How John Cade, superintendent of the Bundoora Repatriation Hospital in a suburb of Melbourne, made his discovery of the almost wondrous effectiveness of lithium in mania is familiar and won’t be repeated here. Full-blast mania is a terrible illness that in former times was often fatal, as the patients would simply agitate themselves to death. A treatment had arrived for it! Yet Cade’s article in the Medical Journal of Australia in 1949 had almost no international impact. At the time, lithium had recently come into bad odor as a cardiac drug, and most psychiatrists had little interest in prescribing it.

Other isolated and little-noticed reports followed Cade’s article. It was actually to the young Danish psychiatrist Mogens Schou (pronounced “Sk-oh” as in “hoe”) that credit is owing for putting lithium on the international radar. In 1952 Erik Stromgren, head of the Aarhus university psychiatric clinic in Risskov, suggested to Schou that he look into Cade’s findings. There had been a tradition in Danish psychiatry of using lithium that went back to physician Carl Georg Lange’s work in 1886. Schou had a particular interest in the subject, as his father, psychiatrist Hans Jacob Schou, had been interested in manic-depressive illness, and a history of the illness ran in Schou’s family. With the help of colleagues, Schou organized a double-blind trial, randomizing the patients to lithium or placebo with the flip of a coin; the results were published in a British journal in 1954. It was one of the early randomized trials in psychiatry. Previously, Schou said, the only effective treatment of mania had been ECT. “The lithium therapy appears to offer a useful alternative since many patients can be kept in a normal state by administration of a maintenance dose.”

Yet this important finding was almost drowned in a sea of British doubt. The most influential training center in the United Kingdom since the 1930s has been the Maudsley Hospital in London, where after the Second World War Aubrey Lewis was professor of psychiatry and Michael Shepherd an influential senior figure. The Maudsley had a historic tradition of skepticism about pharmacotherapy, and emphasized social and community, not biological, psychiatry. Lewis and Shepherd both were said to have scorned lithium treatment as “dangerous nonsense.” Lithium was in fact tricky to use without experience, which is probably the reason the revolution in psychopharmacology began with chlorpromazine and not lithium. But these English clinicians exaggerated its dangerousness. Shepherd sneered that Schou had so much faith in lithium that he put his whole family on it. As for Schou’s concept of “the prophylaxis of depression,” Shepherd considered it risible. Alec Coppen, at a different hospital across London, sought to explain these attitudes: “A lot of people who are in psychiatry are not really interested in the medical model. They went into psychiatry to get away from it… Psychiatric illnesses are seen as a sort of… social illness that should be treated by social methods.” Coppen called this “out-of-date science dating back to the 1950s.” But it was alive and well at the Maudsley, and threw a blanket of doubt over early lithium research.

Lithium arrived on the Continent earlier than in the United States. Of the main players, France was first to license lithium, as lithium gluco-nate, in 1961. For years previously, the French had swallowed “Docteur Gustin’s Lithium,” which, one observer speculated, was why you don’t have a lot of manic-depressed patients in Marseille.

Lithium’s arrival in the United States occurred in the late 1950s, when Sydney-educated Samuel Gershon imported the lithium concept from Australia while working in Ralph Gerard’s big program at Ypsi-lanti State Hospital in Michigan on the biology of schizophrenia, sponsored by the National Institute of Mental Health and the University of Michigan. Gershon and Edward Kingstone, a young Canadian psychiatrist at the Allan Memorial Institute in Montreal, published almost simultaneously in 1960 on lithium in mania.

Even before the appearance of these articles, in 1959 Ronald Fieve, a resident at the New York State Psychiatric Institute, became fascinated with Cade’s and Schou’s work. Aided by a Coleman flame photometer (introduced in 1958 to measure blood levels of this potentially quite toxic substance), Fieve and another resident began an open-label trial with lithium, which they procured from a chemical corporation in Connecticut. Fieve later wrote, “In 1964, when lithium was still commercially unavailable in the U.S., a psychiatric group at the University of Texas had heard of my work, and sent a manic professor to see me. He was psychotic, working on 40 papers and 2 books, euphoric, overtalkative and charming. After I treated him for several weeks with lithium, he returned to Texas completely normal in mood and behavior.” In 1966 Fieve reported on 19 manic patients they had treated with lithium: 44 percent “had good responses.” Two years later Fieve and coworkers said lithium had “a mild antidepressant effect” as well.

To put lithium into patients, of course, it was necessary to get an IND from the FDA. Jonathan Cole was one of two clinicians in the Boston area with an IND for lithium, and was among the several psychiatrists who got Merle Gibson, head of psychopharmacology at the FDA, interested in approving the drug for broader use through the New Drug Application (NDA) process. Cole said, “The Rowell company, a pharmaceutical company in northern Minnesota [Baudette], was the first drug company to be interested in making up lithium carbonate in capsules for double blind placebo and whatnot. They were a small company that sat on a lake [Lake of the Woods] in northern Minnesota because some fish in the lake produced something like cod liver oil, and so they were way ahead of any other company… Then finally the FDA talked Smith Kline French and Pfizer into marketing the drug.” The FDA held up approval of Rowell’s NDA until the two big companies were ready, and all three manufacturers’ lithium carbonate entered the market in 1970.^l

What Cole didn’t say, but is also true, is that in 1969 the FDA greatly agonized about the approval of lithium, especially whether it should be used for long-term prophylaxis after an acute episode, or only for acute mania. As one insider put it, “there was a good deal of caution and a good deal of worry” about its long-term effects. It was thus a systematic lobbying campaign by Frank Ayd and Oregon psychiatrist Paul Blachly, who threatened to prescribe it approved or not, that put lithium over the top at the FDA.

Over the years, no drug has made as much of a difference in the lives of patients with manic-depressive illness, now known as bipolar disorder, as lithium; it not only effectively treats the acute manic phase of the illness, but it also prevents relapse from both mania and depression. In 1971 an English group led by Alec Coppen established the extraordinary effectiveness of lithium in the prophylaxis of unipolar as well as bipolar depression. It’s not clear that these really are two different types of depression, and readers should take this distinction with a grain of salt. Melancholia is melancholia, whatever its polarity.

Nonetheless, the distinction was thought important at the time. In 2000, Sam Gershon, looking back on the launch of lithium 30 years previously, said, “Lithium is in general the most effective medication for acute mania, producing improvement in about 70 to 80 percent of cases.” Its effectiveness in the prophylaxis of bipolar disorder was also beyond doubt, he said. In short, “lithium is the most effective agent in the treatment of bipolar disorder.” Gershon made this comment at a time when the shelves of the pharmacies were becoming crowded with other mood stabilizers for the exploding number of people, including children, being diagnosed with bipolar disorder.

In retrospect, “The miracle of lithium was not its treatment of acute mania,” as Dennis Charney, a veteran psychopharmacologist then at Yale University, put it in 1995 at a meeting of the Psychopharmacologic Drugs Advisory Committee of the FDA. “Neuroleptics, and even high-dose benzodiazepines, are quite effective for the treatment of acute mania… The issue is prevention of relapse.” The committee was meeting to consider Abbott’s application for the approval of its mood-stabilizer drug Depakote (divalproex sodium) for bipolar disorder. Charney thought it appropriate that this fact about lithium should be mentioned in the Depakote label.

It wasn’t.

In the late 1940s the Geigy Company in Basel was following the same antihistamine trail that produced chlorpromazine. But not wanting just to duplicate the phenothiazine nucleus, in 1949 they took an “imminodibenzyl nucleus” (first synthesized in 1898, with an ethylene in place of the sulfur atom), then played around with the side chains, and put some of the resulting compounds into trials as hypnotics. Among other trialists, Geigy was in contact with Roland Kuhn, a psychiatrist at the Munsterlingen Cantonal Asylum (Thurgau Canton) in Switzerland. Kuhn, in his early 40s, had a strong interest in psychotherapy and philosophy and was anything but a psychopharmacologist. He had been losing confidence in psychoanalysis, and began to realize that many “neurotic” or “psychosomatic” patients were really depressed. But how to help them? Admitting them to hospital and doing ECT seemed a bit much to him. “How often I thought,” he later said, “we should improve the opium treatment. But how?”

Mindful of the success of chlorpromazine, in 1953 Kuhn contacted Robert Domenjoz, head of pharmacology at Geigy, and asked him if one of the earlier imminodibenzyl compounds they had looked at (G 22150) might be worth a second look. Yet 2 years later, in March 1955, Kuhn told Domenjoz that G 22150 had been ineffective in schizophrenia and manic-depressive illness.

