A History of Intravenous Anesthesia

Fig. 47.1
This apparatus was designed by the Kny-Scheerer Co. for use with any type of IV anesthetic drug. As described in a report on Intravenous Anesthesia by two New York surgeons, William Honan and J Wylis Hassler, read at the American Association of Anaesthetists meeting in June, 1913: “the narrow tube by the side of the reservoir was for the introduction of hedonal or isopral preliminary to the employment of a solution of ether.”

Hedonal, Ether, Paraldehyde, Magnesium, and Alcohol

In 1909, Russian pharmacologist Nicholas Krawkow (1865–1924) demonstrated the anesthetic properties of hedonal in St. Petersburg. Hedonal was a urethane derivative used primarily to treat insomnia. It was the first agent given intravenously that produced surgical anesthesia with “a moderate degree of safety” [7]. Unfortunately, hedonal dissolved minimally in water, had a slow onset, and a long duration of hypnotic action. In 1909, Ludwig Burkhardt in Wurzburg, Germany, described IV administration of ether and chloroform for surgical anesthesia [8]. Subsequently, in 1911, Felix Rood in England gave a 5% solution of ether in saline intravenously to 18 patients [9], reporting that it enabled great flexibility of anesthetic depth, and was accompanied by minimal postoperative vomiting or pulmonary irritation. In 1913, HLC Noel and HS Souttar described the anesthetic effects of IV paraldehyde (Paral) [10], and shortly after, CH Peck and SJ Meltzer described the anesthetic effects of magnesium sulfate given by continuous IV infusion [11]. K Naragawa used ethyl alcohol for IV anesthesia in 1921 [12], as did H Cardot and H Laugier in 1922 [13]. However, none of these anesthetics became popular in clinical practice.

The First Effective Barbiturates

In 1903, Nobel laureate Emil Fischer (1852–1919) and Joseph von Mering (1849–1908), two German chemists working at Bayer, synthesized the first derivative of barbituric acid (diethyl-barbituric acid) possessing sedative and hypnotic properties, an oral sedative named “Veronal”, named after Mt Verona where Fischer had recently spent a holiday [14]. In 1921, the French anesthesiologist, Daniel Bardet reported the first clinical use of an IV barbiturate, Somnifen, a mixture of diethyl-barbituric acid (Veronal) and diallyl-barbituric acid (Dial), for induction of anesthesia [15]. However, low solubility in water and a long duration of hypnotic action limited the clinical use of Somnifen. In 1929, JH Fjelde reported his use of a smaller dose of Somnifen for ‘induction’ before ethylene anesthesia [16].

“The patient is put to sleep right in his room, a distinct advantage which further relieves him of further thoughts of the coming operation. When awakening they are back in their own bed and have no recollection of having been to the operating room at all. A wonderful feature (of the drug)!”

Introduced in 1927, a more water-soluble barbituric acid derivative called Pernoston (originally Pernocton), gained acceptance for induction of general anesthesia’as part of the increasing acceptance of intravenous induction of anesthesia [17]. H Weese and W Scharpff synthesized the first rapid and short-acting barbiturate, hexobarbital (Evipan, Evipal) in 1932 [18]. Despite producing prominent excitatory side effects, hexobarbital was widely used in Europe. LG Zerfas described the intravenous use of soluble sodium amytal (amobarbital) in 1930 [19]. Although Zerfas cautioned users regarding the prominent cardiorespiratory depressant properties of amobarbital, it became the most popular intravenous anesthetic in North America until the introduction of thiopental.

