Banned Read online

  Diethylene glycol

  72 percent

  Sulfanilamide

  10 percent weight/volume

  Water

  15.6 percent

  Furthermore, the chemists conducted spectrographic examination that failed to reveal the presence of known poisonous substances, such as lead, bismuth, mercury, or arsenic. In essence, the chemical analysis of Elixir Sulfanilamide confirmed the statements of the Massengill Company regarding the composition of the drug.58

  E. M. K. Geiling, chair and professor in the Department of Pharmacology at the University of Chicago, also conducted toxicity studies of Elixir Sulfanilamide. Geiling determined the toxic agent by carrying out toxicity experiments on rats, rabbits, and dogs using the following substances: pure diethylene glycol, pure sulfanilamide, Elixir Sulfanilamide-Massengill, and “synthetic” elixir of sulfanilamide (produced by the AMA Chemical Laboratory with pure substances in approximately the same proportions as found in the Massengill elixir). Through a series of experiments, Geiling hoped to determine three things:

  1. The toxic and lethal doses of each of the substances when given in relatively small doses three times daily. This information seems particularly necessary since we were not able to find any data in the literature on this specific point.

  2. Our experiments were further planned with the hope of being able to reproduce in healthy experimental animals, in about the same time, the clinical and pathologic picture as presented by patients who had taken fatal doses of the Elixir of Sulfanilamide-Massengill.

  3. Through our experiments we hoped to discern the toxic ingredient in the Massengill elixir.59

  Such a strategy laid the foundation for future toxicological investigations. Geiling’s results left no doubt as to the cause of the deaths. In rats fed via a stomach tube three times daily, the mortality rate rose to 100 percent after eight or nine doses of two cc’s of diethylene glycol, Elixir Sulfanilamide, or synthetic elixir. By contrast, none of the rats treated with sulfanilamide or water died during the experiment. Thus Geiling and his collaborators concluded that diethylene glycol was the toxic agent in Elixir of Sulfanilamide-Massengill since animals treated with the chemical exhibited the same symptoms as animals treated with Elixir Sulfanilamide and a synthetic elixir formulated from the same ingredients in the same proportions.60

  As further confirmation, they noted that sulfanilamide alone did not prove fatal to rats, rabbits, or dogs, but the drug did cause convulsions in some of the animals, and the researchers raised the possibility that the drug contributed to tissue damage in animals or human beings with impaired renal function. The final element of the conclusion contained significant implications for future policy, specifically premarket testing on animals: “Our experiments emphasize the importance of administering drugs in divided doses to experimental animals when it becomes necessary to know whether or not a drug has cumulative effects. Errors resulting from an oversight of this important pharmacologic principle may be costly to human lives.”61

  Given that drug therapy often required repeated daily doses, Geiling believed that testing new therapies should reflect such procedures if cumulative effects were to be understood. Such a view began to broaden pharmacologists’ perspective to include chronic effects. Geiling further clarified his point: “We can confirm the finding of Haag and Ambrose that the ingestion of 15 cc. of diethylene glycol per kilogram in a single dose by stomach tube proves fatal to rats. This figure, however, is no index of the toxic and possible fatal effects of the drug, if administered in small divided doses, especially since neither the fate nor the mechanism of detoxification is known.”62 The earlier research had isolated the lethal dose of diethylene glycol, which is to say, its acute toxicity, but Geiling distinguished this piece of data from the equally important effects of repeated doses, or chronic toxicity. In this sense, the toxicity of diethylene glycol had broader implications for the science and policy of newly introduced drugs.

  Finally, Paul R. Cannon, M.D., also from the University of Chicago, conducted a pathological evaluation of rats, dogs, and rabbits that had died following toxic doses of diethylene glycol, Elixir of Sulfanilamide-Massengill, synthetic elixir, and sulfanilamide alone. Cannon found a remarkable similarity between the pathological effects of a toxic dose of diethylene glycol, the synthetic elixir, or Elixir of Sulfanilamide-Massengill in animals and a lethal dose of Elixir Sulfanilamide in humans.63 In addition to the research of Geiling and Cannon, Edwin P. Laug and his colleagues at the FDA published the first toxicological benchmark: the LD50 (see below).