Sometime later that year, Domenjoz met Kuhn in a hotel in Zurich and showed him “a large sheet of paper on which were drawn about 40 chemical formulas of the substances which were available for clinical evaluation.” Kuhn said he wanted the molecule with the same side chain as chlorpromazine, and selected it from the chart. It was G 22355. Over the next year and a half Kuhn subjected G 22355 to trials for a number of different indications. It had some success in psychosis, but it showed an even more important feature: After putting G 22355 into around 150 patients, Kuhn wrote Domenjoz in August 1956 that it “has an obvious effect on depression. The vital depression visibly improves.” Vital depression, a concept introduced by German psychiatrist Kurt Schneider in 1920, was a form of melancholia with severe physical sensations of pain and fatigue. Kuhn went on to tell Domenjoz, “The patients feel less tired, the sensation of weight decreases, the inhibitions become less pronounced and the mood improves.” Kuhn had, essentially, discovered the effectiveness of an important new agent in melancholia.

Yet Geigy was not really interested in Kuhn’s discovery. Domenjoz remained fixated upon schizophrenia, and sent G 22355 out to a number of other trialists for further exploration as an antischizophrenic, saying nothing about Kuhn’s discovery in depression.

In the meantime, the Second World Congress of Psychiatry loomed, to be held in Zurich in September 1957 (the first had been in Paris in 1950). Kuhn was invited to contribute a paper and wrote up his findings about G 22355, which Geigy had brand-named Tofranil, generically imipramine. When Kuhn gave his lecture at the congress, there were perhaps a dozen people in the audience and no questions were asked afterward. Among those present was Frank Ayd, who later said,

Well, it was dramatic. Kuhn is a rather tall man, slender, very soft spoken, very cultured, very dignified and very erudite…. He didn’t say “this is a good antidepressant.” He said “this is a good drug for depressed patients who have these symptoms…” I’m not sure how many people in that room appreciated that we were hearing the first announcement of a drug that was going to revolutionize the treatment of affective disorders — and do more than that. If one thinks of what imipramine can do, it’s not just an antidepressant, it’s an anxiolytic, it’s an anti-panic. We would have never had all these things if Kuhn hadn’t given a very lucid and convincing paper. I’ll tell you… you want to read the English translation of his paper — it’s as good as the Gettysburg address.

Published in the first week of September 1957 in the Swiss Medical Weekly, the paper reported a striking effect on depressive symptoms: “The patients become generally livelier, their depressive whisper voices become louder, the patients appear more social, the yammering and crying come to an end.”

When Tofranil was launched in Switzerland late in 1957 — in the United States in 1959 — it was described as a “thymoleptic,” literally drugs that “take hold of the mood,” a term long in use in Spanish psychiatry that Geigy had also been employing internally. Geigy still did not really get behind the drug, however, and the next year, 1958, when Kuhn attended a psychopharmacology conference in Rome, the company representatives basically cold-shouldered him. The turning point came only when at the conference Robert Boehringer, one of the prominent Geigy shareholders (and an eminence grise in the firm), asked Kuhn if there was anything that might serve for his wife at home ill with depression. Kuhn recommended Tofranil; she readily recovered, and thereafter Geigy warmed to the drug.

That Rome conference, the first International Congress of Neuro-Pharmacology in September 1958, was interesting for another reason. There, a group from the Parisian Val-de-Grace military hospital presented a paper in which they reported finding in their depressed patients who had been given imipramine “a resumption of contact with the outside world, a suppression of internal inhibition, liberation from anguish [by which the French usually meant somatic anxiety].” A year earlier, Geigy had given Jean Delay and his disciples at the Ste Anne Mental Hospital in Paris a supply of imipramine to try, but having never heard of Geigy, Delay evidently dismissed the drug from his mind. Upon seeing at the Rome conference that they had been neatly scooped by the Val-de-Grace group, Delay is said to have chewed his assistant Pierre Deniker out roundly in public for having had the drug for a year and not conducting trials.

Internationally, the first big imipramine trial was Heinz Lehmann’s in Montreal, published in October 1958. Lehmann had little use for statistics, preferring, like Kuhn, the close daily observation of his patients (rather than the use of rating scales in large trials). Lehmann concluded of his results: “In the setting of a closed hospital we were able to give effective relief to most of our depressed patients with the drug alone [as opposed to drug plus ECT], even if they were greatly disturbed.”

In England, it was psychoanalyst Hilda Abraham, the granddaughter of Freud-intimate Karl Abraham, who conducted one of the first imipramine trials. She told Alan Broadhurst, who worked for Geigy’s English branch, “I haven’t ever really thought about using medication in the treatment of depression, can we talk about it?”

In the United States, Joe Schildkraut remembers trying imipramine at the psychoanalytically oriented Massachusetts Mental Health Hospital in those years. “These drugs seemed like magic to me,” he said. “I became aware that there was a new world out there, a world of psychiatry informed by pharmacology.” Said Don Klein of an early trial at Hillside Hospital in Queens, “We assumed [imipramine] would be some sort of supercocaine, blasting the patients out of their rut. Remarkably, these anhedonic, anorexic, insomniac patients began to sleep better, eat better, after several weeks … saying ‘the veil has been lifted.’”

Later in 1958, Jean Delay and Pierre Deniker did get around to conducting a controlled trial comparing imipramine to the MAOI iproniazid (Marsilid). In therapeutic terms, they showed about equivalent results: About two-thirds of “depressed” patients improved on each drug, although the responders to the oxidase inhibitors might have been a different group from the responders to the tricyclic antidepressants such as imipramine (which acted, it was later discovered, by inhibiting the reuptake of neurotransmitters). The investigators found among their 137 depressed outpatients that 74 percent of those with “melancholic depression” got better on imipramine. What to call this new drug? Delay and Deniker scorned Geigy’s label thymoleptic and preferred instead “psycho-analeptics,” or mood stimulants. At a March 1959 conference at McGill University in Montreal, they said, “It seems appropriate to class these drugs together as anti-depressive agents falling within the category of psycho-analeptics.” Psychoanaleptic did not catch on any better than thymoleptic, but antidepressants did.

There were numerous other small trials, most of which were uncontrolled, a matter of secondary importance because even the few controlled trials enrolled so few patients as to make the comparison with placebo meaningless. The whole concept of statistical “power” had not yet caught on in psychiatry (as opposed to “statistical significance” — all these little trials were “significant”). Yet for what it’s worth, in 1965 Jonathan Cole, director of the National Institute of Mental Health’s Psychopharmacology Service Center, which undertook large controlled trials of new drugs, and Gerald Klerman, previously a colleague of Cole’s at the PSC and now a faculty member at Harvard, undertook an overview of the 23 controlled studies that by then had been published on imipramine compared to another active drug or to placebo. They found the effectiveness of imipramine clear though not overwhelming: “… Sixty-five percent of the 550 patients treated with imipramine improved while thirty-one percent of the 459 control patients improved. On this basis, imipramine seems superior to placebo.”

What established imipramine as the mainstay of antidepressant therapy for decades was not really this margin over placebo. As Tom Ban points out, “What made imipramine acceptable for clinical use was that in some patients, however few, it really worked. If one took the patient off prematurely the patient relapsed, and if the patient who responded to imipramine for the first time had a recurrence episode, he/ she responded again to the drug. The nagging question was and has remained how to identify the responding patient without exposing non-responding patients to the iatrogenic effects of antidepressants, and the expense involved in purchasing them.”

Which patients? It was exactly the same problem as with Marsilid: How does one define the group that will do well on them? It would certainly not be all patients with “depression.” The view was gelling that imipramine — and the host of other drugs introduced over the next decade as tricyclic antidepressants (TCAs) in reference to their chemical structure of three fused rings — had some specificity for melancholia. As Fritz Freyhan, another member of the pioneering generation of psychopharmacologists who, like Heinz Lehmann and Frank Berger, were German-Jewish emigres, told the same conference in 1959 at McGill University, where Delay and Deniker gave their above-mentioned findings: “Tofranil, by way of a crude analogy, may be compared with an antibiotic, the action of which only becomes apparent if it hits a suitable bacterial target. Thus, Tofranil can be thought of as a key which fits only certain keyholes. If it fits, it opens a door through which the depressive psychopathology departs and disappears.” In Freyhan’s own trial at Delaware State Hospital in Farnhurst, patients with classic “manic-depressive illness” did best on Tofranil. At this conference there was a general agreement that, as Frank Ayd put it, “Tofranil is not a tranquilizer [like chlorpromazine] but instead a specific antidepressant.”

