This chapter is relevant to Section G8(i) of the 2017 CICM Primary Syllabus, which asks the exam candidate to "understand the detailed pharmacology of inotropes and vasopressors". Levosimendan is an inotrope and vasodilator with a unique mechanism of action (it acts as a "calcium sensitiser" by stabilising calcium-bound troponin C) and an unusually long duration of action due to its long-lived active metabolite (OR1896). It has appeared in two CICM primary exam papers, and will probably appear in others, as its popularity in critical care remains undiminished by either cost or trial results.
- Question 10 from the second paper of 2013 (digoxin vs. levosimendan)
- Question 8 from the first paper of 2012 (adrenaline vs. levosimendan)
- Question 15 from the second paper of 2020 (dobutamine vs. levosimendan)
Class Inodilator Chemistry Pyridazinone-dinitrile derivative Routes of administration IV Absorption High oral bioavailability (85%) Solubility pKa 6.3, minimally water soluble Distribution VOD = 0.3 L/kg; 99% protein-bound Target receptor Troponin C Metabolism Metabolised mainly by the liver (95% into inactive metabolites, and 5% into OR1896 which has a long half-life and potent activity) Elimination Eliminated mainly as renally excreted metabolites Time course of action Levosimendan itself has a half life of around 1 hour, but OR1896 has a half-life of over 80 hours. Mechanism of action By binding to troponin C, levosimendan stabilises its open state, allowing muscle contraction. This increases contractility. It also vasodilates by activating ATP-sensitive potassium channels in vascular smooth muscle (like hydralazine). Additionally, at high doses,it acts as a phosphodiesterase (PDE3) inhibitor. Clinical effects Increased cardiac contractility, increased heart rate, significant arterial and venous vasodilation (including pulmonary arterial vasodilation), decreased afterload, increased arrhythmogenicity. Purported cardioprotective effect. Single best reference for further information Antila et al (2007)
This is a popular drug, and there is certainly plenty of material out there ready to promote its positive properties. Antila et al (2007) is probably the most readable short overview which contains most of the relevant details for CICM exam purposes. Innes et al (2003) and Figgitt et al (2001) are also useful, as they offer some details regarding The manufacturer propaganda from Orion Pharma is also potentially useful.
Chemical structure and chemical properties of levosimendan
Levosimendan is difficult to characterise chemically. It is unhelpfully described as a hydrozone, a pyridazinone and a nitrile, which are all rather broad groups of rather heterogeneous chemicals. Hydrazones are basically ketones where the oxygen has been replaced with a NNH2 functional group, which is a huge fundamental class of substances that are relatively scarce in nature because of their reactivity and instability. Pyridazines are heterocyclic molecules with adjacent nitrogen atoms, also scarce in nature because their synthesis requires pnictogen hydrides, hideously toxic and highly reactive molecules which only occur naturally in the nightmares of chemists. Nitriles are even more ubiquitous, as the term technically describes any organic compound that has a −C≡N functional group.
As you can plainly see, it has all of these features, but it would be hard to group it into the same category as the stuff they make blue gloves out of. The best short classification was offered by Innes & Wagstaff (2003) who called it a "pyridazinone-dinitrile derivative". Specifically, it is simendan which is the dinitrile derivative, and levosimendan is its active levo-enantiomer. Dextrosimendan is also active, but is 47 times less potent (Kaheinen et al, 2006). Apart from this brother-molecule, pharmacological relatives of levosimendan (i.e. substances which are chemically distinct but which have a similar mechanism of action) include vesnarinone and pimobendan, as well as a series of experimental compounds which are only known by their candidate numbers (MCI154, EMD-53998 and EMD-57033). They all vasodilate, calcium-sensitize, and phosphodiesterase-inhibit.
