Looking Upstream
by Jonathan King

In December 2005, the National Institutes of Health (NIH) announced that it would be funding The Cancer Genome Atlas, beginning “with a pilot project to… systematically explore the universe of genomic changes involved in all types of human cancer.” The initiative applies the newest gene-sequencing technologies to identify the different forms of gene and chromosome damage observed in human cancers. Such information is of great interest to the patient, medical and pharmaceutical communities because it might open new paths to developing therapeutic and anti-tumor agents specific for the altered cell functions in the diverse forms of cancer.

However the potential health significance is even greater, since if properly conceived and applied, the detailed characterization of the genetic damage could be used to identify the mutagens and carcinogens causing this damage. Rather than accepting the incidence of cancer as a fact and focusing solely on treatment, researchers could identify the etiological agents of many human cancers, opening the possibility of the only true method of cancer prevention, the elimination of exposure to carcinogens.

Disappointingly, but predictably, signs suggest the project will be blind to such a preventive strategy. Singularly lacking from the 1,200 word NIH announcement of their decision to fund of this project, and from all of the follow-ups, was any mention of identifying actual etiological agents of cancer, the primarily chemical mutagens or carcinogens that cause the mutations and other damage. Also lacking was any mention of the existing Atlas of United States Mortality by County, which captures some of the environmental and occupational origins of cancers in the US. [Pickle, Linda W., et al. (1996) US. Dept of Health and Human Services, Hyatssville, Maryland (DHHS Pub # (PHS)97-1015)]

The subtle message transmitted by the public presentation of the cancer genome project is that the cancer problem is located in our genes. Readers of GeneWatch are familiar with this tendency to blame our genes for our condition of life. According to NIH genetics chief Francis Collins, the project will “enumerate the complete list of genomic insurgents that lead to cancer.” Articles in newspapers across the country described the program as intended to identify the “cancer genes” associated with most human cancers. The message transmitted to the reader is that cancer is due to tissues that carry altered genes — sometimes called oncogenes — and the NIH effort will identify these damaged genes in the various tumors that affect us. The effort to identify the genetic damage in cancer cells is indeed critical to progress in preventing cancer and in generating better treatments for those afflicted. However, the description of the genome atlas does not bode well for the use of this information for preventing
cancer.

The distortion of the current representation is particularly obvious when we consider lung cancer. Clearly, any program that identifies the damaged genes in lung epithelial cells that are altered in lung cancer is valuable. However, to fail to point out that such genetic changes are due to exposure to carcinogens present in cigarette smoke masks the real problem — the carcinogens in cigarette smoke.

Why do we get cancer?

Do we get cancer because control of the proliferation of our cells depends on genes whose damage leads to cancer? Or do we get cancer because we are exposed to environmental carcinogens? The first model feeds into the current interests of the pharmaceutical industry and modern medicine, i.e. to accept the occurrence of disease in the population and sell the patient some therapy, treatment or procedure to alleviate the condition. This also conforms to modern medical practice, which focuses on the illness of the individual, and generally ignores the conditions that led to the disease in the population. The focus on the diseased individual further shifts liability and responsibility away from the corporations, institutions or processes that generated the carcinogens inducing the cancer.

An alternative formulation is to identify the agents that cause the disease, and either remove them from the human environment or help people protect themselves from exposure. However, this policy brings biomedical scientists and physicians into conflict with the social and economic processes that generate the carcinogens: the tobacco industry with respect to cigarettes, chemical manufacturers and the petrochemical industry with respect to industrial chemicals, pesticides and herbicides. Vivid examples of this resistance have been described in books like Paul Brodeur's Expendable Americans, Samuel Epstein's The Politics of Cancer and Robert Proctor's Cancer Wars. Even earlier, Walter Hueper, the principal scientist attempting to bring the relationship between synthetic aniline dyes and bladder cancer to medical attention and a leader in the development of the National Cancer Institute, was prevented from becoming its Director due to opposition from the industry.

Molecular basis of tumor development

The NIH announcement and much National Cancer Institute (NCI) literature refer to cancer as a complex disease. There is some truth in this; for most tumors, multiple alterations are needed to transform the normal cells into cancer cells. Since the many regulatory processes controlling and inhibiting cell proliferation differ from tissue to tissue, the damage pathway to cancer is different for different tissues. However, one of the great contributions of NIH sponsored research over the last three decades has been to reveal the underlying nature of cancer. It may be useful to review this, so as to demystify the disease.

Cancer represents the uncontrolled proliferation of a particular cell type. The actual pathology and course of the disease depends on the cells and the tissue of origin. Lung cancer has a different course than liver cancer, but they both represent cells that are dividing when and where they should not be. Malignancy generally represents the ability of the cancer cells to overcome the constraining signals imposed by normal contact inhibition by cells that surround it in the affected tissues and organs.