One more qualification: “The TCAs worked beautifully in melancholies,” said Max Fink, in a conversation in 2006 looking back, “so long as they weren’t psychotic melancholies.”^ Melancholic patients who are psychotic do not respond well to tricyclic antidepressants alone and require urgently electroconvulsive therapy or, if that is not available, other chemical agents such as an antipsychotic or lithium. (Whether a combination of antipsychotic and tricyclic is the right treatment remains unsettled at this writing.)

But among drug treatments for nonpsychotic melancholia, the tricyclics were the treatment of choice. I have not made a great deal about their serious side effects, of which there were some, simply because later advocates of the SSRIs have overstressed the side effects of the tricyclics. Older readers who are clinicians may recall wrestling with cardiac arrhythmias induced by the tricyclics. But such serious adverse events were uncommon, and the most frequent side effects were “anti-cholinergic” in nature, such as dry mouth, blurred vision, and the like. As with any drug class, one has to weigh therapeutic benefits against risks, and the benefits were great.

Alfred Pletscher attempted to account for Tofranil’s “victory march” in the 1960s and after, when the MAOIs were vanishing from the radar. Kuhn’s original observations, Pletscher said, had been confirmed by others: The MAOIs had already shown “that mental depression was a condition accessible to chemotherapy.” Yet imipramine did not have the side effects of the MAOIs; and researchers in basic neuro-sciences were coming up with evidence that these tricyclic antidepressants affected brain neurotransmitters differently, which gave them a convincing story.

Pletscher was right: The door that Tofranil had opened in the brain saw much of the kind of psychopathology rush out that the amphetamines, history’s first mild antidepressants, couldn’t touch.

Looking back, Alexander Glassman at the New York State Psychiatric Institute said, “Actually, in the old days, the drugs were very effective. The old tricyclics were very effective drugs. They certainly had side effects that killed people in overdose, but they were very effective compounds. And if you got a good blood level, we really were getting 75-80 percent of the patients that were hospitalized in those days for depression better.” This is not bad.

The MAOIs were the first drug class based explicitly on a theory: that the biological brain “amines” such as serotonin and norepinephrine were neurotransmitters, and that correcting imbalances in neurotransmitters could make psychiatric illness better. In an advertisement in 1957, Hoffmann-La Roche flogged Marsilid as “an amine oxidase inhibitor which affects the metabolism of serotonin, epinephrine, norepinephrine and other amines.”‘ How it works: These neurotransmitters are discharged by the upstream neuron into the synapse, the space between upstream and downstream neurons, in order to make the downstream neuron fire. In the world of academic psychopharmacology, the theory was that psychiatric illnesses were caused by too little norepinephrine and serotonin, so here’s a great idea! We’ll prolong the presence of the amine neurotransmitters (the “monoamines”) in the synapse by inhibiting the action of the enzyme that destroys them, namely, monoamine oxidase. Hence, the presence of these vital monoamines in the synapse would be prolonged to do good; fewer would be taken back up into the upstream neuron (via the reuptake mechanism), and mental illness would be checked. Thus the concept of the monoamine oxidase inhibitors as therapeutic agents was born, although the first reference to monoamine oxidase inhibition appeared almost 20 years earlier, in 1938, when John Gaddum, then at University College London, said that the drug ephedrine increased the amount of epinephrine by inhibiting amine oxidase. Even today, it’s not clear whether this concept of reuptake inhibition offers the open sesame to the neurochemistry of psychiatric affliction. But it makes a good story, and billions of dollars worth of drugs — latterly the selective serotonin reuptake inhibitors (SSRIs) — have been sold in its name.

The real story of iproniazid and the MAOIs is as follows: Hoffmann-La Roche had procured large supplies of the chemical hydrazine used as rocket fuel in Germany during the war.’ On the basis of this supply, the company synthesized two antituberculosis drugs from isonicotinic acid hydrazide, and launched them in 1951: isoniazid (Rimifon) and iproniazid (Marsilid). An early trial of iproniazid took place at the Sea View Hospital, a sanatorium in the New York borough of Staten Island. The Sea View trials were a success; the drugs reversed the systematic toxicity of tuberculosis; the patients’ fevers subsided, and they began gaining weight. But there was one problem: In addition to the expected side effects such as muscle twitching and hyperreflexia, many of the patients became “mildly euphoric.”‘ TB is not a fun disease, yet later news accounts had patients dancing about the halls at Sea View.’

In the meantime, Swiss chemist E. Albert Zeller, at Northwestern University in Chicago, had discovered in 1952 that iproniazid, but not isoniazid, inhibited the action of the enzyme monoamine oxidase. Interest in the monoamines was just starting to quicken in psychiatry in these years, and any drug that had an effect on a brain chemical was going to be scrutinized clinically. But in 1953 Gordon Kamman, at the Fergus Falls State Hospital in Minnesota, put iproniazid into a group of chronic schizophrenic patients and found it virtually useless. Three years later, George Crane, at the Westchester Division of New York’s Montefiore Hospital, supplied a dyspeptic account of the “psychiatric side-effects of iproniazid”: “Concomitant with changes of outward behavior [hyperactivity, euphoria] were ambivalence and hostility, particularly evident in the female patients.” Normally, such negative reports would put an end to a proposed psychiatric indication. Yet interest in psychiatric uses of iproniazid continued to percolate.

The question of who in fact did discover the psychiatric uses of MAOIs is clouded in mystery. Nathan Kline’s name is usually associated with the discovery, but here the trail becomes foggy. Did pharmacologist John C. Saunders, then a research associate at Ciba’s offices in Summit, New Jersey, really claim at a conference in 1955 that his work on reserpine had led him to the monoamine theory of depression? He later said, in another conference in 1958, that it was he who earlier had proposed “the application of amine oxidase inhibitors to psychiatry on the theoretical basis of their action on amine metabolism in the brain, with the probability that they would alleviate depressions.” Kline himself, director of research at Rockland State Hospital in Orangeburg, New York, was at the 1958 conference, and presumably sat in the audience as Saunders, by now also working at Rockland, said he had made this earlier claim — a big claim, essentially to have discovered the role of neurochemistry in psychiatry. Unfortunately, the transcript of the 1955 conference made no note of Saunders’s participation.

Meanwhile, a second powder trail: In 1955, Alfred Pletscher, a scientist at Hoffmann-La Roche’s Basel headquarters, was a visiting scholar in Bernard Brodie’s lab in the National Heart Institute at Bethesda, Maryland, a lab that was rather surprisingly doing research on psychopharmacology. Kline visited Pletscher, who told Kline about the new drug iproniazid and its promise in psychiatry. Kline is said to have experimented first on himself. Then in 1956, together with Saunders and Harry P. Loomer, who was head of the clinical service at Rockland State, Kline administered the drug to 17 chronic female patients in the hospital who were “withdrawn, regressed, deteriorated, colorless, and of flattened affect.” Kline also gave iproniazid to patients in his private practice in Manhattan. At the end of 5 weeks, 47 percent of these chronically depressed patients had improved; by 5 months “a minimum of 70 percent of the patients have shown measurable response.” These results, reported in 1957, were pretty impressive, given that hitherto only ECT worked in such a population. The authors billed iproniazid as a “psychic energizer.” Loomer was senior author of the paper, then Saunders, then Kline. (George Simpson, at Rockland State at the time, recalls them as such sloppy researchers that, “It was remarkable that they ever discovered anything.”)

Just prior to the publication of their results, Frank Ayd reported early in 1957 that he had given Marsilid to 39 psychiatric patients at Franklin Square Hospital in Baltimore who had a mixed bag of diagnoses. He found that over half of them “did not benefit from this drug.” When Kline saw this negative one-pager in the American journal of Psychiatry, he “blew his cork,” as Ayd said later. Ayd told Kline, “Nate, the truth of the matter is I didn’t know you were working with iproniazid. This was an idea that came from the Chief of Medicine who works with TB patients and I just tried it.”