Thiopental and Methohexital

Fischer and von Mering described the sulfur-containing barbiturates in 1903 [17]. They administered a single dose to a dog which promptly died, leading them to conclude that sulfur made the compound toxic. They immediately discontinued this avenue of research. In the early 1930s, Ernest Volwiler and Donalee Tabern, working at Abbott Laboratories, resumed studies on the sulfur-containing barbiturates. After trials in animals, two compounds were presented to the American anesthesiologist John Lundy (1894–1973) to investigate at the Mayo Clinic, namely sodium allyl-sec-butyl-thiobarbiturate (also known as thiosebutal), and sodium ethyl (1-methyl-butyl)-thiobarbiturate, a sulfur derivative of pentobarbital (Nembutal), which was called thionembutal or “Hypnotic 8064”. In 1934, Lundy and Ralph Tovell reported their initial clinical experience with thionembutal in 2207 patients [20]. They observed that rapid injection of thionembutal led to respiratory depression which was monitored by taping a fluffy piece of cotton to the patient’s upper lip and using the movement of the cotton to measure the respiratory rate. This monitoring technique was subsequently referred to as “Lundy’s butterfly”. Two years later, Ralph Waters’ group in Madison, Wisconsin published results of their study of thionembutal [21], subsequently renamed sodium thiopental or thiopentone [Pentothal]). Thiopental rapidly replaced other barbiturates for intravenous induction of anesthesia, particularly in the Americas.



Anesthesia folklore suggested that after the Japanese attack on Pearl Harbor in 1941, “more USA servicemen died from thiopental than war-related injuries”, but no direct evidence supports that assertion. It is true that large doses of thiopental administered to hypovolemic but otherwise healthy traumatized patients could produce profound hypotension. When administered as the sole anesthetic during World War II, the cardiorespiratory depressant effects led to deaths, and FJ Halford, a surgeon, suggested that IV anesthesia with thiopental was “an ideal method for euthanasia” [22]. He also wrote “While intravenous anesthesia would seem ideal for war injuries because of its compactness, ease of preparation, and non-explosive characteristics, it should clearly be recognized that under war conditions, anesthetics cannot always be given by highly skilled anesthesiologists, but have to be given by doctors, nurses and even orderlies, for whom the art is strange….” What a wonderfully perceptive statement! But an analysis by FE Bennetts [23] suggested that the number killed, missing or fatally injured as a result of the attack on Pearl Harbor was 2403 and despite shortages of oxygen and transfusion supplies, thiopental caused few deaths.

The German pharmacologist Helmut Weese (1897–1954), the director of research at the Bayer Co., introduced another thiobarbiturate, known as Baytenal or buthalitone, with a ‘brief’ recovery time. In Britain, it was marketed as Transithal or Ulbreval. As with hexobarbital, a high incidence of involuntary muscle movement led to its abandonment. In Germany, ethyl-secbutyl-thiobarbiturate, known as Inaktin, produced actions similar to thiopental and was briefly popular in the 1950s. Pharmacologists Zima, von Werder and Hotovy attempted to accelerate the breakdown of the barbiturate molecule by substituting a methyl thioethyl group at position five [24]. The resulting drug, called Nercval in the US and Thiogenal in Germany, was very short-acting but less effective than thiopental as an induction agent.

The rapid onset of hexobarbital’s action was considered to be due to the methyl group on the molecule. This stimulated further research into methylated barbiturates, and researchers at Eli Lilly Laboratories developed Lilly 22,451 in 1956. The drug showed greater potency and faster recovery than thiopental, but also possessed convulsive properties. Further modification led to Lilly 25,398, known as methohexital (Brevital). Methohexital continues as a popular IV anesthetic for dental and short procedures, in particular where less suppression of CNS electrical activity is desirable (e.g., surgery for seizure disorders and electroconvulsive therapy). Although methohexital allowed a more rapid recovery of consciousness than thiopental (or its chemical analog, thiamylal, also known as Surital), the occurrence of excitatory side effects made methohexital less attractive for short-lasting surgical procedures following the discovery of proprofol. Despite its well-known cardiovascular and respiratory depressant side effects, thiopental dominated IV anesthesia from the mid-1930s until the introduction of propofol in the mid-1980s. In 1982, Dundee and McIlroy wrote “Thiopentone…‥will be hard to replace by non-barbiturate agents… methohexitone still holds a place for brief procedures, but this agent may be replaced by a shorter-acting non-barbiturate” [14]. They did not mention propofol.

Steroid Anesthetics

In the early 1940s, Hans Selye (1907–82) discovered that glucocorticoid steroids such as progesterone could produce sleep in rodents [25]. Selye found anesthetic properties required the presence of an oxygen atom at either end of the steroid molecule. While pregnanedione was the most potent of the steroids tested, it was poorly soluble in water and irritated veins’and never reached clinical practice. Hydroxydione (marketed as Viadril and Presuren), a water-soluble steroid, came into clinical practice in the mid-1950s. However, its slow onset (3–5 min), high incidence of thrombophlebitis, and prolonged recovery times limited acceptance.