  An appreciation of the impact of the Elixir Sulfanilamide case on the evolution of toxicology in the United States requires a review of the history of the Division of Pharmacology at the FDA. The FDA was established in 1927, when the activities of the Bureau of Chemistry within the USDA became distinct from agricultural chemistry and the work of the department as a whole. Regulators and regulatory chemists were moved to a separate agency called the Food, Drug, and Insecticide Administration (later, the FDA).64 However, jurisdiction for the Federal Insecticide Act remained under the USDA. This move, noted Christopher Bosso, separated regulatory activities on behalf of consumers from those on behalf of farmers.65 The first commissioner of the FDA, Walter Campbell was determined to examine the toxicology of lead and arsenic in response to the widespread use of these chemicals as agricultural insecticides. To oversee this project, he appointed Erwin Nelson, a pharmacologist on leave from the University of Michigan, as the acting chief of the new division.66

  In 1935 Nelson devoted himself to raising pharmacology, hitherto part of the Division of Medicine at the FDA, to division status. To accomplish his goal, Nelson canvassed various universities for experts in critical aspects of toxicology. Several individuals from the original branch of pharmacology made up the core of the new division: Harold Morris, Herman Morris, Howard Lightbody (a biochemist specializing in the study of enzymes), and W. T. McCloskey. Nelson recruited several additional scientists for the new division: Edwin P. Laug, from the University of Pennsylvania; Lloyd C. Miller, a lipid biochemist trained at the University of Rochester; and Herbert Braun, from the University of Wisconsin. Rather than selecting trained pharmacologists, Nelson sought scientists whose specific expertise could contribute to the toxicological analysis of lead and arsenic. Thus he assembled experts in analytical work, enzymology, animal studies, and pathology.67

  Herbert O. Calvery joined the Division of Pharmacology as Nelson’s replacement as acting chief. Other scientists joined the new division. Geoffrey Woodard entered the division as a laboratory apprentice, but he played an important role in acute toxicity studies. Harold Morris moved to the National Cancer Institute, where he became widely respected for his research on cancer. O. Garth Fitzhugh, a specialist in the study of chronic toxicity, replaced him in the area of chronic toxicology in 1938. These individuals transformed toxicology from the study of the effect of a single dose on a single animal to the sophisticated statistical analysis of dose response curves necessary to understand the toxic effects of drugs and other chemicals on various animals and humans. Such precision arose as a direct response to the Elixir Sulfanilamide tragedy.68

  The FDA mobilized its field scientists against Elixir Sulfanilamide and now through the new Division of Pharmacology it could also respond through laboratory analysis. In reviewing the existing literature, including Geiling’s research, Edwin Laug and the other FDA toxicologists realized that no one had developed a method for comparing the toxicity of one substance to another. Although Laug and others suspected that diethylene glycol was the cause of the many deaths associated with Elixir Sulfanilamide, they had to develop an approach that would confirm their suspicions statistically. To do this, the FDA toxicologists were assisted by the ground-breaking research of Chester I. Bliss, who was brought to the FDA as a part-time consultant by Herbert O. Calvery. Bliss had studied with R. A. Fisher, the great English bio-statistician who developed and promoted the use of statistics in biology. />
  Bliss had published a seminal paper, “The Calculation of the Dosage-Mortality Curve,” in which he demonstrated the sigmoid character of the typical dosage-mortality curve as established by numerous toxicological studies of a large number of organisms by many biologists. Acknowledging his debt to R. A. Fisher and the trends in biostatistics Fisher inspired, Bliss selected his procedures on the basis of their statistical accuracy and efficiency. Bliss intended to present his techniques for calculating the transformed dosage-mortality curve in sufficient detail so that biologists with limited knowledge of statistics could use them. The value of the dosage-mortality curve to biologists in general, and toxicologists in particular, arose from its ability to describe the variation in susceptibility between individuals of a population. In line with the theory that any given population would show a range of susceptibilities to any given toxin, Bliss demonstrated that mortality could be plotted against dose in a way that would indicate what percentage of any given population would be killed by a particular dose. This experimental technique could determine the precise minimum lethal dose for each organism, since a dosage above minimum effectively killed susceptible individuals. Bliss recommended exposing a series of sample groups of organisms to graduated doses and recording the percentage killed for each sample group. When Bliss plotted this data, it resulted in a sigmoid curve (exactly as predicted). Thus he concluded: “The sigmoid dosage-mortality curve, secured so commonly in toxicity tests upon multicellular organisms, is interpreted as a cumulative normal frequency distribution of the variation among the individuals of a population in their susceptibility to a toxic agent, which susceptibility is inversely proportional to the logarithm of the dose applied.”69 Bliss’s explanation and technique were enhanced by a note from Fisher. It examined the case in which there were few or no survivors and appended a method to address this problem.