Kline henceforth announced himself to be the discoverer of the effectiveness of the MAOIs in psychiatry. For this achievement he received the coveted Albert Lasker Award for Clinical Medical Research in 1964 (he had received a first Lasker Prize in 1957 for discovering the effectiveness of reserpine in psychiatry). Loomer and Saunders, their noses very much out of joint, sued to share in the monetary award. Loomer dropped out of the action, but Saunders ultimately collected a third of the prize.

Curiously, Kline was said to have had “quite a job” convincing Hoffmann-La Roche to go for the psychiatric indication. According to London psychopharmacologist Merton Sandier, who was more or less an eyewitness, Elmer Severinghaus, medical director of the American branch of Roche, “was unimpressed by the case Kline put forward and vetoed further clinical study of the problem.” But Kline went over Severinghaus’s head, had lunch with G. David Barney, the president of the American office of Hoffmann-La Roche, and persuaded him to move forward on a pilot study.

Subsequently, the evidence of Marsilid’s effectiveness in psychiatry was substantial. In 1959, C. M. B. Pare and Merton Sandier at the Bethlem Royal and Maudsley Hospitals in London conducted the first large controlled trial of Marsilid, finding that, in 50 depressed patients “who had been considered suitable for electroconvulsive therapy,” 26 were improved, 12 of them due to the drug and in some cases quite dramatically so. “Perhaps of most interest,” they said, “is the question whether iproniazid is a general euphoriant or whether it acts only on patients whose depression is due to a specific metabolic abnormality which has yet to be identified. The striking response to iproniazid noted in some patients compared with the apparent ineffectiveness of the drug in others suggests that the latter may be true.” It is a striking comment on the failure of clinical psychopharmacology to progress that 50 years after they penned those lines, the biochemistry of depression remains a riddle.

By 1958, much experience had been collected. People spoke of Marsilid as eliciting a “Mona Lisa smile.” Carl Breitner, at Arizona State Hospital in Phoenix, concluded that “catatonic schizophrenia” and severe depression were really the same illness because both responded well to Marsilid and ECT — this at a time when the only consistently effective treatment for catatonia and melancholia had been ECT. At Mercywood Neuropsychiatric Hospital in Ann Arbor, over 70 percent of the patients with psychotic depression did well on Marsilid. (Today, there is no drug with remotely comparable effectiveness in this condition.)” Said Zale Yanof, a Toledo, Ohio, internist at a symposium in 1958, “… I soon found that the Marsilid effect was one that I had never seen clinically before. It is a potent psychic energizer and remarkable mood normalizer…”

In a trial beginning May 1957 of 142 psychiatry outpatients, Boston psychiatrist Samuel Joel discovered Marsilid to be the first drug effective in obsessive-compulsive disorder. Kline had already mentioned this anecdotally,” but Joel presented the numbers more systematically. “A group of six severe obsessive-compulsive cases have shown marked improvement with iproniazid treatment, where other therapies were ineffective.”"

Kline himself gave Marsilid in high doses together with amphetamine for acute depression. One of his patients, “a professor of medicine in a well known university, despite annoying side effects, cannot be persuaded to reduce the dose of iproniazid below 75 mg, since he would ‘rather suffer the side effects‘ and be free of the depression that has plagued him for twelve years than ‘take any chances.’”" Kline’s private patients on the Marsilid-amphetamine combo occasionally responded “in twenty-four to forty-eight hours.” (A horse named “Marsilid” won in the ninth at Belmont racetrack outside of New York on September 4, 1959. It is tempting to think that its owner might have been among Dr. Kline’s grateful patients.)

Marsilid seemed to have an elective effect not just in Nate Kline’s “acute depression” but in melancholia as well, reducing the need for ECT by up to 70 percent, as Theodore Robie with a private psychiatric practice in East Orange, New Jersey, found in 1958. Moreover, at St. Thomas’s Hospital in London, medical staff preferred iproniazid in patients “showing somewhat atypical depressive states, sometimes resembling anxiety hysteria with secondary depression, who seem to be specifically and almost completely relieved of their disabling symptoms by iproniazid after the failure of all other forms of treatment.” This 1959 report, by two of the students of William Sargant, head of psychological medicine at St. Thomas’s Hospital and one of the founders of biological psychiatry, was the beginning of the tradition of treating “atypical depression” with MAOIs.

So, Marsilid was a big therapeutic success, probably superior to any agent on the market today for serious depression or obsessive-compulsive disorder. It became almost a kind of cult drug, able to deliver therapeutic benefit in very sick patients who responded to nothing else. Unhappily, it was dogged by liver toxicity. In 1958 Hoffmann-La Roche sought to corral the problem by changing the recommended dose on the label. Practitioners could smell withdrawal coming; physicians at the Veterans Administration as well as the National Institutes of Health argued “that the product was so important in the treatment of certain mental health cases that it should remain on the market…” There existed, said the FDA in a private communication to a physician at the Philadelphia College of Pharmacy and Science, a core of “specialists familiar with its use… adamant in their opinion that Marsilid is a particularly useful drug and should be kept available for use where indicated.” The risk of liver toxicity was, the agency said, about one per three or four thousand patients, with about a 20 percent fatality rate. This jaundice side effect didn’t appear “when the drug was used at high doses in over 300,000 TB cases,” the Pink Sheet noted.

Nevertheless, in January 1961 the FDA ordered Marsilid withdrawn. It continued to be available for investigational purposes under a so-called IND, but was no longer in the pharmacies. An IND is FDA-speak for Notice of Claimed Investigational Exemption for a New Drug; they are filed by individual physicians seeking special permission to treat a patient with a drug not under an effective NDA, or New Drug Application.

Yet Marsilid’s reputation as something of a wonder drug lingered on. In 1965 a psychiatrist in Batavia, New York, sought an IND for permission to treat Irene X, who had been a patient in Rochester State Hospital in 1958 because of “recurring psychotic depressions of an endogenous nature.” Guy Walters, the director of Rochester State, told Mrs. X’s treating psychiatrist in Batavia that, “Mrs. X failed to benefit from the usual anti-depressants and shock treatment, and she underwent a protracted hospitalization. Finally, in March 1959, she was started on Marsilid, 50 mgms. t.i.d. [three times a day] with remarkable improvement, and within a month, was released.” Mrs. X stayed well on Marsilid for several years, but relapsed in 1962 after it ceased to be available; she was readmitted to Rochester State. “Since her readmission, Mrs. X has again received the gamut of accepted treatment for depression including Parnate [tranylcypromine], Tofranil [imipramine], Ritalin [methylphenidate], Elavil [amitriptyline], various tranquilizers and shock treatment, all without any lasting success.” Now it was time, again, for Marsilid.

By the late 1950s, a slew of other MAOIs had appeared, several with exceptional clinical promise. When the American Medical Association met in June 1959 in Atlantic City, Warner-Chilcott’s Nardil (phenelzine) had just cleared FDA approval, Pfizer’s Niamid (nialamide) was just around the corner, and Lakeside’s Catron (pheniprazine) was in the wings. “The entire group,” said the Pink Sheet, “are regarded as potent agents, and are being watched closely.”

The rest of this story takes us beyond the 1950s and the first drug set. Yet it was in the 1950s that this drug class established its reputation, despite often horrendous side effects, as something of a miracle cure for mood disorders. There was a certain population out there, deuced hard to define, that responded beautifully to Marsilid. Mrs. X’s story is clearly one example of this.

Moreover, intriguing little niche diagnoses bobbed to the surface, forgotten today but not necessarily inexistent, for which Marsilid proved effective. Leo Alexander, a noted Boston psychopharmacologist (and adviser to the U.S. government during the Nuremberg Trials), thought that “depression” was probably a cluster of different disorders. One of them was anhedonia, a “joyless inhibited state” that meant the inability to experience pleasure. He thought Marsilid a specific for that. Anhedonic patients were fatigued, listless, and filled with self-pity, unlike the self-reproach typical of melancholia. Alexander called this the “inert psychasthenic anhedonic reaction,” echoing the “psychasthenia” of turn-of-the-century Paris psychiatrist Pierre Janet. In one group of 54 depressed patients, Alexander had 16 with psychasthenic anhedonia. Of these 16 patients, 88 percent responded well to iproniazid (versus 26 percent of the nonanhedonic depressed).

