What is the link between chemicals and cancer?
Forty years ago, Bruce Ames was a young microbiologist working at NIH in the day and enjoying Scottish country dancing in the evening, when he had an inspiration: to use the rapid growth of bacteria as a method for determining whether a particular chemical was able to cause mutations. If the chemical was positive — i.e., was mutagenic — it might be considered as a possible cause of cancer. This method, soon called “the Ames test”, became widely used. It was cheap, fast, and sensitive. One of the first discoveries was that a dye commonly used in children’s pyjamas had mutagenic properties. Bruce Ames became a hero to the environmental movement when he led a successful campaign to ban such dyes.
Ames was more interested in reducing the death toll from cancer than he was in attacking new chemical technology. As more results from the Ames test accumulated, he realised that many naturally occurring chemicals were also giving positive results. Even more disturbing, the number of chemicals that seemed to be positive in high-dose tests on mice and rats was, he felt, excessive. In an extensive series of important reviews, published in prestigious journals such as Science and Proceedings of the National Academy of Sciences, he has attempted a quantitative estimate of the difference in human cancer. Because his figures show manmade chemicals in food and the environment to be quite insignificant compared to natural or self-inflicted factors, the name of Bruce Ames is now anathema to the same environmental movement that once applauded him. Nevertheless many professional scientists believe that Ames’ position is basically correct. If the inventor of the Ames test now says that most methods for detecting carcinogenicity are invalid, it is certainly not a case of sour grapes. This article is an attempt to summarise his beliefs. Those who are sufficiently interested should read some of the papers listed in the bibliography.
(1) What do we know about the incidence of cancer?
First, cancer risk increases according to the 5th power of age. That is, a 40-year-old is 100,000 times more likely to be cancerous than a 20-year-old. There are more cancer cases per 100,000 population simply because we are living longer and no longer dying of infectious diseases.
Second, the age-corrected mortality (death rate) from cancer has been declining since 1950 except in those over 84. Overall decline has been 13%. Naturally much of this decline is caused by improved detection and treatment. The only exceptions are lung and skin cancer, clearly caused by tobacco smoking and by increased exposure to sunlight. There are occasional claims that certain types of cancer are increasing slightly, but improved methods for detection are probably responsible.
Thirdly, some mostly unknown environmental factors have a major influence on the types of cancers that are likely. Japanese, for instance, have a high incidence of stomach cancer, yet Americans and Japanese-Americans have a low incidence. On the other hand, American men have much more likelihood of prostate cancer than do Japanese.
(2) What are the major known causal factors in cancer?
The single most important factor is smoking. This accounts for one-third of all US cancer deaths, not to mention one-fourth of heart disease. Each year, smoking causes 400,000 premature deaths in the US and 3 million deaths around the world.
Chronic infections contribute to about one-third of cancer on a world- wide basis. As mentioned below, any factor that causes body cells to divide increases the likelihood of cancer. Hepatitis B and C infect 500 million people, mainly in Asia and Africa. This liver infection is a major cause of “hepatocellular carcinoma”. Two different Schistosomiasis worms infect Chinese colons and Egyptian bladders, being associated with increased cancer risk in those two organs. Liver flukes cause chronic inflammation of the biliary tract, hence risk of cholangiocarcinoma. A bacterium, Helicobacter pylori, is adapted to living in the human stomach and is now believed to be a major cause of stomach cancer, ulcers and gastritis. (So much for the classical psychogenic explanation for ulcers!)
Overall about 70% of cancers might be caused by environmental factors, but pinpointing the exact causes is very difficult. There remains some 30% that cannot be ascribed to any factor other than age and bad luck.
(3) How does cancer develop?
The first requirement is that a dividing cell suffer some sort of damage to its DNA. (DNA is the basic material of our genes.) DNA damage occurs all the time, but our bodies have excellent repair mechanisms to detect and destroy damaged DNA. Based on the amount of DNA breakdown products in the urine, Ames and co-workers estimate about 10,000 “hits” on DNA every single day in an adult. These repair mechanisms are not 100% perfect, and some damaged DNA does escape.
DNA damage is mostly caused by oxidants. The oxidants in turn arise from both internal and external sources. Internal oxidants come from mitochondria, peroxisomes, cytochrome P450 enzymes, and phagocytic destruction of infected cells. The production of oxidants when infected cells are destroyed may be a factor in the connection between chronic infection and cancer. External sources of oxidants include the nitrogen oxides of tobacco smoke, iron and copper salts, and natural plant phenolics like chlorogenic and caffeic acid.
If oxidants are bad, then antioxidants should be good. They are: antioxidants protect against disease. Natural antioxidants include ascorbic acid (vitamin C) and tocopherol (vitamin E). Synthetic antioxidants are also good. One worker estimated about 5% reduction in cancer because of approved antioxidants added to our food.
