Explorer-scientists, including Charles-Marie De La Condamine, Edwin Bancroft, and Charles Waterton, travelled extensively in different areas of South America. They provided the raw material to European pharmacologists and physiologists who undertook the initial studies of curare. These hardy explorers often put themselves at personal risk; the Amazonian forest was an inhospitable place, and working with the raw material was dangerous. The French scientist FD Herissant describes how he and his assistant became weak and almost unconscious while boiling the crude curare extract in order to create a concentrated mixture. “A pint of good wine and sugar” cured both investigators.

The eighteenth century physician, Edwin Bancroft, brought back samples of crude curare from South America for study. He was the first to thoroughly report the diverse tribal methods of preparation and storage of the curare. The natives used a variety of vessels for storage of the curare preparation: a clay pot (pot curare), a calabash gourd (calabash curare), or a hollow piece of bamboo (tube curare, hence “tubocurare”). His son, Edward Bancroft provided samples to the English surgeon, Benjamin Brodie, who made the crucial observation that animals poisoned with curare could be kept alive by inflating their lungs [5]. Brodie also suggested treatment of tetanus with curare [6].

Despite poor therapeutic experiences, many considered using the curare “poison” for medical purposes, but its inherent dangers made it unsuitable as a drug fortrial and error experimentation. Nonetheless, we were getting nearer to the discovery of the beneficial uses of curare.

The psychiatrist Abram Bennet was a colleague of Michael Burman, an orthopedic surgeon who had reported his use of curare in childhood spasticity. Bennet saw a potential for the use of curare to minimize the traumatic consequences of convulsive shock therapy [2,20]. Metrazol-induced convulsions produced a 40–50% incidence of compressive spinal fractures. As reported by Bennet, curare prevented these fractures [20], Such results showed the effectiveness of curare, and probably preserved the practice of convulsive therapy.

In 1942, a modest paper, brief and to the point, altered forever the practice of anesthesia [21]. With little fanfare, no IRB approval, no oversight, and no previous experience, Harold Griffith and Enid Johnson at the Homeopathic Hospital of Montreal administered Intocostrin to 25 patients anesthetized with cyclopropane. All patients breathed spontaneously throughout surgery, none had their tracheas intubated, and some were challenging (e.g., a 112 kg patient for hemorrhoidectomy and an “…extremely obese woman…” for dilatation and curettage.) Griffith opined: “we have been so much impressed by the dramatic effect produced in every one of our patients that we believe this investigation should be continued.” It is tempting to speculate on what might have followed had any of these patients died from what Beecher might have labeled a “curare death” (see below).

In 1943, two chemists at Squibb purified d-tubocurarine from the bark of Chondodendron Tomentosum. King had extracted the same compound in 1935 but had not known the source of his preparation. The purified d-tubocurarine had 6 times the potency of Intocostrin, and soon replaced it.

To better understand the patient’s potential perception of curare’s effects, some doughty clinicians submitted to paralysis while awake. In 1943, Helen Barnes underwent laryngoscopy after her colleagues gave her 4 ml of Intocostrin IV while she was awake. Her colleagues passed a laryngoscope and viewed her vocal cords with ease. She then used Intocostrin to facilitate tracheal intubation in 5 patients [25], recognizing that these doses of curare necessitated pharmacologic reversal with neostigmine. In 1946, Frederick Prescott, a pharmacologist at the Westminster hospital in London, submitted to awake paralysis with Intocostrin, a reportedly unpleasant experience during which he remained aware [24].

Despite the reports of Barnes and Prescott, some clinicians believed that curare depressed the central nervous system and could produce anesthesia, and many patients suffered awareness under anesthesia in those early years. “I…was awakened by the most excruciating pain in my tummy. It felt as if my whole insides were being pulled out…” [26].