Here’s another niche diagnosis, today completely off the radar, whose responsiveness to Marsilid commands a second look: In 1959 Martin Roth, professor of psychiatry in Newcastle upon Tyne, described a syndrome he called “the phobic anxiety-depersonalization syndrome,” involving patients who are “dependent and immature but also scrupulous, fastidious and inflexible.” The syndrome did not get taken into the official American diagnostic manual, DSM-III, in 1980. Yet it has wide international credence. In any event, the syndrome responds impressively to MAOIs. Thomas Ban said, “I’ve never seen a case of phobic anxiety-depersonalization for which MAOIs didn’t work. What Martin Roth discovered is a distinct disorder.” Thus Marsilid served as a kind of psychopharmacologic torch for carving out diseases.

The MAOIs “opened new vistas for psychiatry,” as Alfred Pletscher put it. “At a time when biological psychiatry was still in its infancy, the idea of treating mental disturbances, such as psychic depression, by interfering with biochemical processes in the brain, was quite sensational.” But by the summer of 1959, as the American Medical Association met in Atlantic City, there was one more development. “Very much in the limelight at last week’s meeting,” said the Pink Sheet, was “Geigy’s Tofranil, an anti-depressive though not a monoamine oxidase inhibitor, cleared by the govt. a few weeks ago.” The MAOIs would shortly shuffle offstage because of rare “hypertensive crises,” elevations of blood pressure that occur when monoamine oxidase is prevented from breaking down primary amines such as tyramine in the gut; these exaggerated fears represent the sort of herdlike behavior among psychiatrists that often unfairly sinks psychiatric drugs.

There now danced into the limelight a different kind of antidepressant entirely: the “tricyclics.”

The first antipsychotic in psychiatry was not chlorpromazine but a predecessor called promethazine (marketed in the United States in 1951 as Phenergan). The French drug house Rhone-Poulenc synthesized it in 1944 as part of a whole series of phenothiazine antihistamines they hoped to clean up with. Of course nobody called it an antipsychotic in those days because the term had not yet been coined. The French phrase for this new drug class was neuroplegics, literally drugs that struck the neurons of the brain. Promethazine, the first of the neuroplegics, was used in French mental hospitals in the early 1950s. Yet promethazine, though useful as a sedative in psychotic patients, was not terribly effective; it was said to make the patients so sleepy they could hardly move, and they were worse when they woke up. Promethazine was forgotten in psychiatry as soon as Rhone-Poulenc, once put on the psychiatry trail, synthesized a much more powerful member of the antihistamine series: chlorpromazine.

In a story too well known to merit much retelling, chlorpromazine was used for psychiatric indications for the first time in France in 1952. It was approved in the United States by the FDA in March 1954 and launched as Thorazine for “intractable pain” and vomiting. Only in July 1955 did Smith, Kline & French begin indicating it for psychiatry. The first ad: “Thorazine reduces need for electroshock therapy.”

Chlorpromazine was marketed in Europe as Largactil, or a drug having a large action, which was literally true: Chlorpromazine is effective for a wide variety of illness conditions, not merely for hallucinations, delusions, and psychotic agitation. It does abolish hallucinations and delusional thinking, and sedate the assaultiveness, the pacing, and the general agitation of many patients with schizophrenia. Its antipsychotic action is certainly no myth. “Anyone who would have bothered to spend just two nights on call prior to CPZ [chlorpromazine] on an admission service of a psychiatric hospital would have to be blind and deaf not to see what the introduction of CPZ did,” said one senior psychiatrist.

Yet chlorpromazine, and many of the other antipsychotics in the phenothiazine class, was also effective against pain, and was used, for example, as an obstetrical analgesic. It is a splendid antianxiety drug, and probably the treatment of choice in severe anxiety — if one is prepared to accept, in low doses, a less than one percent risk of extrapyramidal motor symptoms (e.g., tremors, rigidity) in the balance. It is also an effective antidepressant, and as we scan the first drug set for evidence of antidepressant action that was later forgotten or minimized in the craze for later patent-protected agents, we must glance at the antipsychotics as antidepressants.

A qualifier: Electroconvulsive (“shock”) therapy, originated in 1938, remains the most effective treatment of serious, melancholic depression. So reprising the usefulness of antipsychotics in depression might unwittingly convey the impression that they represent a superior treatment. They do not. Jean Delay, in whose Paris clinic chlorpromazine was first tried systematically, pointed out in 1955 the drug’s tremendous success in treating mania, sudden-onset psychosis, and schizophrenia. It was unsuccessful, however, in treating eleven depressed patients who then responded to ECT.

But not everybody wishes to undergo ECT, owing to the great stigma that is attached to it. Antipsychotic treatment of depression should be considered only in those patients in whom a good seizure through ECT cannot be obtained, as occurs in those with alcoholism, or when for cultural reasons ECT is rejected. For many with depression, chlorpromazine represents an effective treatment option. Jean Sigwald in Paris discovered this with his outpatients in 1953, a year after the drug’s first trials. In eight cases of “melancholia with anxiety,” Sigwald had very good or good results in five. A typical patient was

a woman of 46, experiencing an anxious melancholia for eight months, suicidal ideation with a plan, mental and physical slowing. From May 1952 on, for 37 days, we gave her 125 mg of chlorpromazine a day, a dose that we had to reduce to 100 mg following somnolence and tachycardia [accelerated heart rate]. In a few days the melancholic state disappeared, her suicidal ideation vanished completely, and the resumption of her earlier daily routines occurred progressively, becoming complete. We saw the patient nine months after the beginning of the treatment, and are maintaining her with a daily dose of 25 to 50 mg.

Sigwald said of the trial as a whole, “In general, chlorpromazine treatment almost always improves the patient’s mood, and one observes the disappearance of sadness and depression, in some cases giving way to a certain euphoria.” Later, he called chlorpromazine “the insulin of the nervous.”

Among the “neurotics” in the practice of English psychiatrist John Hutchinson, “The patients who benefit from [chlorpromazine] most are those who suffer from mixed states of anxiety and depression,” he said in 1956. “The states of agitated depression seen frequently in middle-aged patients respond dramatically to Largactil therapy. Deep intramuscular injections of 50 mg. up to twice daily bring the condition swiftly under control.” Admittedly, this is qualitative testimony, not a quantitative finding based on the random assignment of patients to a treatment and a control group. But clearly from Dr. Hutchinson’s perspective, chlorpromazine was close to a miracle drug in his older depressed outpatients.

These early investigators were often ecstatic about the results of chlorpromazine in schizophrenia, yet in the same breath they might mention their depressed patients. When Basel psychiatry professor John Staehelin convoked a symposium on chlorpromazine in November 1953, there was great enthusiasm about its effect in schizophrenia. Yet his staffer Paul Kielholz pointed out that it was really quite effective in endogenous, or melancholic, depression as well.

Having read a Rhone-Poulenc brochure while in the bathtub, Heinz Lehmann decided to introduce chlorpromazine in 1953 to several patients at the Verdun psychiatric hospital in Montreal where he served as director:

We included schizophrenics, depressed patients and we also had some organic dementias; we didn’t know who to give it to. We gave it for agitation… And two or three of the acute schizophrenics became symptom free. Now I had never seen that before. I thought it was a fluke — something that would never happen again but anyway there they were. At the end of four or five weeks, there were a lot of symptom-free patients. By this I mean that a lot of hallucinations, delusions and thought disorder had disappeared. In 1953 there just wasn’t anything that ever produced something like this — a remission from schizophrenia in weeks.

But then Lehmann added “something which is not often mentioned nowadays [1996], but quite a few other investigators had found the same: there were quite a few depressed patients who got better too, quicker than they would ordinarily have done.”

Lehmann thought little of ECT. But many ECT practitioners started adding chlorpromazine to their treatment of depression. In 1955 Douglas Goldman, director of the Longview State Hospital in Cincinnati, described a 35-year-old woman with a history of manic-depressive illness who had been on small doses of chlorpromazine. She then tried to commit suicide by driving her car into a tractor-trailer truck, an outcome avoided only by the skill of the truck driver. Afterward, “Patient told her husband she had planned her demise at home, but apparently the opportunity afforded by the truck seemed immediately attractive.” Goldman gave her a course of ECT, continuing the chlorpromazine at 300 mg a day, whereupon she recovered. Goldman said, “It was entirely clear to those taking care of the patient, that the combination of chlorpromazine with electric shock treatment accelerated significantly her recovery from the depressive state.”