  Bliss’s paper appeared in 1935, but the FDA pharmacologists were already familiar with his work, since he worked in their laboratory. The dosage-mortality curve provided an excellent method to evaluate the toxicity of a given chemical. Laug and his colleagues in the Division of Pharmacology were among the first to apply Bliss’s method to an actual case when they evaluated the toxicology of glycols and derivatives in response to the Elixir Sulfanilamide disaster. Prior to this study, toxicologists had estimated the doses that were lethal to 10 percent, 50 percent, and 90 percent of a given population, but the results were unsatisfactory, particularly when toxicologists used such data to interpolate a sigmoid curve through a number of points. Laug and the FDA pharmacologists used Bliss’s method to calculate and plot the dose-mortality curve for rats, mice, and guinea pigs. They determined that the most useful parameter was the lethal dose for 50 percent of the population (LD50), as this value required the fewest animals for calculation (a minimum of 10). In contrast, arriving at the same level of precision for 99 percent deaths required at least 103 animals. Moreover, Laug and his team calculated standard errors for the LD50 and found they could determine the dosage at which 50 percent of the population would be killed in nineteen out of twenty experiments.

  Citing the studies of Geiling and others at the University of Chicago, Laug reflected on the importance and probable significance of his team’s findings for humans. Although the FDA researchers had confirmed the findings of Geiling and others, Laug advised restraint in applying the results to humans: “It seems proper at this point to reemphasize the inadvisability of attempts to interpret experimental data on laboratory animals directly in terms of man. It is entirely too dangerous. The present investigation confirms again the wide variations that may occur between species.”70

  Although they presented the derivation of LD50 as a valuable new tool in toxicology, the FDA researchers acknowledged that neither acute nor chronic toxicity gave a complete picture and scientists should assess both dimensions of toxicology. They argued for establishing the most complete toxicological profile possible. Ideally, such a profile would include both acute and chronic toxicity studies as well as pathological evaluations. In addition, certain glycols produced acute effects at high doses without producing chronic effects at lower doses. Laug contrasted this toxicological behavior with that of the well-studied heavy metals and narcotics: “Again, although small doses of lead, mercury, selenium, fluorine, narcotics, etc., are acutely toxic, it is the insidious character of their chronic toxic effects that makes them even more dangerous.”71 As of the 1930s, the study of chronic toxicity had not progressed as far as the study of acute toxicity, except in the case of heavy metals and certain drugs.72 Nevertheless, the model of chronic toxicity provided by heavy metals was hardly universal, as Laug noted in the statement above.

  Methods for deriving dose-response curves and LD50s were certainly among the most important legacies of the Elixir Sulfanilamide tragedy. Not only did the approach transform regulatory science as it was conducted at the FDA, industry adopted the same procedures for its analyses of new chemicals (at the FDA’s direction). Long after he retired, Laug would recall the significance of the division’s initial exploration of LD50 as one of the first successful applications of statistical approach to toxicology: “I think it was the most significant thing that we did … in those days there was not much precision when determining toxicity. And what we did was by the use of statistics, we made it possible that when you treated animals with something toxic, you could create a curve, a slope, and the significance of that was that you could then compare it to something else and that was the point, LD50.”73 A small group of FDA toxicologists in the Division of Pharmacology transformed the field of toxicology. More than any other single procedure, LD50 became the benchmark in most initial studies of toxicity for pharmaceuticals and environmental chemicals.

