Even three decades ago, there was widespread scientific agreement that the production of cancer could be explained by one of two biological processes. One was called “somatic mutation,” proposing a genetic change in the body cells as the basic mechanism; the other was the invasion of the cell by a virus, endowing the infected cells with cancer properties. Essentially the same choice between the two hypotheses confronts us today, but we can now begin to consider with some intelligence the molecular level of the reaction, and to conceptualize a reasonable reconciliation of these two seemingly divergent view points. Even more important, we can now test many of our hypotheses with actual experiments, using the new tools of electron microscopy, X-ray diffraction, immunochemistry, electrophoresis and other exact physicochemical methods of separation of complex molecules.
In order to account for the basic characteristics of fully developed cancer, with its anaplasia and autonomy, passed on from generation to generation of cells, some permanent change seemingly must occur in the DNA or RNA molecules that compose the genic directorate of the cell. The recent discovery that nucleic acids from the Shope virus become incorporated in the DNA structure of the nucleus offers a reasonable model for the carcinogenic action of at least some of the viruses. As we have noted before, only further investigations will establish whether simpler chemicals, such as the carcinogenic hydrocarbons, can act directly on the cellular DNA or on messenger RNA, or whether an activation of some virus like entity as an intermediate step has to occur. The evidence, it seems to me, is against this latter hypothesis as an essential requirement. For one thing, both ionizing radiation and many chemical carcinogens can produce mutations in the genetic apparatus of the sex cells, and it is hard to imagine that a virus like intermediary must be involved in changes that direct evolution.
Another key question of yesteryear, “One cancer or many?” appears to favor the latter. The diversity of viruses and chemicals that can evoke the neoplastic transformation, and the diversity and behavior of tumors that are produced would suggest different chemical reactions and different sites of the molecular lesion. It must be admitted, however, that much of the diversity could be attributable to the quantitative extent of the lesion.
We must again emphasize that although for the convenience of discussion we refer to cancer, there is no sharp all-or-none difference between normal cells and cancer, but rather a series of transformation from the temporary, benign hyperplastic masses, to benign tumors that remain localized and noninvasive, through dependent neoplasms that can continue to grow and invade only if certain hormonal or other conditions of the host are met, to cancers with varying degrees of normal appearance and normal functions, to the relatively completely independent anaplastic cancer without any recognizable normal functions. This spectrum of characteristics resides in the cells of these growths, and may be maintained for many generations of transplantation to other hosts. In general, however, tumors tend to become more anaplastic with serial transplantation, a development that may be attributable to selection as well as to further cellular alteration.
We have also referred many times to “the cancer cell.” Actually, a single cell in a multicellular organism is almost an abstraction. The stimuli that convert normal cells to cancer cells are not limited to an individual cell, but must affect a population of cells, and it is more than probable that similar changes then take place in many cells of the cellular population. In an investigation of the globulin patterns of tumors arising from plasma cells, the myelomas, it was found that the tumors showed individual globulin patterns, rather than containing a wide variety of such patterns, which suggested single cell points of origin. On the other hand, analysis of the origin of skin cancers following ultraviolet radiation leads to the conclusion that the simultaneous stimulation of a large number of cells is involved. It is likely that both situations hold in different circumstances. Under the specialized conditions of tissue culture, several lines of sarcomas have been established from cell populations that were started from single connective tissue cells.
What are the morphological and the biochemical characteristics of cancer cells that distinguish them from normal cells? A massive amount of observations that have been recorded does not allow us to talk of absolute characteristics that would describe all cancers. Perhaps more realistic, although less general attributes to be sought would be the differences between normal cells of some specific tissue and the cancers that arise from such a tissue. In this narrower range, a few cancer characteristics can be found. Cancer cells tend to be larger than normal cells; the nucleus particularly is enlarged and contains more haematoxylin or Feulgen-staining material. As a consequence, the nuclear-cytoplasmic ratio is often smaller than that of comparable normal cells. A greater proportion of the cells at any given time appear to be in normal or abnormal mitosis. The nucleolus is usually more prominent in cancer cells, and rather often more than one nucleolus is observed.
Biochemically, one of the characteristics of many cancer cells was discovered by Otto Warburg of Berlin over 30 years ago, and is still incompletely exploited in investigational approaches. This is the way that cancer cells utilize glycogen as a source of energy. In almost all normal cells (and, as usual, there are a few vexing exceptions), glycogen is broken down to water and carbon dioxide through a series of steps if the cells have an adequate supply of oxygen. This is known as aerobic glycolysis. If normal cells are deprived of oxygen, the process of sugar breakdown stops at an earlier stage, of lactic acid; this is known as anaerobic glycolysis. Cancer cells, in general, often display anaerobic glycolysis even when adequate oxygen is available, and this may be attributable to abnormalities in enzymes such as phosphohexose isomerase. As a consequence of this incomplete breakdown of glycogen, cancers form and accumulate an excess of lactic acid. When a rat with a transplantable cancer is given a large amount of glucose, there is a drop in the pH of the tumor because of the increase of lactic acid. A recent extension of these observations is the finding of increased lactic dehydrogenase in cancer tissue and in the blood of some animals and patients with cancer.
