or verapamil did in fact have detectable neurophysiological abnormalities on their EMG (Ozkul, 2007).
      • Aminoglycosides are also responsible for potentiating the actions of nondepolarising blockers through some degree of calcium channel blockade, as is demonstrated by the fact that their influence is completely reversed by calcium chloride (Paradelis et al, 1988)
• Factors that partially depolarise the myocyte membrane would logically reduce the amount of neuromuscular blocker required to completely depolarise it.
  • Hypermagnesemia: Okamoto (1992) determined that increasing the concentration of magnesium decreased the necessary concentration of both depolarising and non-depolarising neuromuscular junction blockers in a dose-dependent manner. The discussion section is inaccessible, unfortunately, which means the mechanism underlying this phenomenon will remain obscure to our readers until Dr Okamoto answers the emails. Magnesium, as a physiological calcium channel blocker, would also be expected to interfere with acetylcholine release.
  • Hypocalcemia should also theoretically cause a faster onset of neuromuscular junction block, following from the study above.
• "Pre-curarisation" or "priming" is a technique that probably does not belong in this "decrease the necessary dose of blocker" section, because technically that does not happen here. The technique calls for the use of small, sub-clinical dose of neuromuscular junction blocker to be administered first, to be followed by a later "proper" dose which achieves the clinically desired effect more rapidly. This works because 70-80% of receptors must be occupied for a  "proper" clinical effect to develop; which means one could administer enough blocker to occupy a fair percentage of these receptors, and still be breathing relatively well. Then, the anaesthetist can administer the rest of the agent rapidly, and - because the NMJ is already partially blocked - thereby achieve the onset of optimal incubating conditions much faster. Or so the theory goes.
  This technique has been promoted as a solution to the frustratingly slow onset of high-potency agents such as pancuronium and vecuronium. It does actually work: for example, Rao et al (2018) described the use of vecuronium for rapid sequence induction, where 0.02mg/kg of vecuronium - 1.4mg per 70kg person, or about 20% of the normal dose- was administered first, and then the rest was given as a bolus for intubation. The onset of optimal conditions was much faster with this, changing from 149 seconds to 63 seconds on average.
• Defasciculation is a form of "priming" but for a completely different reason. In those cases, a small (10% of ED₅₀) dose of a nondepolarising blocker is administered before the rapid sequence induction for the purposes of reducing fasciculations, and thereby ameliorating the risk of raised ICP and all the other adverse effects of suxamethonium. Of course, at this stage most reasonable people would raise the point that, to completely avoid all the adverse effects of suxamethonium, the operator should have just finished the job with the rest of the nondepolarising agent, and discarded the suxamethonium ampoule.  
• Self-taming is a form of defasciculation where suxamethonium is used for both the "priming" dose and the intubating dose. This is not even anything to do with the speed of onset of block, and therefore does not belong in this section or even in this chapter, but there was seriously nowhere else to put it. The mechanism behind "self-taming" is usually described as "neuromuscular desensitisation or accommodation", without any further explanation. If one digs a little deeper, one can find an explanation of accommodation in textbooks that describe the action of depolarising agents, where it is presented as the phenomenon of surrounding the motor endplate in a "ring" of inactive voltage-gated sodium channels that insulates the depolarised junction from the rest of the muscle, and thereby prevents neurotransmission. Yes, reader, that does sound just like the basic vanilla explanation of how depolarising agents are supposed to work, and does not explain how a small dose of depolarising agent is supposed to prevent fasciculations. Fortunately, Baraka (1977), who first described this phenomenon, did manage to reference a 1968 work by Waud for explanations, where "accommodation" is described as something that occurs variably and dose-dependently across different endplates. The upshot is that enough endplates are "accommodated" by the first small dose of sux, so that when a "proper" dose is given, there are not enough motor units to get excited and produce clinically meaningful fasciculations. And by "clinically meaningful", of course we mean "something the patient may complain about in the follow-up clinic". Wald-Oboussier (1987), testing this method against pancuronium, concluded that both techniques were equally effective in protecting the patient from myalgia.

Factors that could either increase OR decrease the concentration required to produce neuromuscular junction block, depending on the sort of agent used:

• Factors that reduce the number of acetylcholine receptors such as normal old age or myasthenia gravis tend to reduce the necessary dose of non-depolarising muscle relaxant, and increase the dose requirements of depolarising agents. In the case of myasthenia, the receptors are actually destroyed by autoantibodies against the acetylcholine receptor itself. The fewer the receptors, the easier it is to occupy 70-80% of them with fewer drug molecules, thereby producing non-depolarising block. According to Eisenkraft et al (1990), the ED₉₅ for vecuronium in patients with myasthenia is reduced by about 55-40%, i.e. only about half the dose is required to produce the same level of block.
  