IV Inhaled Anesthetics

As Burkhardt, Rood, Honan and Hassler reported in the early 1900s [810], inhaled (volatile) anesthetics such as diethyl ether and chloroform could be given intravenously. A half century later, the modern ‘Godfather’ of inhalation anesthesia, Edmond (“Ted”) Eger from UCSF, confirmed these observations [26]. Although this route of administration avoided airway irritation, the anesthetic state it produced was no better than that produced by inhalation of ether, and required fluid volumes occasionally producing pulmonary edema. Although interest in IV infusion of volatile anesthetics recently returned [27], it is unlikely to gain widespread clinical acceptance in the future.


Henry Price (1922–2002) pioneered the effort to connect the pharmacokinetic (PK) effects of IV anesthetics to their pharmacodynamic (PD) effects. In 1960, he suggested that redistribution of thiopental from vessel-rich tissues (e.g., brain) to less well-perfused tissues (especially muscle)’rather than elimination of the drug from the body or uptake by fat’governed the initial rapid awakening from thiopental [28]. Cedric Prys-Roberts and Chris Hull in England, and Don Stanski in the US, used these pharmacokinetic-dynamic concepts to expand our understanding of the clinical properties of newer IV drugs (see Chapter 40).

Thiopental was used to induce a deep coma-like state in patients with traumatic head injuries (or strokes), the rationale being that thiopental possessed ‘brain protecting effects’ because of its ability to decrease cerebral metabolism and intra-cranial pressure. In 1962, Ellison (“Jeep”) Pierce and colleagues at the University of Pennsylvania demonstrated that thiopental proportionally decreased cerebral metabolic rate and cerebral blood flow, and increased cerebral vascular resistance [29]. In 1973, Jack Michenfelder (1931–2004) and Richard Theye (1923–1977) at the Mayo Clinic, showed that thiopental diminished cerebral energy requirements by reducing metabolic activity, thereby affording some degree of protection of the functioning brain (i.e., the brain in which thiopental could decrease metabolic rate) [30]. Unfortunately, in the absence of brain function (i.e., in the case of severe [global] hypoxic brain injury), thiopental provided no improvement in clinical outcome. Nevertheless, thiopental-induced coma remains a common treatment for acutely head-injured and post-cardiac arrest patients.


The cardio-respiratory depressant effects of thiopental prompted a search for a more ‘ideal’ IV anesthetic, one having thiopental’s good characteristics but not its limitations. In 1964, Paul Janssen (1926–2003) and colleagues reported their synthesis of etomidate, a new IV anesthetic, with a chemical structure unrelated to that of any other IV anesthetic. Alfred Doenicke described early clinical studies with etomidate in 1974 [31]. An aqueous solution of etomidate (Amidate) was unstable at physiologic pH, and etomidate was formulated as a 0.2% solution with 35% propylene glycol at a pH of 6.9, a combination producing pain on injection. In 2006, a lipid emulsion formulation (Etomidate-Lipuro) was introduced in Europe which produced less pain on injection [32]. Etomidate advocates pointed to a major advantage compared to thiopental: minimal cardio-respiratory depression even in patients with clinically-significant cardiovascular disease [33]. Many anesthetists considered it to be the induction agent of choice for high-risk patients who would benefit from preservation of a normal blood pressure during induction of anesthesia (i.e., patients with clinically-significant cardiac and cerebrovascular disease), from so called ‘stress-free’ anesthesia. My research group at Stanford University subsequently demonstrated that etomidate’s ability to maintain a ‘stress-free’ state was not due to its central effect, but rather to a peripheral action, namely inhibition of adrenal steroidogenesis.