Systematic searches for differences in enzyme patterns in cancer as compared with normal tissues have been rather frustrating. One of the larger investigations was undertaken by the late Jesse Greenstein of the National Cancer Institute. He concluded that transplanted cancers tended to converge toward a similar enzyme profile whereas normal tissues reflected their specific functions in distinct enzyme patterns characteristic for the tissue in question. Fifteen years later it is easy to criticize this work, in two ways. The first is that well established transplantable tumors were used for most of his determinations, and these tumors do through selection tend to be rather similar. When “spontaneous” cancers of animals or man are used for such studies, there appears to be as wide a diversity of patterns as among normal tissues. The second criticism is that the methods for measuring enzyme activity were usually limited and, by present standards, now antiquated; for example, many measurements were carried out at a single pH or temperature, rather than over a range of conditions.
A favorite material for comparative biochemical studies has been the liver and the benign and malignant hepatomas that can be induced by azo dyes and other chemicals or that arise “spontaneously.” Harold Morris of the National Cancer Institute, in his observations on the characteristics of a series of transplantable hepatomas, found one which is practically indistinguishable from normal liver in its enzymatic patterns. This “minimal deviation” tumor casts some doubt that as a class cancers will be typified by their enzyme activities.
We have commented on the truism that a carcinogenic chemical or a cancer virus is not cancer. By the same token, a cancer cell is not cancer. For the disease characteristics of cancer are manifested not by single cells but by a large population of such cells and their interaction with the normal processes of the host. The cocarcinogenesis work of Berenblum, for example, suggests that individual cells possessing the potentials of cancer may lie dormant or nonmanifest for protracted periods. And, although leukemia has been transplanted successfully in mice by the use of single cells, usually a dose of several thousand cells is required for the passage of tumors from one genetically identical animal to another. Thus, cancer as an entity starts as a colony of cancer cells, with the resultant development of interplay of factors between the individual cells of the colony, and the reactions between the cancer colony and the host in which the colony exists.
In order to survive and to grow, the cancer must derive a nutrient supply from the host. For most solid tumors, this requires the acquisition of a blood supply. Observations on the larger blood vessels of tumors, by means of X-rays and radio-opaque media, indicate that such vessels are increased in number and tortuosity and many even form patterns that are helpful in the localization of the mass. However, microtechnique studies on skin tissues indicate that the increased blood supply does not represent better oxygenation and nutrition, for the blood flow is stagnant.
At the boundary of the tumor and the normal tissue there is no definite capsule, and no inflammatory reaction. The latter is often a feature in transplantable tumors, especially when there is some degree of genetic difference between tumor and host. Histochemical stains of the interphase indicate the presence of increased amounts of proteolytic enzymes, probably involved in the invasive destruction of normal tissues.
The cancer colony in its general attributes appears to be composed not of virile vandals but of maladjusted compulsive neurotics. The reserve of the cancer cells in regard to recovery after two forms of injury perhaps permits this word picture.
In tissue culture, cancer cells in general are not more sensitive to ionizing radiation than many normal cells, but in tissues many cancers can be destroyed at doses that allow recovery of the normal tissues. George Crile, Jr., of Cleveland recently reported that many cancers are destroyed by heat, at temperatures that are tolerated by the normal tissues.
The cancer colony is less cohesive than organized tissue, and it has been reported that calcium is decreased in cancer tissue. Many cancer cells are motile, as are normal cells under tissue culture conditions. With some transplanted tumors, cancer cells can be isolated from the liver of the animal within hours after the implantation of the tumor under the skin. Such individual or very small clumps of cells are destroyed, since they do not give rise to metastatic masses. Recently considerable interest has been devoted to the identification of cancer cells in the blood and other tissue fluids of patients with cancer. No clear relationship is demonstrable, however, to the recurrence or metastasis of cancer and such detection. This is attributable to at least three factors, the first one being that there is considerable error in the identification of such cells as cancer. The second is that even correct morphological identification does not mean that such cells are alive in the sense of being able to reproduce. The third is that most individual cancer cells are destroyed before they can lodge in a capillary, attach themselves to the blood vessel wall, and penetrate into the deeper tissue where they can divide and set up a new colony, a metastasis.