ONE MORNING this summer I'll drag myself from bed and stumble downstairs to brew my morning coffee. Once it's brewed, I'll sweeten it -- but I can't be certain that the government will let me use the safest sweetener.
The culprit is the Delaney clause in an amendment to the Food, Drug and Cosmetic Act. Introducted by Representative James R. Delaney of New York, and passed with the amendment in 1958, this clause states: "No substance shall be added to food which has been shown, by tests appropriate for the testing of food additives, to cause cancer in man or animals in any amount." While its intention is the laudable one of protecting the American public from cancer, its effect has been to distort our entire food safety research program, to encourage inadequate testing of food additives, and to require the banning of substances which are demonstrably doing more good than harm.
Few Americans had heard of the Delaney clause before August 1970, when the Food and Drug Administration [on the basis of experiments now being reevaluated] banned cyclamate, an artificial sweetener. However, the availability of saccharin, another aritificial sweetener, significantly blunted the impact of this decision. Since many compounds which appear perfectly safe in normal use have been shown to produce cancer when tested under extreme conditions, no one should have been surprised when the Canadian Health Protection Branch found that saccharin produced bladder cancer in rats [if if formed 5 percent of their and their parents' diets]. As a result, in April 1977, the FDA decided to place saccharin on its list of forbidden additives.
The American publicc reacted with astonishment and even indignation. How, they asked, could a substance they had use for years without apparent ill efffect suddenly be dangerous? And how could any responsible person justify banning the only artificial sweetener available, when so many Americans -- who are notoriuous for their sweet tooth -- already weighed far too much?
Congress reacted to these concerns by postponing the ban until May of this year. The legislators clearly hoped the delay would permit additional studies including a persumably authoritatives report by the National Academy of Sciences, to clarify the situation and produce a consensus on what should be done. In fact, nothing has changed. As a staff member of the Senate subcommittee on health and scientific research told Science magazine, "The NAS has given us not one whit of additional help on the policy problem."
It is clear that what we need is not more studies or more research -- especially not more of the type we've been doing -- but a thorough going overhaul of our regulatory philosophy: one that discards the vain quest for absolute safety and seeks instead to offer us the maximum overall protection that can be achieved.
Those who accept the philosophy of the Delaney clause, but argue for the continued use of saccharin -- by pointing to its long history of apparently safe use or by questioning the application to humans of the results of experiments on animals -- are confusing the issue. That a substance produces cancer in animals does not necessarily mean it will do so in humans.Yet, with one possible exception, every substance that has been shown to produce cancer in humans also does so in some animals. The problem is that we cannot readily separate the cancers that saccharin might have caused from those that would have occurred anyway.
In an effort to solve this problem, the FDA and the National Cancer Institute have launched a survey that will involve 9,000 people over an 18-month period at a cost of about $1.375 million. Yet the survey will be able to demonstrate a cancer risk statistically only if saccharin causes at least twice as many cancers in humans as it does in rats, the most sensitive animal species known. So even this study can do no more than set a limit to the risk; it cannot possibly show that saccharin is completely safe.
Those who question the high doses used in the animal experiments are raising a more complex question, one which goes to the heart of the Delaney clause. They implicitly appeal to our experience with ordinary poisons, every one of which -- from aspirin to cyanide -- is harmless below some measurable "threshold dose." Whether the same is true or carcinogens remains intensely controversial within the scientific community. The "threshold dose" theory is certainly not true for radiation: a single particle hitting a critical site in the cell nucleus may cause cancer. Theory suggests the same may be true for chemicals. But for most chemicals, unlike radiation, the likelihood that a molecule will arrive at the critical point is not independent of dose. In some cases, therefore, our experience with chemicals suggests what has been called a "practical threshold": a dose below which the incidence of cancer is negligible. Yet this is not an absolute no-effect level; it does not guarantee total safety. And the Delaney clause regards a practical threshold as no threshold at all.
The very high dose levels used in the saccharin and cyclamate experiments have been questioned on other grounds. They clearly produced bladder stones, an effect not seen with lower doses. Dr. Morris Cranmer, head of the National Center of Toxicological Research, has produced for the FDA an 835-page "Final Report on Saccharin," which argues that what appear to be saccharin-related cancers were probably caused by these bladder stones rather than by saccharin itself. But Dr. Emannuel Farber of the University of Toronto, head of the NAS saccharin panel, though conceding the theoretical possibility that Dr. Cranmer's argument may be correct, is not convinced by it. And in the absence of conclusive evidence, we should probably accept the NAS panel's finding that "saccharin is a carcinogen of low protency."