A chance observation that eugenol, a centrally-active compound derived from the oil of cloves and cinnamon exhibited anesthetic properties, led Jean Thuillier and Robert Domenjoz to use a congener of eugenol for IV anesthesia [34]. Propanidid [Epontol, Fabontal], a phenoxyacetic acid derivative of eugenol, was introduced into clinical practice in Europe in the 1960s as the first non-barbiturate IV anesthetic. This so-called ultra short-acting anesthetic became popular for brief day-surgery (ambulatory) procedures. Unfortunately, this poorly water-soluble compound was formulated in a Cremophor EL solution causing hypersensitivity reactions that led to the withdrawal of propanidid from clinical practice.

Steroid Anesthetics

In the early 1970s, Glaxo Research examined many pregnanedione derivatives, eventually leading to the development of CT-1341, a mixture of two steroids, alphaxalone and alphadolone in a Cremophor EL solution, and eventually marketed in 1973 as Althesin (also Alfathesin) [35]. Alphaxalone provided the anesthetic activity, and alphadolone enhanced the solubility of alphaxolone. Althesin became a popular anesthetic in Europe and elsewhere for short day-surgery procedures, but was never marketed in the US. Its brief duration of action facilitated control over the hypnotic state when administered as a continuous infusion for maintenance of anesthesia [36]. Unfortunately, as with propanidid, the Cremophor EL solubilizing agent produced rare but dangerous hypersensitivity reactions, resulting in Althesin’s withdrawal from the market in 1984.

Minoxalone, a water-soluble, rapid-acting steroid was developed in 1979. However, excitatory reactions and more prolonged recovery times limited its use. Most recently, the steroid anesthetic pregnanolone (eltanolone), a naturally occurring metabolite of progesterone, solubilized in the lipid emulsion used to dissolve propofol, was tested for its anesthetic properties. Although it rapidly induced an hypnotic state [37], eltanolone produced longer recovery times than propofol, and it was never approved for clinical use.

Phencyclidine (PCP) and Ketamine

PCP was synthesized in 1928. Investigations into the clinical use of cyclohexylamines began in 1958, with two compounds produced by the Parke-Davis Company, CI 395 and CI 400. Both drugs produced disorientation, agitation and hallucinations. According to Vincent Collins, CI 400 produced “reverbigeration” (i.e., selecting a phrase that would then be repeated rapidly, sometimes for the entire duration of the procedure) [38]. Although rejected for use in humans because it produced hallucinations, mania, delirium and disorientation, CI 395 (Sernyl, [the name it was given prior to clinical testing is said to be derived from ‘serenity’]) was approved for use in veterinary medicine in 1962 [39].

Calvin Stevens, working at Parke Davis Laboratories, synthesized the ‘sister’ drug of Sernyl known as CI 581, later called ketamine (Ketalar or Ketaject). After relocating to the US and entering the practice of anesthesia, Guenter Corssen (1916–1990) became an active investigator and passionate advocate of new IV anesthetic drugs and delivery systems. In 1965, he and Ed Domino (father of Karen Domino (author of chapter 18 in this book) at the University of Michigan first used ketamine clinically [40]. They noted that “CI 581 appears to produce a state resembling cataplexy where patients felt as though they were in outer space, or had no arms and legs.” They proposed the term “dissociative anesthesia” to describe an unconscious state with profound analgesia but devoid of the cardio-respiratory depressant effects associated with the commonly used barbiturate compounds [41].

In the early 1970s, as a graduate student in pharmacology at the University of California, San Francisco, I became interested in IV anesthesia, initially investigating the pharmacologic properties of ketamine and its interactions with the volatile anesthetics [42]. Ketamine was a racemic mixture of two optical isomers and in a preliminary clinical trial we performed at UCSF in the late 1970s, it was found that the S(+) isomer possessed more potent hypnotic and analgesic properties than the R(−) isomer [43]. The S(+) enantiomer also produced less prominent psychomimetic properties and allowed a faster recovery of cognitive functioning [44].

Subanesthetic doses of ketamine produced profound analgesia and amnesia without suppressing consciousness and protective reflexes, useful properties in patients having emergency surgery. Analogous to PCP, anesthetic doses of ketamine could produce untoward psychomimetic side effects. A surgical colleague asked a young Eger to anesthetize an elderly VIP with diabetes and significant vascular disease, for debridement of a gangrenous big toe. Eger suggested to the patient that a spinal anesthetic might be just what was needed, but the patient demurred, observing that the surgeon had suggested this new anesthetic, called ketamine. Prompted by youth, arrogance, and an absence of any experience with ketamine, Eger agreed, supplying what generously might be called an “uneven” anesthetic. Nonetheless, he and the patient survived. A month later, his surgical colleague called on Eger to again anesthetize this patient. Eger had barely entered the room for his preoperative visit when the patient pointed his finger at him and said “Don’t you give me that stuff again. I wasn’t right in my head for two days.”