However compelling these stories and data from open trials, they do not represent the gold standard of psychiatric evidence, the randomly controlled trial, or RCT. In 1962, Max Fink and Donald Klein at Hillside Hospital in the New York borough of Queens, where Fink was director of the Department of Experimental Psychiatry, provided definitive evidence of the efficacy of the phenothiazines as antidepressants. Between October 1958 and October 1959, in the largest placebo-controlled study ever done in one place up to that point, they randomly assigned two hundred patients referred for pharmacotherapy to a phenothiazine antipsychotic or to the new tricyclic antidepressant imipramine. They were stunned by the results from the phenothiazine group: “The greatest change occurred in the patients who had the greatest affective expression… The depressive state was markedly alleviated, although a mild apathy persisted.” This report was an early warning flare that the distinction then being established between “depression” and “schizophrenia” might be an artificial one, and certainly that the phenothiazines were not specific for “schizophrenia.”

Nonetheless, the distinction between “antidepressants” and “antipsychotics” in the late 1950s shattered the concept of “tranquilizers,” despite the fact that many new drugs, it turned out, could be launched as either (and thus were de facto tranquilizers). Among French researchers, the phenothiazines were seen as having strong antidepressant activity. As Jean Delay’s associate Pierre Deniker told an English audience in 1959, “Some of the new neuroleptics … seem to have a definite effect on depressive states. Levopromazine [levomepromazine] has been commonly used.” The supposed neuroleptic levomepromazine, synthesized by Rhone-Poulenc in 1958 and launched in France as Nozinan, went on to acquire a reputation as an antimelancholic.

Sometimes, the choice of whether to identify an agent as an antipsychotic or antidepressant was more a business decision than a scientific flip of the coin. When Pfizer developed thiothixene in the mid-1960s, they had little idea of the drug’s uses. Nathan Kline said, “We ran trials on thiothixene when it first came out and the manufacturers did not know what it was good for; [the trials] showed that it was quite effective as an antidepressant. They subsequently produced doxepin, marketing that as their antidepressant and switched thiothixene to use in the treatment of psychotics.” Indeed, when Pfizer floated thiothixene as Navane in the United States market in 1967, it was for “acute and chronic schizophrenia.” The story is actually even more colorful, for when Pfizer produced doxepin (Sinequan) in 1969, they are said to have envisioned it as an anxiolytic, then switched it to an antidepressant. For Pfizer, the concepts of anxiolytic, antidepressant, and antischizophrenic really turned out to be marketing slogans.

Drugs marketed as antipsychotics often beat recognized antidepressants and anxiolytics in depression trials. In 1966 a team led by John Overall at the University of Texas Medical Branch at Galveston — and including pioneer psychopharmacologists Leo Hollister in Palo Alto and Veronica Pennington in Jackson, Mississippi — concluded that anxious depression was really quite different from retarded depression. The logic: the antipsychotic drug thioridazine (Mellaril) — a latecomer phenothiazine (1959) to the first drug set — had soundly beaten the tricyclic antidepressant imipramine (Tofranil) in anxious depression; yet in retarded depression imipramine beat thioridazine.

There was lots of evidence that the supposed antipsychotics of the first drug set made serviceable antidepressants.’ When Saul Rosenthal and Charles Bowden at the University of Texas Medical School in San Antonio put thioridazine head to head with the antianxiety drug di-azepam (Valium) in 1973, thioridazine “was significantly superior to diazepam in a group of symptoms representing depressive symptomatology, including suicide, psychomotor retardation, helplessness, worthlessness, and guilt feelings.”‘ Valium, unsurprisingly, did well in this trial as an antianxiety drug, yet Valium also had a long history of effectiveness in nonmelancholic depression. It is interesting that in both these trials, thioridazine, an “antipsychotic,” was able to treat depression effectively; in theory depression was a quite different disease from schizophrenia.

Thus, the tranquilizer concept of the first drug set turned out to be quite robust; drugs belonging to different therapeutic categories — such as antidepressant, antianxiety, and antipsychotic — were actually more or less interchangeable. How could this be, in disease categories that in theory, by the late 1950s, were increasingly considered mutually exclusive and homogeneous?

Other terminology also vied for currency. The term antidepressant, as noted in Site, had existed since the commercial publicity for the amphetamines in the late 1940s, but it was not generally accepted in psychiatry at that point. The term acquired the keys to the city only with the American launch of imipramine (Tofranil), the first of the tricyclic antidepressant drugs, in 1959; initially called a “thymoleptic,” it was relabeled “antidepressant” in 196o.3 Four years later, antianxiety surfaced as a separate category with Hoffmann-La Roche’s advertisements for Librium (chlordiazepoxide), a member of the benzodiazepine class that the company insisted were not tranquilizers.

All these terms represented the splintering of the unity of “tranquilizer.” Indeed, the very existence of a concept such as “tranquilizer” menaced the effort to cut psychiatric illness into neat nosological categories and to assign a pharmaceutical treatment for each.

Methylphenidate (Ritalin) and meprobamate (Miltown) were quite different drug classes. Methylphenidate is a stimulant whose structure, a “phenyl-ethylamine backbone,” is close to that of amphetamine. Meprobamate is a dicarbamate, a compound based on carbamic acid, composed of a simple nitrogen-carbon molecule. Yet meprobamate, like methylphenidate, was an effective agent for nonmelancholic depression.

Wallace Labs billed Miltown explicitly as a tranquilizer, and said that it was specific for anxiety. Effective in “anxiety, tension and mental stress” claimed the first ads in 1956. It was Wallace’s benactyzine-meprobamate combo Deprol, launched in 1958, that they marketed for depression. (Benactyzine is synthetic atropine, an anticholinergic that American Cyanamid had patented in 1946 and, as mentioned, Merck brought out as Suavitil in 1957; it was thought, like all anticholinergics, to have some antidepressant properties, and atropine was commonly used in Europe as an antidepressant.) Deprol is rather difficult to defend as a superior product, given that the psychopharmacologists of the day widely scorned it and the American Medical Association’s Council on Drugs called it in 1971 “an irrational mixture,” declaring it “not recommended.” Yet many practitioners have fond memories of Deprol, and it must have had qualities that escaped the academic psychiatrists.

The blockbuster meprobamate itself served quite well as a mood drug, which makes it sound like an antidepressant but in fact it was probably best for “nerves,” in that large space between antipsychotics and aspirin. Frank Ayd, chief psychiatrist at the large private psychiatric hospital Taylor Manor in Ellicott City, Maryland, and active drug trialist, later said in an interview, “… There are people out there who are not psychotic but who are very miserable and who are quite willing to pay good money and go to a lot of inconvenience to get some relief. They knew they were never going to end up in institutions, although they often feared that, but they knew it was impacting on their married lives, social lives and their ability to work… God knows, there was enough overwhelming evidence that the barbiturates were not drugs that you could give out in a cavalier way for a minor condition.” This was the advantage, he said, of meprobamate. “So now when you had meprobamate with very small companies in New Jersey and then Wyeth of course had connections, international and what not, it became world wide very quickly.”

Meprobamate did well in trials. Lowell Selling, a psychiatrist in Orlando, gave it to 187 patients with problems of various descriptions who came into his office between January 1953 and April 1954. In the course of the study, he dispensed over 54,000 tablets of the drug (four tablets per day was the standard dose). Ninety-five percent of those with tension improved or recovered; anxiety, 90 percent; involutional (midlife) depression, 80 percent; and so on. He found no withdrawal problems in the patients.

Leo Hollister at the Veterans Administration Hospital in Palo Alto, California, among the best known investigators in the young field of psychopharmacology, gave meprobamate or a placebo to 37 inpatients; he also tried the drug in an open (uncontrolled) study of a further 191 chronically hospitalized patients. Results in the controlled trial: In the placebo arm, 2 of 15 improved (13 percent); in the meprobamate arm, 13 of 22 improved (59 percent). In the open trial: Of patients with affective disorders, 74 percent got better, as did 74 percent of those with anxiety; 40 percent of those with “mild” schizophrenia also improved. Hollister concluded, “The results from treating patients with anxiety reactions or affective disorders were quite gratifying. In these patients meprobamate appears to be the drug of choice.”