The health benefits of antioxidants, provided mostly by fruits and vegetables, are statistically highly significant. The quarter of the US population with the lowest intake of fruits and vegetables has double the cancer rate of the quarter with the highest intake. This applied to “epithelial” cancers (lung, mouth, larynx, oesophagus, stomach, pancreas, cervix, bladder, and colorectal) plus ovarian cancer. Breast and prostate cancer, on the other hand, is less affected by fruit and vegetable diets. (Although there is at least a statistical link between fat/calorie intake and breast cancer.)
Persons taking daily tocopherol or ascorbate had one-third the risk of developing cataracts. In contrast, smoking and radiation (both well known oxidative stresses) are strong risk factors for cataracts. Smoking seems to destroy ascorbate: smokers need to take double or triple amounts of ascorbic acid to achieve the same blood levels as non-smokers. Incidentally, smoking by the father seems to affect sperm production and health; smoking fathers increase the risk of birth defects and childhood cancer in their offspring.
Excess food, at least in rats, is “the most striking rodent carcinogen ever discovered”. Even a 20% increase in calories over the optimal results in shorter life, with more endocrine and mammary tumours.
Excessive cell proliferation (cell division) is a very important factor in cancer production. This has been mentioned above in relation to chronic infection. Major dietary factors, such as salty pickles in the Japanese diet, have been hypothesised to be involved in the high rates of stomach cancer in this population. Even table salt, at high enough concentrations, can cause stomach cancer.
That cell proliferation predisposes to cancer is a major source of false positives in chemical screening as normally carried out. Test chemicals are repeatedly applied to animals at the “MTD” (maximum tolerated dosage). This is like chronic wounding, “which is known to be both a promoter of carcinogenesis in animals and a risk factor for cancer in humans”. Many chemicals that purportedly have caused cancer at high dose (MTD) levels, may therefore not be true carcinogens. The infamous saccharine tests are a case in point: only female mice dosed with nearly toxic levels of saccharine showed an increase in bladder tumours.
For these chemicals that “cause cancer” at high doses only by tissue irritation, a tenfold reduction of dose in a rat or mouse experiment would show much more than a tenfold reduction in risk. This seems to have been confirmed. One analysis of 52 tests showed that two-thirds of the purportedly positive results for carcinogenicity would not have been found if the dosage had been cut even by one-half! (I suspect that commercial cancer-screening laboratories get new contracts in direct relationship to how many “successes” they have had previously.)
(4) How do synthetic and natural chemicals line up as causes of cancer?
The conventional cancer-screening techniques are, as stated above, too sensitive. There are not merely a few chemicals that show up as carcinogenic. Instead, nearly one-half of all chemicals tested seem to be positive in these tests. The ratio is the same for both natural and manmade chemicals, even though very few natural chemicals have been tested. Thus we cannot generalise that natural chemicals are inherently safer or riskier than synthetic chemicals. We must look instead at the quantities of chemicals ingested.
Plants contain surprisingly large quantities of natural pesticides. One of Ames’ greatest achievements, in my opinion, has been to compile convincing evidence about how many natural chemicals have pesticidal functions. (In my youth, the question of the function of different “secondary” plant products was much debated. Some thought that products like alkaloids and lectins were mere accidents of metabolism, a plant process gone wild. I personally thought that the main role of these chemicals was to provide research material for young biochemists.) Ames pointed out that up to 5% of the fresh weight of vegetables can be natural pesticides.
The list is very long, and a sample limited just to non-toxic plants would include: the sharp flavours of mustard and other cabbage-family crops; piperine (10% of weight of black pepper); light-sensitising psoralens in parsnip and celery; chlorogenic and caffeic acid in coffee beans; nerve-poisoning alkaloids in potatoes, tomatoes and eggplants. The cat-attracting chemicals in catnip are actually very good insect repellents. The vast majority of plants are inedible by us. Even so we are at risk of poisoning if cattle or sheep graze on them. Abraham Lincoln’s mother died when she drank milk of cows that had grazed on snakeroot. A California infant was born deformed when fed milk from a goat that had been eating lupin. The concept that “natural is harmless” is simply false.
Ames has published numerous estimates of the amounts of natural pesticides that we eat every day. He calculates that we eat about 10,000 times more natural pesticides than synthetic pesticides. More usefully, he and his coworkers have attempted to estimate the relationship between the amounts of different chemicals we are exposed to, and their potency as carcinogens. After all, it is the dosage that makes the poison, to coin a phrase. Some of his calculations are shown in Table 1, rewritten from Ames et al., 1987. The last column (HERP%) is a relative risk. A 5% HERP doesn’t mean a 5% risk of cancer!