Levosimendan presents as a liquid concentrate (2.5mg/ml). The drug itself is a yellow crystalline powder that is not stable in aqueous solutions at physiological pH and readily undergoes hydrolytic decomposition in water. As the result, the pharmaceutical presentation of commercially available levosimendan is with anhydrous (100%) ethanol, as well as povidone as a solubilising agent which makes it even yellower (in fact apparently it can turn orange with age, without any loss of effect). Citrate is used to keep this yellow glug at a pH of 3.0, and even then the drug must be kept in the fridge at 2-8º C. Most drug handbooks and PI documents recommend it to be diluted with 5% dextrose, but this does not seem to have a solid scientific basis, and in fact some studies of levosimendan used saline as a diluent and still reported positive results.
Pharmacokinetics of levosimendan
Both pharmacokinetics and pharmacodynamics of this drug are reasonably unique, which means it pays to know them in some detail for exam purposes.
Administration and absorption
Levosimendan is usually administered as a slow intravenous infusion. It can be administered either centrally (i.e. via a central line) or peripherally through a cannula. Though it is diluted in alcohol and buffered to a pH of 3.0, the infusion is usually sufficiently slow, and thrombophlebitis is not considered a major risk (or, at least, case reports describing skin necrosis following levosimendan extravasation are rare, insofar as the author could not find one). As it is frequently administered to critically ill patients who already have central lines for other reasons, this is usually a non-issue.
Having said all this, the bioavailability of levosimendan is around 85%. Sandell et al (1995) fed a tiny bit (0.5mg) to healthy volunteers and determined that it underwent minimal first pass metabolism. That means that, at least theoretically, it would be possible to gradually load somebody with this drug using a purely enteric formulation. This is not a novel idea. For example, Markku et al (2008) tried it in 307 heart failure patients with NYHA IV heart failure. Ironically, they called it the PERSIST study. Levosimendan was administered orally (1mg bd) and the investigators looked at an exploratory surrogate endpoint (the "Patient Journey" composite) similar to what was used by the COMET trial. The total number of days alive and out of hospital was basically similar for all groups, and the raw mortality data is difficult to interpret (it was higher in the low dose levo group, but lower in the high-dose group and placebo). Given the terrible experience we have had with other highly arrhythmogenic oral inotropes, one might forgive the investigators for not persisting with further trials of oral levosimendan.
Rate of infusion and the idea of loading
Levosimendan is supposed to be administered as an infusion with an initial loading dose. Or at least, the prescriber information booklets seem to ask for this. For instance, the Simdax sheet from medsafe.govt.nz says:
"The treatment may be initiated with a loading dose of 6-12 μg/kg infused over 10 minutes. ... The lower loading dose of 6 μg/kg is recommended for patients on concomitant intravenous vasodilators or inotropes or both at the start of the infusion."
And this is indeed what seems to happen in many published studies. For instance, in the LIDO trial (Follath, 2002), "an initial loading dose of levosimendan of 24 μg/kg was infused over 10 min". However, about 9% of these patients developed the sort of hypotension that is worrying enough to stop the study drug for an hour or so, to recommence at half the normal rate.
This loading dose is, fortunately, not essential. The haemodynamic effects of this drug are maximal about 24-48 hours following an infusion, i.e. they take a while to develop anyway, which means there is absolutely no reason to rush things. For example, Lunghetti et al (2011) omitted loading and determined that the haemodynamic effects were exactly the same after enough time had passed, except that without loading "we did not observe severe hypotensive episodes compared to standard therapy". Looking at a selection of levosimendan trials (eg. from this 2017 meta-analysis by Silvetti), only two used a loading dose, and the rest (four) used a continuous infusion. In short, the daftness of giving a high dose of a potent vasodilator just as the patient's cardiovascular system is at its most unstable is clearly visible to most trial designers. On the other hand, if one's haemodynamics are sufficiently robust to withstand the assault of 24 μg/kg of levosimendan over 10 minutes, then does one really need the levosimendan? In short, if in their own practice one decided to omit the loading dose in haemodynamically imperfect patients, one's cowardice would not be criticised.
Protein binding and distribution
Levosimendan is highly protein-bound. After administration, the vast majority of it (~ 99%) ends up attached to albumin (Chu et al, 2014). Using radiolabelled drug on healthy volunteers, Sandell et al (1995) determined that it distributes into a rather small volume, something like 29.1L (or, 0.3 L/kg). It has a pKa of 6.3, and its terrible water solubility is already discussed above.