Meprobamate found all kinds of uses. Here is Arthur J. McComiskey, an ear-nose-and-throat specialist in New Orleans, testifying at an FDA hearing in 1966: “I see quite a number of people who feel they just can’t swallow, they say they have a lump in their throat, anything from a lump in their throat to a ball, or a cocklebur, some of them even say a pine cone.” So he checks them out, does maybe a bit of throat dilatation, then reassures them and gives them something to let them relax. “Most of all I think the medication [meprobamate] I give them to relax, to relieve anxiety does more for them than my actual treatment. I have seen it work again and again.” Other of Dr. McComiskey’s patients fear they can’t breathe. “They are usually terrified and feel they are going to smother that night. I not only have to assure them but give them something to let them relax and lay this thing aside. That is where I use meprobamate too… They absorb it all and get a good result with it, and they can go to bed…”

An imputation that haunted meprobamate from the beginning was addictiveness. This would result in a vastly unfair FDA hearing in 1966, discussed in the next chapter. Yet from the get-go, the scientific evidence of meprobamate’s addictiveness was meager. A typical study: In 1957, Joseph Borrus, a psychiatrist in New Brunswick, found meprobamate free of addiction problems. Nor did he find evidence of tolerance, that is, of patients needing increasing doses to get the same effect, or of withdrawal difficulties: “… In a few instances, approximately 1 percent, a strong dependence upon meprobamate will take place with a definite reluctance to give up the drug. This usually occurs in extremely dependent, emotionally immature persons who will grasp upon any means to maintain themselves free of their inner tension… The vast majority of patients, perhaps 75 to 85 percent in my own experience and that of others, will discontinue using meprobamate as they begin to feel better.” The remaining group “will discontinue gradually at the physician’s suggestion.”

In a study in 1964, Leonard Goldberg, an alcoholism researcher at the Karolinska Institute in Stockholm, found that, unlike heroin or morphine, meprobamate did not create dependence at therapeutic doses, and that the risk of dependence was less than 0.1 percent, far less than the dependency potential of alcohol. When dependency occurred, the time to its onset was measured in months rather than weeks as for other drugs (alcohol, however, was measured in years, heroin in days); the frequency of dependency per million users was on the order of 1-10, as opposed to 5,000-20,000 for alcohol. Frank Berger later said, with something of a snort, of the SSRI antidepressant paroxetine (Paxil), “Paxil is truly addictive. If you have somebody on Paxil, it’s not so easy to get him off…. This is not the case with Librium, Valium and Miltown.”

The uptake of meprobamate was stunning. In the months after the drug’s marketing in 1955, as Frank Ayd later remarked, “the demand for Miltown … far exceeded that for any drug previously marketed in the United States. For a time [television comedian] Milton Berle was renamed Miltown Berle, magicians pulled Miltown instead of rabbits from their magical hats … and drugstores displayed signs reading ‘out of Miltown’…” By 1965, around the time of the first break in the Miltown sales curve under hammering from the benzodiazepines, Carter-Wallace (as the firm became known) had sold about 14 billion tablets of meprobamate, supplying drug for around 500 million meprobamate prescriptions written in the United States for some 100 million patients.

More than anything else, it was the tremendous sales of meprobamate that crystallized a political reaction against “the tranquilizers.” At the generally hostile hearings of the Senate Appropriations subcommittee in June 1957 on the budget of the National Institutes of Health, psychiatrist Nathan Kline was the only witness to defend meprobamate, particularly against media statements “about side effects and withdrawal symptoms.” At hearings that Estes Kefauver, Democratic senator from Tennessee, convened in 1960, a horrified Kefauver remarked that in the United States alone 500 tons of meprobamate had been produced in 1958, “enough to give every adult male in the United States 40 hours a week of medication of this drug.”

The three drugs that symbolized the tranquilizer era had all been launched in or close to 1955: meprobamate, the antipsychotic reserpine, and the antipsychotic chlorpromazine. Although other drugs were called tranquilizers as well, these were the core products. They were seen as anchoring a spectrum of tranquilization for nervous conditions ranging from psychosis to nonmelancholic depression and anxiety, with chlorpromazine at one end in the “serious drug” category, and meprobamate at the other for the griefs of “everyday practice.” Later, science would distinguish between antipsychotics on the one hand and anxiolytics and antidepressants on the other, three separate categories, never shall they meet. But throughout most of the 1950s they were all lumped into one: the tranquilizers. In 1966 Ayd stated at the FDA hearings on meprobamate that, of these big three, reserpine and chlorpromazine were “major tranquilizers” — not everybody’s cup of tea. “You would not normally prescribe such drugs for the anxiety patients. All we had for them was the barbiturates. When meprobamate came along… this was a welcomed addition. This is one of the reasons why in a very short period of time the whole world … began to prescribe both the major and the minor tranquilizers.” These drugs meant “the ability of psychiatric patients to be treated in general hospitals, or in the office without even the necessity of going into a hospital.” This was significant progress.

In the early 1950s, a stream of new psychiatry drugs began to pour onto the market, although the psychoanalytically oriented psychiatrists of the day were really the last to prescribe them, and family doctors and internists the first. Today, the well seems to have run dry and new drugs in psychiatry are seldom. It seems astonishing that in the 1950s and early ’60s medicinal chemistry could have devised so many agents that conferred a benefit and had few conspicuous side effects. Granted, these early drugs didn’t have to run the gamut of animal safety tests and randomized placebo-controlled clinical trials that are required today, but they did undergo some investigation that, along with informed clinical opinion, served as valid sources of evidence that the effectiveness of these drugs was not in doubt, especially their impact upon depressed mood.

In 1951, Schering Labs brought out the first of the nonbarbiturate drugs for insomnia and anxiety. Dormison (methylparafynol), synthesized in Germany in 1913, had a modest uptake in the United States. A modified version, synthesized in 1955 by British Schering and launched that same year in France as N-Oblivon, enjoyed huge success. The drug was quickly elbowed aside by the benzodiazepines after 1960 and forgotten. But it’s a sign that what were now being called the tranquilizers were picking up.

In 1954 Ciba launched a drug against depression of much greater historical heft, even though its antidepressant activity later became forgotten in the furor about hyperactivity in children: Ritalin (methylphenidate). It was marketed as a stimulant rather than a tranquilizer, although it fit the part. The story begins in 1944 when Leandro Panizzon, one of Ciba’s medicinal chemists in Basel — born in 1907 and raised in Milan — was synthesizing nitrogen-containing compounds. Ciba’s inhouse pharmacology showed it to be a mild stimulant — indeed it had an amphetamine-like structure. Because it was almost a matter of honor in those days that chemists experimented on themselves, Panizzon and his wife Marguerite both took the new agent. It made no particular impact on him, but Marguerite felt excited and adventuresome under the drug’s influence. “I used to take it before a tennis match,” she later said. Ritalin was in fact named after her (“Rita”). Ciba marketed it in Switzerland as a “psychotonic” in 1954, and brought it out in 1956 in the United States from its headquarters in Summit, New Jersey, as a drug “to lift the depressed patient up to normal without fear of over-stimulation” (a disadvantage of the amphetamines). By 1957 Ciba was flogging it as a “mild smooth-acting antidepressant and stimulant.”

There was evidence for this claim. In animal studies, Ritalin reversed the side effects of reserpine in monkeys (reserpine caused a depression-like syndrome, and reversing it was a standard pharmacological test of antidepressant efficacy). In the first clinical trial, in 60 patients on a general medical ward in the Berlin-Charlottenburg Municipal Hospital in 1954, the drug caused considerable increases in mental aptitude, as measured in math experiments, plus produced a “good to euphoric” mood in three-quarters of the patients. The amount of Ritalin imported to the United States rose from 22 pounds in 1954 to 1,215 pounds in 1955, and it was in the United States that the main work was done establishing Ritalin as an antidepressant.

In 1955, at the Traverse City State Hospital in Michigan, there were about 500 patients on the antipsychotic drug reserpine, many of whom fell into reserpine-induced depressions or were sedated to the point of seeming asleep all the time. Of the 25 given a reserpine-methylphenidate combo, 22 improved. “Some have even been sent home…. In our opinion,” the study concluded, “phenidylate [methylphenidate] is well worth further clinical trial as an analeptic [stimulant], particularly in the chronic, regressed, negativistic psychotics.” Many of these backward patients may have had psychotic depression. The Traverse City statistics were eye-openers.

Evidence started to accumulate of Ritalin’s effectiveness in fatigue and dysphoria, symptoms associated with the kind of nonmelancholic patient in the community who today would probably be diagnosed with the catchall label of “major depression.” Of 39 “fatigued, tired, depressed” outpatients in the private practice of Milwaukee physician Adolph Natenshon, 27 had an “excellent” response to Ritalin and 7 a “good.” “Their worries seemed to disappear,” said Natenshon. “They were alert, fatigue disappeared, and they could go all day without tiring.” Although these 39 patients, “depressed due to pressures of modern day living,” were not melancholic, they certainly corresponded to the popular conception of depression in our own day.