Scott Smith, a pediatric anesthesiologist in Salt Lake City thought that curare anesthetized babies because they remained immobile during surgery. To test his belief he had colleagues completely paralyze him with curare with support of his respiration by bag and mask [27]. In contrast to what he expected, he perceived touch and pain throughout his paralysis and required tracheal intubation.

Many anesthetists initially thought that the use of muscle relaxants was inherently unsafe, a correct perception for partially paralyzed patients allowed to breathe spontaneously and whose tracheas were not intubated. Some suggested that curare had inherent cardiotoxicity [28]. However in the ten years following its introduction, it became clear that inappropriate use rather than curare caused the problems.

The decade opened propitiously, with the 1951 clinical introduction of succinylcholine. Although succinylcholine was first synthesized in 1906, it fell to Bovet in 1949 to observe its paralyzing action [30]. This observation, and his earlier discovery of gallamine, resulted in his 1957 Nobel Prize for Medicine.

In contrast to Intocostrin, succinylcholine received a warm welcome. Frances Foldes stated “Its use is not accompanied by unwanted side-effects and the incidence of post-operative complications after its use is low”[31]. But first impressions can be deceptive. Subsequently, numerous adverse effects of succinylcholine were identified [6]. One, prolonged block in patients with plasma cholinesterase deficiency, was an early example of the influence of pharmacogenetics, a term coined in 1957 (see Chapter44 on A History of Pharmacogenomics Related to Anesthesiology). The familial nature of prolonged block was soon identified [32], and its relation to atypical cholinesterase demonstrated [33].

In 1954, the “honeymoon” ended. Beecher and Todd reported a six-fold increase in “anesthesia-related” deaths in patients receiving “curare” [28]. To Beecher and Todd, “curare” encompassed d-tubocurarine, Intocostrin, dimethyl-tubocurarine (metubine), gallamine, succinylcholine, and decamethonium: the overall risk of death from anesthesia at that time was 1 in 2100 (far greater than today), but in a patient given curare the risk increased to 1 in 370. No cause was determined, and the increased mortality was not related to the patient’s medical status, experience of the provider, or any other factor that could be identified. Reversal with an anticholinesterase was not reported in the study and presumably was not accomplished. The failure to administer neostigmine almost certainly underlay the lethality of the administration of curare, but in the opinion of the authors, most of the anesthesia-related deaths were from cardiovascular collapse, not respiratory insufficiency. This finding could have ended the clinical use of muscle relaxants, but Beecher and Todd wisely avoided simplistic conclusions. Although the use of relaxants increased anesthesia-related mortality, it did not appear to increase, indeed possibly decreased, overall operative mortality, a point argued forcefully by Abajian and colleagues in a rebuttal paper [34]. Beecher and Todd presciently reasoned that if curare caused an anesthesia-specific problem then the professionalism of anesthesiologists would compel them to investigate, understand, and resolve it.

The depolarizing action of succinylcholine was not immediately appreciated. William Paton delineated relaxants into depolarizing and non-depolarizing drugs in 1956 [35]. Nondepolarizing drugs had a bulky structure (were pachycurares); depolarizing drugs had a long narrow structure (were leptocurares). We are still not sure how relaxants work. Do they enter and plug up ion channels? Do they compete with acetylcholine for receptors on the motor nerve terminal? Do they block the release of acetylcholine? Probably all of the above, and more [36].

In the body, succinylcholine quickly degraded to the naturally occurring compounds choline and succinate. Their natural nature made them nearly undetectable and possibly made succinylcholine the “ideal murder weapon.” Lurid cases of suspected murder by succinylcholine have occurred. In 2007 a Nevada jury convicted critical care nurse, Chaz Higgs, of first degree murder in the death of his wife, Kathy Augustine, a prominent Nevada political figure. The murder weapon was a syringe of succinylcholine [37].