  Still, the question remains: Why did the Division of Pharmacology, which was established to evaluate the hazards of lead and arsenic in insecticides, develop the LD50 in response to the Elixir Sulfanilamide disaster? The Division of Pharmacology lost its funds to investigate lead and arsenic in 1937 or 1938. According to Geoffrey Woodard, another FDA pharmacologist, “Well, the insecticide problem was the original basis for having set up a division. Now as Ed [Laug] said, it was a political football and I remember we worked through ’37 or ’38 until New Year’s Eve—right up until midnight on New Year’s. Because Congress cut off our funds and said there was to be no more work on lead and arsenic.”74 In Woodard’s view, which was shared by the other FDA pharmacologists, lead and arsenic insecticides had become politically charged.75 In response to increased scrutiny, apple growers had appealed to their congressmen, who voted to transfer the authority for the examination of insecticides to the Public Health Service (PHS) at the National Institutes of Health (NIH) by revising the appropriation act of the FDA so that none of the funds could be used for the study of toxicity of lead and arsenic. The restriction of funds did not stop Calvery and Laug from publishing on the risks of lead compounds, however.76 Deprived of its work on the toxicology of lead and arsenic insecticides, the Division of Pharmacology concentrated its considerable effort on the study of Elixir Sulfanilamide, the glycols, and eventually other chemicals. Christopher Bosso revealed that shifting research on insecticides from the FDA to the PHS meant a move away from long-term laboratory studies based on experiments with laboratory animals to extrapolate chronic effects. In contrast, the PHS emphasized field surveys that questioned farmers about their health, which could reveal acutely toxic effects, but for the FDA, the PHS approach was inadequate for determining longer-term chronic effects.77

  While the FDA strove to eliminate the risks posed by Elixir Sulfanilamide, and the AMA with the assistance of Geiling and other University of Chicago faculty conducted the chemical and toxicological analysis of the elixir, Massengill attempted to defend its product. The company had earlier been convicted and paid fines for violations of the Food and Drugs Act, in September 1934 and March 1937. H. C. Watkins, Massengill’s chief chemist, had been cited by the solicitor of the Post Office for
distributing a medicine alleged to reduce weight, to bring about “perfect slenderness,” and to cause the body to acquire “a trim, youthful, athletic look.” To avoid charges of fraud, Watkins filed a stipulation agreeing that the sale of the product would be abandoned and not resumed at any future time.78 Unfortunately, under the PFDA, the only basis for action against the interstate distribution of the “elixir” was the allegation that the word implied an alcoholic solution, whereas the product was a glycol solution. It was for this reason that S. E. Massengill, owner of the eponymous company, could claim in a letter to the AMA, “I have violated no law.” Although Massengill’s statement conformed to the letter of the law, regulatory bodies generally believed that most drug manufacturers recognized a greater responsibility to the public. Such ethical obligations did not concern Massengill. He refused to take responsibility for the deaths, instead blaming them on the “bad effects” of sulfanilamide.79 Massengill’s statement betrayed his complete lack of knowledge regarding the toxicity of Elixir Sulfanilamide. Even in the absence of controlled animal experiments, like those conducted at the University of Chicago and the FDA, a simple literature review would have revealed a careful analysis of ethylene glycol (the active ingredient in antifreeze) and some of its derivatives, including diethylene glycol.

  Specifically, Massengill and Watkins would have discovered a paper written by Oettingen, one of the fathers of pharmacology in America. Oettingen, who at the time taught pharmacology at Western Reserve University in Cleveland (and later served as the director of the Haskell Laboratory), tested the toxicity of the glycols on various animals, including rats and frogs.80 Regarding the therapeutic potential of diethylene glycol and ethylene glycol, Oettingen wrote: “Ethylene glycol and diethylene glycol may be of interest for therapeutic use, as solvent and as vehicle. With these substances the local irritation is comparatively small. Their application to the skin seems to be without risk. Given orally in larger doses, they may produce severe gastro enteritis and systemic symptoms.”81 Still, even if they had completed a literature review, Massengill and Watkins might have argued that the benefit of the drug sulfanilamide outweighed the unknown risk of gastro enteritis. It is also possible that they had experience with one of the nontoxic glycols, such as polyethylene glycol, which is nontoxic and an effective laxative (widely marketed today as an over-the-counter therapy).