In drug studies, special efficacy is spied in the “dose-response” relationship: the higher the dose, the better the response. Of 77 patients with diagnoses mainly of fatigue and neurotic depression seen at Hahnemann Medical College Hospital in Philadelphia in 1960, 44 percent of those on the 20-mg daily dose of Ritalin had a good response, 70 percent of those on the 60-mg dose.

In those days there were few controlled trials against placebo, so the kind of data one might expect today are simply not available for these historic periods. Moreover, the few controlled trials that were undertaken enrolled typically such small numbers of patients (being “underpowered”) that they were unable to spot anything less than penicillin-size differences. Nonetheless, there were some useful trials. In 1970 veteran psychopharmacologist Karl Rickels at the University of Pennsylvania studied Ritalin against placebo in 42 “mildly depressed” outpatients versus 34 on placebo. He and co-trialists found Ritalin significantly more effective than the sham drug. He concluded that Ritalin “may be of value in the treatment of mildly to moderately depressed patients who are treated by general practitioners, whose main target symptoms are fatigue, apathy, or anorexia…”

In 1957 the Council on Drugs of the American Medical Association found Ritalin “useful as a mild cortical stimulant in the treatment of various types of depression… Neurotic patients appear to respond better than those with frank psychoses.” This judgment is really the bottom line: Ritalin was quite effective in nonmelancholic depression, rather less so in hospital-type melancholia.

“Methylphenidate in 1956 was the first new drug used for the treatment of depressive states,” said pharmaceutical market-researcher Paul De Haen in 1973. It replaced the amphetamines, but when it became overshadowed by the tricyclic antidepressants, it was switched to childhood hyperactivity, its function as an antidepressant forgotten. Yet Ritalin is a reminder that there is gold in them thar hills, that much of the first drug set was quite useful in mood disorders.

Yet it was difficult for trialists to overlook that this antihypertensive drug also produced “a calming, tranquilizing effect, without the drowsiness so frequently associated with barbiturates,” as Riker Labs claimed in 1953 of its product “Rauwiloid,” an alkaloid fraction of the Rauwolfia serpentina plant. Boston cardiologist Robert Wilkins, who undertook the first clinical study of reserpine, reported comments of hypertensive patients, such as, “I’ve never felt as well,” “I haven’t felt this good for years,” “Nothing bothers me anymore,” and “I just don’t give a damn.” Wilkins added, “Of course, this is gratifying to the physician, but more important, it may give an indication of how the drug may act not only in the neurotic hypertensive but in other neurotic patients as well. I have told many psychiatrists… that ‘Rauwolfia is good psychotherapy in pill form.’”

In 1954 Riker Labs linked the concept of “tranquilizing” to frankly psychiatric indications in an ad for its new drug “Rauwidrine,” a combo of Rauwiloid and amphetamine: “Mood elevation needed? Here is a better approach.” The company praised “the “tranquilizing, mildly sedative action of Rauwiloid” combined with the stimulant effects of amphetamine. Nothing was said about hypertension. This marked the beginning of the era of advertised tranquilization in psychiatry.

Tranquilization was not just a marketing concept but was believed at the time to have an underlying scientific validity. When in 1957 the FDA sued State Pharmacal, a Chicago firm, to stop advertising their over-the-counter sedative “Tranquil” as a “tranquilizer,” it was on the logic that tranquilizers were something more than sedatives. The FDA believed there were four groups of tranquilizers: phenothiazine derivatives such as chlorpromazine (Thorazine); rauwolfia and its alkaloids such as Serpasil (reserpine); antihistamines with a diphenylmethane structure, such as benactyzine, which Merck brought out in 1957 as the “antiphobic” Suavitil; and meprobamate-style agents. It is thus interesting to see these highly diverse groups of drugs gathered together under the same umbrella on the grounds that they produced “tranquilization,” a concept that has now vanished from psychopharmacology. But much else of value has vanished as well.

In mephenesin, Berger had discovered modern history’s first “tranquilizer,” a drug that relaxed muscle, calmed the mind, and conferred the balm of sleep, without causing daytime sedation. He did not, however, realize at first the importance of his discovery. He then did some research adding mephenesin to anesthetics to observe its muscle-relaxant qualities. He also successfully used mephenesin to diminish the cramps and spasms in tetanus. At this point he didn’t think that mephenesin acted on the brain.

Then in 1947 he immigrated to the United States. Berger’s first post was at the University of Rochester Medical School in the department of pediatrics. It was the only job he could get at the time. Many of the patients had diseases involving movement disorders, such as cerebral palsy. Berger said to himself, “Let’s try mephenesin on some of these.” He had great success. Later he said, “I had never seen anything like mephenesin. I wanted to pursue this. The only thing wrong with it is, it’s not long acting. If you give it to a patient with cerebral palsy, it will diminish the shaking and tremors for half an hour completely. Then it will gradually come back over three or four hours.”

Berger didn’t have a solid British patent for mephenesin. Squibb became interested in the drug, and learned that it allayed anxiety as well as relaxed muscle spasms; the FDA approved it in September 1948 and Squibb brought it out in 1954 as Tolserol. Quoting from a clinical trial published in the Journal of the American Medical Association in 1949, the company claimed it as “… the only drug we have seen that allays anxiety without clouding consciousness.” It was also useful in the treatment of alcoholics: Wean them off alcohol and onto mephenesin, then withdraw them from the mephenesin. Squibb told Berger, “Look here, your British patent is no good. We won’t pay you any royalties, and if you don’t like it you can sue us.”*

Decades after these events and after everyone had forgotten about mephenesin, Berger — by this time a distinguished figure in psychopharmacology — said somewhat ruefully, “Mephenesin was the product.” He considered it superior to all the antineurotic drugs that came later. “It’s the drug that totally works.” As far as Berger was concerned, the apex of the mood drugs had been reached with the very first attempt.

There is no guide, chart, or computer software program for clinicians to clearly identify or quickly predict which drugs interact with the CYP enzymes and create clinically significant drug interactions in patients. More research and clinical drug trials need to be conducted and reported on these enzymes and their interactions. With the knowledge of how cytochrome P450 enzymes enzymes work and of their physiologic role in DIs, pharmacists can better predict significant interactions that are likely to occur or identify potentially problematic drugs.

Understanding which P450 isozyme is responsible for the metabolism of a drug will be essential when trying to predict and understand the magnitude of drug interactions. Some drug metabolism inhibitors are highly selective for certain CYP isozymes. Some drugs that are highly selective enzyme inhibitors may also be substrates for that same enzyme system and may cause an interaction by being a competitive inhibitor. Obviously, if it is known that a new drug is metabolized by a specific CYP isozyme system, it is logical to assume that the drug will exhibit DIs with known inducers and inhibitors of specific CYP isozyme(s).

Drugs having the highest capability for producing life-threatening or serious clinical consequences include those with a narrow therapeutic index. These drugs include warfarin, carbamazepine, lithium, procainamide, phenytoin, quinidine, theophylline, tricyclic antidepressants, and valproic acid, and/or highly protein bound agents such as warfarin, phenytoin, oral hypoglycemic agents, sulfonamides, and NSAIDs that may be substrates, inhibitors and/or inducers of either CYP1A2, CYP2C9, CYP2C19, CYP2D6, and/or CYP3A4 isozymes. Management of patients in a clinical setting can be simplified if drugs that are known to produce harmful DIs with each other are avoided or limited and the patient is closely monitored by a clinical pharmacist and/or physician.

If a drug interaction is suspected, the following approach may be helpful in determining the offending agents. First, determine that the drug interaction is not related to poor patient compliance or improper dosing of a drug. Second, determine if genetic polymorphism could be involved. Third, determine what new drugs have been recently added to or discontinued from the patient’s regimen and if the adverse effect seen is commonly associated with the new drug or any drug the patient routinely receives. Fourth, determine how each drug the patient receives is metabolized or eliminated from the body. Fifth, determine if any drug the patient receives is a substrate, inhibitor or inducer of any CYP isozymes. Lastly, after reflecting on the adverse effects the patient is experiencing, determine if these are commonly associated with any of the drugs. Consulting Tables 1–4 in this article can help identify possible interacting drugs. After the problem is identified, the drug(s) should be quickly replaced with an alternative drug not likely to produce the adverse effects.