|Material||Carcinogen, dose to 70kg person||Rodent Potency||Risk (HERP%)|
|Tap Water||Chloroform, 85 ug||90||0.001*|
|Contaminated Well water||Trichloroethylene, 2800 ug||940||0.004|
|Home air||Formaldehyde, 598 ug||1.5-44||0.6|
|PCB’s, daily||PCB’s 0.2 ug (US average)||1.7-9.6||0.0002*|
|DDT/DDE, daily||DDE, 2.2 ug (US average)||13||0.0003*|
|Bacon, cooked||Nitrosamines, 0.4 ug||0.2||0.003-.006|
|Peanut butter||Aflatoxin, 64 ng/sandwich||0.003||0.03|
|Brown mustard||Allyl isothiocyanate, 5 mg||96||0.07|
|Mushroom, 1 raw||Hydrazines||20-300||0.1|
|Beer, 350 ml||Ethyl alcohol, 18 ml||9110||2.8*|
|Wine, 250 ml||Ethyl alcohol, 30 ml||9110||4.7*|
|Comfrey-pepsin tablets, 9/day||Comfrey root||626||6.2|
|Diet Cola, 350 ml||Saccharin, 95 mg||2143||0.06*|
|Phenacetin pill||Phenacetin, 300 mg||1246-2137||0.3**|
|Phenobarbital, 1 sleeping pill||Phenobaribital, 60 mg||5.5||16***|
|Formaldehyde, industrial||Formaldehyde, 6.1 mg||1.5-44||5.8|
|EDB, industrial exposure||Ethylene dibromide, 150 mg||1.5-5.1||140|
Table 1: Calculated risk factors for common chemicals.
* Material not believed to be gene-damaging; that is, acting as a carcinogen only by irritation or damage at high concentrations.
** Some evidence for increased kidney (renal) cancer after long-term use.
*** Apparently no cancer risk to people taking it for decades.
How then do these theoretical risks relate to the “real world”? A few links can be found. There have been perhaps dozens of cases of liver damage from comfrey-pepsin tablets, although this has been as “hepato-occlusive disease” rather than cancer. These comfrey-pepsin tablets have a risk factor (HERP%) of about six.
Although alcohol is a low-potency carcinogen, large quantities are consumed by some people. Alcoholics have significantly increased risk of cancer in the mouth and throat. Thus HERP’s around five seem to be genuine risks. On the other hand, the HERP value of 16 for one phenobarbital sleeping pill is apparently not connected with any risk of cancer. (Note that phenobarbital is one of the numerous so-called carcinogens that shows up as positive only at tissue-irritating concentrations.)
One interesting point is that TCDD (the dreaded “dioxin” of milk cartons and teabags) is known to cause most of its effects by reacting with an animal component called “Ah receptor”. There are chemicals in broccoli, mainly indole-carbinol, that also react with the Ah receptor. Both chemicals can protect against cancer if administered before challenge with a carcinogen. Both chemicals can promote cancer if administered after the carcinogen has already acted.
Taking potency into account, a 100 g portion of broccoli has 20,000 times more effect on the Ah receptor than a legally allowable TCDD intake of six femtograms/kg/day. (Perhaps it is not surprising, then, that experiments in which rats given a carcinogen were protected by including broccoli or cabbage in their diet. There is evidence that humans too are protected by these vegetables: People who are high-crucifer eaters are significantly less likely to wind up in cancer wards.)
(5) How pesticide regulations and chemical scares diminish public health.
Diet is one of the key routes to better health. Only 9% of the US population eats sufficient fruit and vegetables, higher consumption of these would decrease cancer as well as other diseases. There is plenty of margin to increase fruit and vegetable eating.
To discourage consumption of vegetables and fruits is to diminish public health. Excessively strict limits on harmless levels of synthetic pesticides act to increase vegetable and fruit prices, by reducing production and by increasing cost of production. Thus these regulatory restrictions may well be harming health rather than helping it.
Similar comments could be made about the attacks on Alar a few years ago, when apples disappeared from the lunchboxes of many children.
This then is one reason why Bruce Ames is hated by many “environmentalist” groups. He has shown that they are, in all likelihood, damaging public health under the guise of protecting it against non-existent or unimportant risks.
This review was inspired by an article by Dr Arthur B Robinson in Access to Energy, April 1994.
B.N. Ames. 1983. Dietary carcinogens and anticarcinogens. Science 221: 1256-1262.
B.N. Ames, R. Magaw, and L.S. Gold. 1987. Ranking possible carcinogenic hazards. Science 236: 271-280.
B.N. Ames and L.S. Gold. 1990. Environmental pollution and cancer: some misconceptions. In: Science and the Law (Ed. Peter Huber).
B.N. Ames and L.S. Gold. 1990. Too many rodent carcinogens: mitogenesis increases mutagenesis. PNAS 87: 7772-7776.
B.N. Ames, M. Profet and L.S. Gold. 1990. Dietary pesticides (99.99% all natural), mitogenesis, mutagenesis, and carcinogenesis. PNAS 87: 7777-7781.
B.N. Ames, M. Profet and L.S. Gold. 1990. Nature’s chemicals and synthetic chemicals: comparative toxicology. PNAS 87: 7782-7786.
B.N. Ames, M.K. Shigenaga and T.M. Hagen. 1993. Oxidants, antioxidants, and the degenerative diseases of aging. PNAS 90: 7915-7922.
B.N. Ames. n.d. Does current cancer risk assessment harm health? Published by The George C Marshall Institute, 1730 M Street, N. W., Suite 502, Washington, D. C. 20036-4505. ($US 5.00) [Not seen by me yet — JDM]