This decade saw structure-function analysis applied to drug design. The discovery of a steroidal alkaloid malouetine prompted development of steroid molecules with curare-like effects. In 1935, King had described the structure he thought was required in a neuromuscular blocking drug: two quaternary nitrogen groups separated by a distance (interonium) of 1.4 to 1.5 nm. The interonium distance proved to be more flexible than King predicted, for pancuronium being 1.05 nm and for fazadinium 0.7 nm [39,40]. The general structure had two acetylcholine-like groups separated by an appropriate distance, and had to be bulky (pachycurare) to confer non-depolarizing characteristics. William “Bill” Bowman was a gifted pharmacologist at the University of Strathclyde in Glasgow, Scotland, who had worked with Paton on curare-like drugs. Bowman considered that the rigid steroidal androstane molecule was the perfect skeleton on which to place the quaternary nitrogen groups to create a muscle relaxant drug.

It was these structural “rules” and Bowman’s idea to use the androstane nucleus that led to the development of the aminosteroid relaxant pancuronium by Hewett and Savage [41]. Pancuronium obeyed the rules except for its short interonium distance. This shortness conferred high potency and was long enough to avoid producing ganglionic blockade. Soon after the report of its clinical use in 1967 [42], pancuronium supplanted curare and gallamine for use in patients. Although devoid of ganglionic blocking effects, it had both vagolytic and sympathomimetic activity, but did not produce the hypotension associated with d-tubocurarine, or as much tachycardia as seen with gallamine.

The success of pancuronium heralded the emergence of West Scotland as the epicenter for muscle relaxant development for the next 25 years. Leading this effort was David Savage, the principal chemist-scientist working on these compounds at Organon’s research laboratories (Fig. 50.4). He had notable and inventive enthusiasm. He claimed, a bit (but just a bit) tongue in cheek, that using an androstane (steroid) backbone he could create a drug that would work at any receptor. He said that for fun he had made and tested a steroidal appetite suppressant in rats. It was all too good; the rats starved to death! He was also abon viveur. For one particular experiment he needed a small pressure vessel to contain the reaction. The ideal vessel turned out to be a champagne bottle, and of course the contents had to be drunk first.

Savage had the then novel insight that scientists and clinicians should work together to develop drugs clinicians needed. So he worked closely with anesthesiologists: Leslie Baird at Glasgow Royal Infirmary, and Ronald Miller in San Francisco. There is a story that one of the brain storming sessions occurred at a local bar. Purely by chance, David Savage and a local anesthesiologist struck up a conversation, and started discussing curare and the neuromuscular junction. In the course of the conversation many novel drug structures, probably including a pancuronium prototype, were written down on paper napkins. So engrossed were the two men that they had to be asked to leave several hours later, as the bartender wanted to close up. Many clinician colleagues of David Savage describe similar “design” sessions, where they would articulate their needs for a drug, and David would draw putative structures. To further scientist/clinician interaction, Savage organized the 1^st International Neuromuscular Meeting, bringing clinicians and scientists together in London in 1975.

Like its peers, pancuronium had too long a duration of action because its elimination was primarily via the kidney. Continued research at the Scottish facility solved this problem, leading to the development of two successful drugs, vecuronium and rocuronium, and one ill-fated drug, rapacuronium [43].

Up to the 1970s, clinicians had no practical way of monitoring relaxant effect, except for subjective clinical signs such as abdominal muscle relaxation or adequacy of spontaneous respiration. The need to develop a method of monitoring the residual effects of such powerful drugs was addressed in the next decade.

In the 1950s and 1960s, clinical acumen and judicious dosing guided the administration of muscle relaxants. The first monitors of neuromuscular function were simple nerve stimulators [44]. They required subcutaneous needles for nerve stimulation, The need to define a control response before relaxant administration made them impractical for normal clinical use. In the 1970s, this led Hassan Ali to develop the train-of-four ratio (TOF; Fig. 50.5) [45]. Ali used the known effect of repeated stimulation (repeated stimulation produces a decreasing contraction in a partly paralyzed muscle) in a clinically useful way [46]. Patients received four stimuli at 2 Hz (0.5 s intervals) and the ratio of the fourth response (T4) to the first (T1) was measured. A material decrease, especially below 0.7, indicated residual paralysis.