  The opposite is true for depolarising agents. The ED₉₅ of suxamethonium in myasthenia gravis is 2.6 times higher than in non-myasthenic patients, according to Eisenkraft et al (1988); which means 2.0 or 2.5mg/kg would be required for rapid sequence induction. The destruction and loss of receptors is blamed for this - fewer receptors means a larger fraction of them needs to be occupied and opened by the suxamethonium molecules to produce the same degree of depolarisation, and therefore more molecules are required. Wainwright & Broderick (1987) present this idea while lamenting that there is no supporting evidence to back it. 
  
  Similarly, in the elderly, the postsynaptic membrane contains fewer receptors, and less acetylcholine ends up being released with each action potential, according to Khosa et al (2019). The consequence of this should be an increase in sensitivity to NMJ blocking agents, and indeed this is what is seen clinically. Anaesthesia textbooks (and therefore anaesthesia exams) tend to discuss this in terms of receptor number, but in fact it appears that pharmacokinetic factors play a more important role (Cope & Hunter, 2003), and this is supported by the finding that the sensitivity to suxamethonium is unchanged in the elderly, whereas you would expect it to decrease if the receptor numbers were the main influence. 
• Hyperkalemia hastens the onset of depolarising, and delays the onset of nondepolarising neuromuscular blocking agents. This appears rather often in the Annals of Things Anaesthetists Offhandedly Tell Their Trainees, but is not described very well by official textbooks or review papers. Even Wilbanks et al (2005), a case report that purports to describe and discuss how hyperkalemia affects NMJ blocker activity, spends pages and pages discussing irrelevant things like TOF monitoring and only throws a couple of lines in the direction of the actual physiological mechanism that is the topic of their paper:

  "Hyperkalemia potentiates the neuromuscular blockade induced by muscle relaxants by decreasing the excitability of the skeletal muscle...
  Initially hyperkalemia causes hyperexcitability of cellular membranes by moving the resting membrane potential closer to threshold potential, a smaller stimuli is needed to initiate a contraction. Eventually the Na,K-ATPase pumps begin to fatigue from the excessive depolarizations, and cellular membranes become less excitable"

  This is not supported by anything more than a reference to an old textbook by Stoelting & Dierdorf (2002) , but following the trail of crumbs one ultimately arrives at a 1980 paper by Douglas and Barbara Waud. The Wauds threw guinea-pig lumbrical muscles into baths with toasters, and recorded the effects of different concentrations of bath potassium.  The results produced clearly indicated that higher extracellular potassium increased the dose requirements of the neuromuscular blocking agent:
  
  In short, going by these ED₅₀ values, the difference in the dose requirements between a hypokalemic patient and a hyperkalemic patient may be as great as 36%. But these were the toe muscles of cute little mammals. What about a whole organism? Spectacularly, it appears the Wauds were reproducing an experiment that Hill et al (1978) had already performed in whole living dogs, which arrived at basically the same conclusion.

Factors that have no influence on the speed of onset of neuromuscular junction blockers

It felt reasonable to list these because they are often mentioned by "model" college answers, and this makes them at least as important as the other factors.

To borrow a representative statement from the examiners, "these drugs are charged molecules which do not cross cell membranes and have a low volume of distribution. Absorption from GIT, Lipid solubility, pKa, metabolism and clearance have minimal relevance to speed of onset".  This statement is largely accurate, but, as is explained in the main chapter on the pharmacology of these agents, it is hard to track down a reliable resource containing these data. Still, it is worth knowing why the pharmacokinetics are largely meaningless in determining the speed of onset of these agents:

• They are minimally bioavailable when administered orally, which means South American natives were able to enjoy the meat of the animals they killed with curare darts without experiencing any neuromuscular junction effects.
•  The volumes of distribution of these drugs are indeed rather small, as their lipid solubility is extremely poor and they are generally confined to the extracellular fluid (vecuronium, the most lipid soluble of the agents, has the largest VOD, of around 0.27 L/kg). Absorption from
• The pKa of these agents - to remind the careless reader, a measure of the ionisation of the drug, the pH at which 50% of it exists in an ionised form - is therefore so difficult to establish that for many of the agents there is no agreed-upon value. Owing to the charge of the quaternary ammonium they all possess, these drugs are practically 100% ionised at a huge range of pH values. 
• Because of this high water solubility and poor lipid solubility, these drugs are indeed incapable of crossing cell membranes, and even if they did, they would have nothing to do in there, as their effect sites are extracellular.
• Metabolism and clearance have a lot to do with the offset of these drugs, but not very much to do with the onset, as the rate of their metabolism does not really factor in to their activity at the effect site (where none of them are metabolised).