We live in a world where some risks are inevitable. regulating food additives as though this were not true, the Delaney clause actually increases our risk in three ways.
First, it diverts scarce scientific resources from study of the mechanisms by which cancers are produced. Both Dr. Cranmer and Dr. Farber agree that only by understanding these mechanisms can we mount an effective program of cancer prevention. Yet, as Dr. Cranmer says, "The current saccharin dilemma has consumed many more FDA resources than have ever been expended in the toxicology laboratory seeking data."
More significantly, the Delaney clause's emphasis on whether an additive "does" or "does not" produce cancer -- whether it is or isn't absolutely safe -- encourages inadequate testing. The manufacturer's interests lie, not in obtaining data from which an accurate risk assessment may be made, but in preventing the emergence of data which suggests any risk at all. Thus, once he has successfully completely those tests required by the FDA, he will never again voluntarily submit his compound to testing.
What can happen when tests "raise questions" is illustrated in the case of cyclamate, which was banned on the basis of two experiments. One of the experiments was clearly inappropriate, and the other -- sponsored by the manufacturers -- tested a cyclamate-saccharin mixture, thus eliminating the possibility that tests results could be attributed to either compound. Since then, some 20 experiments have been performed, not one of which has produced significant evidence that cyclamate is a complete carcinogen. Nevertheless, last September, an FDA administrative law judge ruled that the safety of cyclamate had not been adequately established. He gave no indication what evidence he might consider adequate.
When this question was put to Dr. Charles Kokoski of the FDA's Division of Toxicology, he could only reply."The answer to that has not really come out. The only thing approximating it was [an experiment twice the sise of the largest ever conducted]. . . which was mind-boggling." He continued, "I guess if [the manufacturer] were to come back and say, 'All right, now we want to do the work. What is needed?' then we would have to sit down. . . and come up with some recommendations."
In other words, even though everyone agrees with the conclusion of a National Cancer Institute committee that "the present evidence does not establish the carcinogenicity of cyclamate," any suggestion of carcinogencity renders approval virtually impossible. This fully justifies Dr. Cranmer's comment, "The all-or-none approach of the Delaney clause provides absolutely no incentive to a petitioner to do comprehensive research; in fact, the better work you do, the greater the chance that your compound will be banned. Who are we kidding?"
A regulatory system which rewards, rather than punishes, thorough safety testing has been proposed by Dr. Nathan Mantel of George Washington University and Dr. Marvin Schneiderman of the National Cancer Institute. This system does not attempt to eliminate all cancers, but insists on proof that the number of cancers traceable to the use of proposed additive will be less than some established safety threshold -- let's say, one in a million. The manufacturer can prove this only by determining the risk accurately. Even if he finds the risk is not zero, we will still be safer than under our present system, where an apparently zero risk may conceal one far higher than the Mantel-Schneiderman system would permit.
Yet the main problem with the Delaney clause still remains: it does not allow us to deal with benefits as well as with risks. In fact, even if saccharin prevented more cancers than it caused that would be irrelevant to the carrying out of the Delaney clause.
We do not, of coure, know that saccharin prevents more cancer than it causes, but this seems possible for three reasons. First, excess weight, in addition to its well-known effects on blood pressure and heart disease, is a major risk factor in the development of cancer. Hence, if saccharin helps people attain and maintain their ideal weight, this in itself reduces cancer risk. Whether saccharin is helpful in weight reductions is a matter of controversy. As the NAS saccharin panel said, we have "no studies that permit objective assessment of the asserted health benefits of saccharin use."
Second, sugar itself has been shown to be a carcinogen. In fact, Dr. Cranmer calculates that sugar is 375 times as likely to produce cancer as the saccharin that would replace it. If these calculations are correct, then this point alone demonstrates that the Delaney cause ban on saccharin will cause more cancers than it prevents.
But the final point is that saccharin directly prevents cancer. That a substance which causes cancer also prevents it seems paradoxical, yet -- so far as animal experiments go -- the evidence is unequivocal. Saccharin is a weak bladder carcinogen and also [in this organ] a fairly powerful promoter, converting potentially cancerous cells into actively growing chancers. In other organs, however, it inhibits cancers. Which of these actions is quantitatively more important we have at present no way of knowing.
Congress's suspension of the saccharin ban has expired. What will it do now? The easy, perhaps the politically expedient, thing would be to continue to treat saccharin as an exception, exempt from the rules governing other food additives. But this would not truly serve the public interest. What Congress should do, what we must hope it will do, is abandon the inherently vain quest for absolute safety, and substitute a regulatory system under which the risks of all food additives are balanced against the benefits they offer.