Midazolam

In summary:

"Midazolam" by Dundee et al (1984) remains an excellent resource for this topic, even though it is at least ten years older than any of the current crop of exam candidates. Its age is nothing to turn your nose up to, as midazolam has remained midazolam during this time, and new pharmacokinetic principles have not magically sprouted since the 1980s. It also has the distinction of being free to read. For a more modern alternative, the reader can turn to Okkola & Ahonen (2008), which also compares it to some of the other benzodiazepines. 

Class and chemical relatives

Midazolam is a benzodiazepine. This whole class of drugs are basically the product of a series of hideous animal torture experiments from the 1960s, when Sternbach and his buddies at Hoffman-La Roche got a hold of a bunch of different aminoketones with fused benzene and diazepine rings, and then substituted random side groups into them until their experimental cats stopped reacting to electrical shocks. Midazolam was the subject of the eighty-fourth paper on these experiments, if you need an estimate of the number of cats they must have gone through. It appeared on the scene in 1977, interestingly the same year that propofol became commercially available. Its chemical relatives are diazepam, lorazepam, oxazepam, and everyotherazepam you can think of.

One chemical difference that distinguishes it from the others is its pH-dependent solubility. Midazolam presents as a hydrochloride salt in an acidic buffered solution with a pH of around 3.0, and it has excellent solubility at that pH. because the diazepine ring is opened and the molecule is polar. As soon as  it is injected and the pH increases, the ring closes, and midazolam becomes readily lipid soluble, just in time to cross the blood-brain barrier. Here, a figure from Gerecke (1983) illustrates its reversible ring opening in the presence of abundant protons:

Pharmacokinetics

Apart from neat carbon chemistry tricks, the pharmacokinetics of midazolam are fairly unremarkable, insofar as it resembles almost every other CNS-active agent. 

• Routes of administration and absorption: Midazolam is an IV agent, but goes down well enough via the IM or subcutaneous route. It can also be administered orally. After about 30 minutes, an oral dose has made it to a peak concentration. Obviously, the dose required is  larger, as a lot of the drug is lost to first pass metabolism, but this is not such a major issue (oral bioavailability is 44%). It is ready absorbed from mucous membranes, and can be administered intranasally and buccally to unhappy-looking epileptics.
• Solubility and distribution: Midazolam is 96% protein-bound. It distributes rapidly (the α-half life is only 15 minutes). 
• Metabolism and elimination: this is a one of those drugs which depends on hepatic perfusion for its elimination, as it has a high extraction ratio. The main metabolite is  α-hydroxymidazolam, a water-soluble renally cleared substance which retains enough of the original drug's activity to become a major problem in patients with renal failure. Fortunately, it too is rapidly conjugated to a glucuronide, making it both rapidly excretable and harmless. Overall, the elimination half-life of midazolam is said to be 1.5 - 3.5 hours. 

Mechanism of action

The pharmacodynamics of midazolam are representative of the pharmacodynamics of its class, or at least it tends to be used as the archetypal benzodiazepine representative when GABA agonist mechanisms are discussed in the literature. Of that literature, Möhler et al (2002) and Griffin et al (2013) are some of the better examples. To save the reader some time, all benzodiazepines act as allosteric modulators of the activity of the GABA_A receptor, a ligand-gated chloride channel (and they have no affinity for the metabotropic GABA_B receptors). By binding to a  pocket created by the α and γ subunits of the pentameric GABA_A receptors, benzodiazepines changes the channel morphology in a way which increases its affinity for GABA, and increases the frequency of its opening. This pocket is exactly the same place where nonbenzodiazepine "Z-drugs" bind, and distinct from the binding site of barbiturates (which increase the duration of opening instead). In either case, increased GABA_A receptor activity leads the neuron down the final common pathway of all sedatives, where the excess of intracellular chloride  makes the membrane hyperpolarised.

Clinical effects

Sedation, obviously, is the clinical effect of midazolam. Occasionally, a CICM exam answer will call for something more thoughtful and detailed. Obviously it has some special characteristics. For example, one might have noticed one's senior colleagues using midazolam instead of propofol or other agents to maintain sedation in haemodynamically unstable patients. Okkola & Ahonen (2008) turned out to be a goldmine for this material.  

• Airway effects: Midazolam, like most sedating agents (apart from that black sheep ketamine) tends to depress airway reflexes and produces airway failure in sufficient doses. In fact, when Norton et al 2006) compared it to propofol, they found that there really was no difference between the two agents' propensity to cause airway obstruction (or, rather, that interindividual variability of their subjects was greater than the difference between the drugs).
• Respiratory effects:  Like opioids, benzodiazepines in general tend to blunt the respiratory response to hypercapnia. For any given PaCO₂ value, the minute volume will be lower while under their influence (i.e. the response is flattened). But there is still a response: unlike opiates, these drugs do not shift the PaCO₂/ventilation relationship curve to the right (Sunzel et al, 1988). Additionally, benzodiazepines tend to make your ventilatory drive less responsive to changes in hypoxia, and to decrease the hypoxia-induced sensitivity of this drive to CO₂. The upshot of these effects is that the midazolated patient will never actually stop breathing, but they certainly won't be winning any medals for respiratory autoregulation, and will likely accumulate a lot of CO₂. 
• Circulatory effects: By taking away the sympathomimetic effects of pain and anxiety, midazolam (and, in high doses, other benzodiazepines) will decrease the blood pressure (by about 15-20% according to Sunzel et al, 1988) and blunt the baroreceptor reflex. These effects are clinically significant, and one should not be lulled into a sense of security by the supposed haemodynamic neutrality of midazolam. In a study by Weinbroum et al (1997), where midazolam and propofol were compared specifically as long term sedation options for ICU patients, midazolam was slightly more hemodynamically benign than propofol, but both still produced an impressive drop in blood pressure:
  
• Cerebral metabolism is affected, but not to the same extent as by barbiturates or propofol. Midazolam or any of the other benzodiazepines can never produce an isoelectric EEG. Additionally, they do not decrease the cerebral oxygen demand as much as other sedating agents, and are therefore less suited for the sedation of patients with raised intracranial pressure; in the sense that they will not decrease the ICP (Papazian et al, 1993).
• Anticonvulsant effects are often seen at doses which are lower than the doses usually required to cause sedation. However, this phenomenon is almost never observed in the ICU clinical setting, as the patients requiring midazolam infusion are often those in whom small intermittent doses and other agents have already failed, i.e. patients with very severe disease.
• Amnesia seems to be a property of all the substances we use to sedate people, but midazolam specifically has earned a reputation for causing more memory loss than other drugs, allowing some truly tasteless preanaesthetic banter to escape a serious complaints process. When compared to other drugs (eg. propofol), midazolam does indeed appear to produce a greater amnesic effect. When Polster et al (1993) gave midazolam and propofol to healthy volunteers, their performance in word recognition memory tests was markedly more impaired after midazolam.