This useful ratio (Fig. 50.6) correlated with other indices of muscle function, including the absolute decrease in strength of the muscle contraction, the decrease in response to tetanic stimulation, and deteriorating pulmonary function [47,48]. Because it provided its own internal control, it did not require a baseline measurement. The response could be either mechanical force, or electromyography. It also discriminated between succinylcholine-induced block (minimal fade) and depolarizing (marked fade) block. TOF assessment persists as the mainstay of clinical monitoring.

In the 1970s, clinicians could choose a short-acting drug (succinylcholine) or longer acting (nondepolarizing) drugs. All had side effects. This unsatisfactory situation led to the concept of the ideal drug by Karis and Gissen in 1971 [49]. Their criteria have stood the test of time and are still quoted today (Table 50.2), with the notable exception that a demand for fast onset was absent.

Savarese and Kitz elaborated on these concepts in 1975 [50], adding their belief that one drug could not be ideal for all clinical situations. They defined three categories of ideal drug, A, B and C. Type A was a non-depolarizing succinylcholine (rapid onset, rapid offset). Type B had a duration of action approximately half that of curare or pancuronium, and soon became a reality with the introduction of vecuronium and atracurium. Type C was long acting and without adverse cardiovascular effects, anticipating pipecuronium and doxacurium which arrived approximately 15 years later. John Savarese has crusaded for a Type A drug, and, as we shall see, his efforts may be close to fruition.

Defining the direction for new drugs pressured the pharmaceutical industry to synthesize products that would meet the challenges implied by the definitions. As noted earlier, Savage facilitated the dialogue between clinicians and industry by organizing the 1^st Neuromuscular Meeting in London in 1975, just when the concept of the “ideal relaxant” was described. Few at the meeting appreciated the impending “revolution” in neuromuscular pharmacology’the era of the intermediate-acting relaxant was about to dawn.

Until the 1970s, there was but a rudimentary understanding of drug actions, and of the interaction of biodisposition (pharmacokinetics) and effect (pharmacodynamics). Developments at the University of California San Francisco (UCSF) changed this (Fig. 50.7). Leslie Benet developed and refined the science of pharmacokinetics and pharmacodynamics, creating the concept of drug “Clearance.” Concurrently, Lewis Sheiner pursued the science of drug concentration-effect relationships. And a young clinician/scientist Ronald Miller, freshly back from Vietnam, had a mind full of questions relating to the clinical behavior of muscle relaxants and their antagonists. These circumstances fortuitously combined a clinician with important questions and scientists with the talents to develop the tools to address those questions. The breakthrough was to measure both the effect of a drug and the concentrations associated with the effect, and to make inferences as to how underlying conditions (e.g., renal or hepatic failure) influenced these. The first paper applying these techniques focused on the impact of renal failure on the pharmacokinetics of tubocurare [51].

A problem remained: the concurrent measurement of the concentration of a drug at its site of action (e.g., the neuromuscular junction) and the effect of that concentration. The researcher couldn’t sample the neuromuscular junction. Surrogate measures (e.g., the concentration in blood) had to serve instead. That was fine (sort of) if there were equilibrium between the blood and the neuromuscular junction, but how could one be sure? Sheiner and Stanski published an important paper in 1979 that provided the “missing link” between concentration and effect [52]. The authors in this paper recognized that initially the concentration at the neuromuscular junction (and the effect) lagged behind what the anesthetist gave and was in the blood. During recovery, the opposite was true. The paper demonstrated how one could calculate the concentration and effect at the neuromuscular junction by knowing the effect during input and output. This model has broad applicability to any situation where the drug concentration is measured in one place and its effect takes place at another.