Metabolism and elimination
Levosimendan undergoes extensive hepatic metabolism by conjugation with glutathione to pointless and inert cysteine or cysteinylglycine conjugates. About 95% (so, $1235AU) of the administered dose is wasted in this way. The remaining 5% undergoes metabolism by a minor pathway, through the reduction of the N-N bond of levosimendan in the intestine. This yields an amine metabolite, OR-1855. This product is further N-acetylated into OR1896, which is the most pharmacologically interesting. part of this discussion.
OR1896 is only about 40% protein-bound (Antila et al, 2004), and so even though only about 5% of levosimendan gets metabolised into OR1896, it has a much more important pharmacodynamic effect. It also has a much longer half-life. Levosimendan itself has a half-life of around 1.4 hours, whereas OR1896 has a half-life of 80 hours, which is also roughly the timeframe over which it reaches peak concentration (Koskinen et al, 2007). OR1986 retains most of the activity of the parent drug, and is slowly metabolised into renally excreted molecules (unimaginatively named M9, M10 and M11).
Pharmacodynamics of levosimendan
This is probably the most exciting aspect of levosimendan, and therefore the most attractive target for exam writers. The following summary is mainly drawn from the excellent articles by Innes et al (2003) and Figgitt et al (2001).
Unique calcium sensitiser effects
Levosimendan is supposed to increase myofilament calcium sensitivity by binding to cardiac troponin C. As you might remember, troponin C is the calcium-binding subunit of troponin, which has four calcium binding sites. When calcium binds to these subunits, troponin C remains open, allowing muscle contraction. Without calcium, troponin C maintains the activated state of an inhibitory region of troponin I, which inhibits muscle contraction.
By binding to troponin C, levosimendan is said to stabilise its calcium-induced conformational change. This should theoretically lead to increased contractility at any given intracellular calcium concentration. Whatever the level of intracellular calcium you have, it should work harder, because once it binds to troponin C it should end up being stabilised in its open form, permitting contraction which would otherwise have been "forbidden".
However, logically, if you say that this drug increases the stability of the troponin-calcium complex, then it should also decrease the dissociation of calcium from troponin C, and therefore delay relaxation. This effect is not seen: in fact, levosimendan is thought of as a lusitrope, i.e. it is supposed to increase relaxation. This is thought to be because of concentration-dependent binding to troponin C. Levosimendan and its metabolites are supposed to bind to troponin C only in the presence of a high concentration of calcium, which occurs only in systole. In diastole, it releases its grip on the troponin molecule and allows relaxation to proceed.
This inotropic effect is also unique because it is thought to be devoid of negative energy balance implications. The same amount of calcium is required, which means ATP-hungry calcium pumps do not need to work any harder. Additionally, there does not appear to be any additional myosin ATPase activity (Haikala et al, 1995). In short, this increase in contractility occurs at no additional energy cost to the myofibre, making this drug particularly attractive in scenarios where cardiac energy supply is limited (eg. in coronary artery disease).
Other effects of levosimendan
Apart from this "essential" effect, levosimendan does a bunch of other helpful and unhelpful stuff, which might be useful to mention in written exam answers. Because much has been made of its many effects, Papp et al (2012) had produced a consensus statement, "intended to serve as a reference when positioning levosimendan among the currently available drugs for the management of acute heart failure". Notwithstanding the fact that many of these experts were shills for Orion Pharma, their article is a useful summary of pharmacological effects:
- ATP-sensitive potassium channel activator: Levosimendan seems to open these channels, acting essentially in the same way as hydralazine and minoxidil. Moreover, Yokoshiki et al (1997) found some increased effect in the presence of increased ADP and decreased ATP, i.e. where there is ischaemia. Höhn et al (2004) also found some venodilating effects, which they demonstrated in the isolated human saphenous vein. It dilated, and this as reversed by glibenclamide (an ATP K-channel antagonist).
- Mitochondrial potassium channel activator: levosimendan apparently stimulates the ATP-sensitive potassium flux into the matrix of cardiac mitochondria (Kopustinskiene et al, 2004). This is supposed to have some sort of cardioprotective effects.