Cancer is not transmissible from individual to individual, and the great majority of cases are not inherited. Rather, like neurological damage from mercury poisoning, anemia from lead poisoning, black lung disease from coal dust, or bysinnosis from cotton dust, cancer is due to cell damage from exposure of the cells to toxic agents. In the case of cancer, the ultimate targets of the damage are the genes and chromosomes of the affected cells. Thus, the progeny of tumor cells express similar properties as their altered parent cells, since they carry similar chromosomal damage.

A major breakthrough in basic research was the recognition that, in cancer cells, damage was to the genes and chromosomes of affected cells. Such damage includes single nucleotide mutations and more complex forms of DNA damage — deletions, translocations, and duplications — in genes controlling cell proliferation. These changes are transmitted to daughter cells, which continue to pass on the damaged genes, eventually generating a tumor. Since different sets and subsets of genes are controlling proliferation of liver cells, as opposed to lung cells, the genetic targets for transforming cells from normal to malignant cells differ in different tissues. Over the past decade, cancer researchers have discovered that many of these genes involved in encoding proteins regulate the cell cycle, and thus cell division.

Another important step forward was the biochemical identification of the pathway to DNA damage. For many organic chemicals, the carcinogenic species are modified forms, activated by detoxifying enzymes of the liver or other tissues, as part of the body's effort to rid itself of these modified species. Transport of the activated species into the nucleus of a cell results in their attack on the nucleotides in DNA, which generates mutations. For a number of human carcinogens, including poly-aromatic hydrocarbons, aflatoxin, acridine dyes and some metals, the chemical character of the DNA modification reactions are known in considerable detail. For these and many other carcinogens, the tumor induction process has been extensively studied in rats, mice and other experimental animals, as well as in cultures of human cells and tissues.

For a number of colon, skin and breast cancers, the damaged genes leading to the malignancy have been identified. The tumor cells in some women with breast cancer exhibit damage to the essential genes, but the rest of the breast cells and other tissues in their bodies carry the normal, functional versions of these genes.
The relatively rare cases of inherited cancers have been important in solidifying this interpretation. In these cases the mutations occurred in the germ line of the parents, or earlier generations, and were inherited by the children. The mutations were therefore transmitted to all the cells of the offspring. Thus, for example, all the cells in women with an inherited propensity to develop breast cancer because of mutations in their BRCA1 and BRCA2 genes have the same genetic damage to the BRCA1 and BRCA2 genes in both their cancers and their other cells. These individuals represent a distinct, but small, fraction of who currently have breast cancer. Examination of the graph of those with lung cancer makes it immediately clear that this change in cancer frequency within a single generation could not be due to inherited mutations, since the disease was coupled to the individuals' exposure to cigarette smoke, not their parents. It has been important to recognize that as long as the genetic damage is not in the germ line — sperm or eggs — damage will not be transmitted to subsequent generations.

When Robert A. Weinberg and coworkers first showed that the tumorigenic property of human bladder carcinoma cells was associated with mutations in a single gene, it was a dramatic step forward in establishing that cancer was due to somatic mutation. The genes carrying these mutations were named “oncogenes.” The experimental work included clear evidence that the oncogenes were not inherited by the affected individuals, but had occurred in the cells generating the tumors. That is, they were almost certainly the product of mutations occurring in the bladder cells of individuals who had contracted bladder cancer. Nonetheless, the dominant message in hundreds of subsequent papers has been that oncogenes themselves are the cause of cancer, rather than that damage to specific genes (which generates oncogenes) is the cause. The difference may seem subtle or semantic, but it is profound in terms of the of the formulation of policies for reducing — not treating — the incidence of human cancer. There is compelling evidence that many human bladder tumors are due to exposure to aromatic amines and related chemical carcinogens.

Known Human Carcinogens

The recent publicity given to inherited forms of cancer has failed to point out, in both the popular and scientific literature, that there is a high correlation of the incidence of cancer with where people live and work, and with conditions and character of their workplace.

One of the earliest recorded identification of a chemical carcinogen was the 1775 description by the London physician Percival Potts that the high incidence of scrotal cancer in young London chimney sweeps was due to their exposure to coal tar and coal combustion products. Thus began the long history of identification of cancers and corresponding carcinogens. Over the next two centuries, many human carcinogens were identified due to high incidences of particular cancers in a variety of workplaces and environments. These resulting cancers included bladder cancer from aniline dyes, liver cancer from aflatoxin produced by fungi, sarcomas in watch dial painters from exposure to radium, mesothelioma from asbestos in shipyard workers, and many other examples.

The medical community was relatively early in recognizing the cancer danger from exposure to radioactive substances. The first mines for radium and uranium were in Joachimstal & Schneeburg in the Black Forest region of Germany. Local physicians Harting and Hesse noted the high mortality among the miners and reported an unusually high frequency of deaths due to lung cancer. Experience with exposure to radioactive elements, together with the medical follow up on the survivors of Hiroshima and Nagasaki, led to increased awareness of sources of radioactive exposure and cancer: uranium from mining and processing, strontium 90 from atomic bomb testing, plutonium from nuclear reactors and spent nuclear fuel.