Luis Vasconez (a prominent plastic surgeon at UCSF) and I, developed a ketamine-based sedation-analgesic technique for patients having cosmetic surgery procedures [45]. The technique consisted of an intravenous benzodiazepine followed by sub-hypnotic doses of IV ketamine in combination with local analgesia. With this technique, patients breathed spontaneously and maintained good oxygenation without requiring ventilatory assistance or supplemental oxygen. Many patients reported vivid dreams (e.g., one elderly San Francisco matriarch commented in a beautifully composed ‘Thank You’ card [that had a flock of multi-colored birds flying through the sky on the cover] that she felt she was flying above the operating table throughout the procedure). Although some patients reported dreams akin to “near death” experiences (e.g., descending down a dark tunnel with a bright light at the end), most patients found the dreams interesting and many reported pleasant experiences.

In 1999, Manzo Suzuki and colleagues described the opioid-sparing effects of low-doses of IV ketamine (50–100 µg/kg) given to supplement opioid analgesics [46]. More recently, several investigative groups have described the use of 100–200 µg/kg to minimize or replace opioids as part of an IV anesthestic technique, thereby decreasing respiratory depression [47]. These small doses of ketamine produce analgesia without the untoward psychological sequelae associated with ‘anesthetic’ doses.


In 1961, Sternbach and Reeder at Hoffman-La Roche Laboratories synthesized the first IV benzodiazepine’diazepam (Valium). Released for clinical use in 1963, psychiatrists used it to relieve chronic anxiety, neurologists to control seizure disorders, and anesthesiologists to relieve acute situational anxiety prior to surgery. In 1965, Stovner and Endresen in Oslo reported on its use IV as an induction agent [48]. However, diazepam’s solubilizing agent, proplylene glycol, produced pain on injection and venous irritation (i.e., thrombophlebitis). These properties and a prolonged recovery after short surgical procedures limited its popularity.

In 1971, lorazepam (Ativan, Temesta) was developed at Wyeth Laboratories and introduced into anesthesia practice for use as a premedication, but this benzodiazepine produced prolonged postoperative amnesia and fell into disfavor when elective surgery shifted from a hospital-based to an ambulatory setting. As with diazepam, lorazepam produced significant pain and venous irritation following IV administration.

In 1976, Rodney Freyer and Armin Walser at Hoffman-LaRoche Laboratories synthesized midazolam (Versed or Dormicum), a water-soluble IV benzodiazepine. Jerry Reves and his associates at the University of Alabama conducted the first trials in 1978, concluding that “Ro 21-3981 is a promising intravenous anesthetic agent. It appears superior to the closely related benzodiazepine, diazepam, in that it is more potent, less irritating, and shorter-acting” [49]. However, induction of anesthesia with midazolam produced a more prolonged recovery than did thiopental (or ketamine) [50], and it never became popular for that purpose. Nevertheless, it quickly became widely used as a sedative-hypnotic agent for premedication and IV sedation.

At a meeting held on Hilton Head Island immediately before the launch of midazolam, Roche’s anesthesia advisory board warned the company that the availability of the higher concentration (5 mg/ml) of midazolam used for induction of anesthesia might lead to excessive dosing of the drug when given for IV sedation. The company ignored these warnings, and the FDA almost withdrew approval of this valuable drug following reports of multiple deaths due to ventilatory depression after IV administration (at doses similar to those used with diazepam) for sedation by non-anesthesiologists. These clinicians were unaware of the more rapid onset, greater potency, ‘steeper’ dose-response curve, and more profound respiratory depressant effects of midazolam compared to diazepam [45].

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Mar 21, 2017 | Posted by in ANESTHESIA | Comments Off on A History of Intravenous Anesthesia

Full access? Get Clinical Tree

Get Clinical Tree app for offline access