- Phosphodiesterase inhibitor: in very high doses, levosimendan acts as a PDE3 inhibitor. Toller et al (2006) report that this is seen in concentrations greater than 0.3 μmol/L, whereas the therapeutic range is probably around 0.03-0.3 μmol/L. Ergo, after multiple doses or with unreasonably high doses, levosimendan will start acting like milrinone.
Clinical effects of levosimendan
Sure, there's plenty of data about the effects of levosimendan in healthy volunteers (Lilleber et al, 1994), but what intensivist could possibly give a flying hoot about those people. They tend to stay out of intensive care units, smoking drinking and making love in nightclubs. Our population of interest tends to trend older, featuring huge dilated ventricles, infarcted coronaries and dehisced sternal wounds. It is therefore more meaningful to describe what happens to a truly broken circulatory system when levosimendan is inflicted upon it.
Effects on haemodynamic indices
Álvarez et al (2006) looked at the effects of levosimendan in patients who had poor cardiac output following cardiac surgery, and Russ et al (2007) looked at its effects in patients with cardiogenic shock which failed to resolve following revascularisation. There was an encouraging increase in the cardiac index, at least some of which must have been the consequence of an increased heart rate, as demonstrated in their original graphs below:
Additional haemodynamic indices are reported by Malliotakis et al (2007). Without reproducing their graphs, it will suffice to say that levosimendan is a systemic and pulmonary vasodilator. Oh, well, let's reproduce those graphs anyway:
Other effects of levosimendan
These tidbits of information are unlikely to ever come up, as they are probably too clinically relevant for the CICM Part One exam, but insufficiently clinically relevant for the CICM Part Two exam. Sure, levosimendan probably does these things, insofar as serious researchers have reported them as effects, but they are generally left out of even uncritically praiseful pharma literature, because the evidence for them is often limited to studies of isolated rat myocytes swimming in brine. For a faithful record of these effects, one may be directed to Kasikcioglu et al (2006). Without wasting any more of the reader's time, these will be summarised here in the shortest most unprofessional way possible, as an unordered list with (possibly broken) links:
- Cardioprotective and coronary vasodilator effects: The mitochondrial and coronary arterial effects of levosimendan, mentioned above, should theoretically improve blood flow and energy consumption efficiency of ischaemic ventricle myocytes. Kasikcioglu et al list several studies which demonstrate the blood flow does in fact improve, but none have gone so far as to demonstrate that this is somehow separate from the overall increase in the cardiac output.
- Anti-inflammatory effect: the decrease in soluble inflammatory and proapoptotic mediators following a levosimendan infusion has been noted, but its significance is unclear and the connection is pretty tenuous. The aforementioned mediators decrease in association with the use of levosimendan, and they are known to increase the cellular loss following ischaemia, so logically levosimendan should somehow protect from this apoptotic deterioration. But really, does it? Experts produce noncommital blather in response, eg. "this speculation, if proved true, would mandate a fundamental paradigm shift", etc etc.
- Pleiotropic effects: Apparently levosimendan does something weird to extracellular matrix metabolism (Tziakas et al, 2005), which should theoretically prevent some sort of destructive failure-associated remodelling in the myocardium. Exactly how this produces constructive changes in the diseased ventricles, or how much levosimendan you need to marinate them in, remains to be established.
- Neurohormonal modulation: levosimendan is said to decrease the perfectly natural increase in catecholamine secretion in response to shock, which is probably logical considering what it does to the cardiac output.
- Reversal of myocardial stunning: Sonntag et al (2004) found some improvement in the motility of previously movement-restricted LV segments in patients suffering from acute myocardial ischaemia.
- Antiarrhythmic effect: Apparently, levosimendan is supposed to decrease the effective refractory period of myocytes and AVT nodal tissue, and that is supposed to have an antiarrhythmic effect. This could not possibly be true, as the REVIVE II study clearly demonstrated an increased propensity to both atrial and ventricular arrhythmias in the levosimendan group.