Public understanding that cancer was induced by external agents developed out of the linking of lung cancer to cigarette smoking. Initially, this causation was denied, as cancer incidence was not well correlated with simultaneous cigarette smoking. As epidemiologists continued to collect data on the relationship of smoking and health, evidence of the 20-30 year lag between initial exposure and cancer began to appear, and it became clear that the lung cancer was closely associated with previous history of smoking. Though this correlation was resisted by the tobacco industry for decades, the scale of the tragedy and extent of the data eventually overwhelmed such resistance.

With the increasing strength of the environmental movement in the 1960s and 1970s, many studies that linked human cancers to specific carcinogens were brought to the public’s attention, including Irving Selikoff's investigation of the relationship between asbestos exposure and mesothelioma and Thomas Mancuso's studies of radiation workers. The collection and publication of cancer incidence data in the US by geographical location was of great importance to the advancement of this understanding. Cancer incidence data revealed variations far too high to be accounted for by genetic variations in the US population. In most of these cases, the efforts to publicize the identity of the carcinogen were steadfastly resisted by the producers of these carcinogens and their allies.

Studies on people who had changed their environment were also important in advancing this understanding. Thus, for example, Japan has a much higher incidence of stomach and liver cancer than America. However, among Japanese people who migrate to the US, there is a lowered incidence. Their children, raised in the US, exhibit the same incidence as the general US population. Conversely, incidence of colon cancer is lower among Japanese males than American males, but it increases in Japanese men who have migrated to California.

Lung cancer is not the only human cancer strongly tied to exposure to distinctive chemical carcinogens. Striking data has long existed describing the relationship between bladder cancer and exposure to aromatic amines, which include aniline dyes. The effort to generate synthetic purple textile dyes to replace natural indigos was key to the launching of the German synthetic chemical industry in 1856. By 1900, workers in the industry and physicians observing them knew that those who worked in dyeing plants had a high probability of contracting cancer of the bladder.

British studies revealed that risk of bladder cancer in aniline dye industry workers was 33 times greater than that of a control population. Before 1930, one in every four workers in the industry contacted bladder cancer; by 1950 the incidence had been reduced, but was still a tragically high one in six. By the 1940s, experimental evidence emerged from work of Walter Hueper and colleagues at the newly established National Cancer Institute (which was the original NIH Institute) where they showed that treating dogs with aniline dyes induced bladder tumors.

After World War II, manufacture of aniline dyes and related aromatic amines was established in the US. In 1955 Melick et. al. reported the first cases of bladder cancer in the US: 19 of 74 men exposed to 4-aminodiphenyl, used as anti-oxidant in the rubber industry. By 1958, the incidence was up to 36% of those exposed. A later study of another cohort of 366 workers revealed bladder cancer in 26%. [Goldwater, Rosso and Kelinfeld (1965) Bladder Tumors in a Coal Tar Dye Plant (Arch. Env. Health, 11, 814-819)]

Indeed, if one examines the map of the bladder cancer incidence in the Atlas of United States Mortality by County, among the highest incidences are found in the region including Essex County, New Jersey. This was the site of the largest aromatic amine-producing plant in the US. There are also hot-spots in a number of midwestern cities, historically associated with the manufacture of tires. One of the main consumers for aromatic amines were tire manufacturers who used them as anti-oxidants in rubber processing and tire production.

The Role of the Medical and Pharmaceutical Industries:

For the past 30 years, the cancer research establishment has moved steadily away from a focus on identifying carcinogens, to a focus on the cancer patient and the development of therapies and cures. Of course, this latter component needs our full attention. Then again, preventing cancer from developing goes much further in preventing and reducing human suffering. Unfortunately, preventing cancer does not generate a market for consumables.

The pharmaceutical industry is currently the most profitable sector in the US economy. This is due in part to an increasing population that either is ill or is concerned over health issues, which provides a wide market.

The intensive interest in patenting human genes by companies such as Myriad, Incyte and Celera is driven by the disease model of "oncogenes cause cancer." The initial market is for diagnostics tools, in which the patent monopoloy insures profitability (See Gene-watch 15-4). Their long term business plans could involve gene replacement or gene therapy, where patent monopolies will also insure substantive profits. Of course companies invested in such business strategies have no interest in research that focuses on the causes of cancer, and reducing the incidence of gene damage in the population.

However, if we want to protect ourselves from the scourge of cancer in the long run, we need to recognize that the largest social and health benefit will accrue from identifying the causative agent and removing it from the human environment.

Jonathan King, PhD, is Professor of Molecular Biology at MIT and a founder of the Council for Responsible Genetics .