If they only knew their place we wouldn't have all of this trouble.

If tumors would not spread (metastasize), the treatment of them would be considerably simplified; they could simply be cut out. It is their ability to spread which makes for their lethality.

Before discussing metastasis, I would like to make a distinction between tumors of white blood cells, such as lymphocytes, that normally wander, and when they become tumors continue to do the same, and cells of the skin, breast, intestine, and so on that have to undergo some change in their normal behavior to be malignant (i.e., they have to invade, and some will metastasize). I am going to confine the discussion in this chapter entirely to those cells that have to change their behavior (epithelium).

A popular misconception is that once the cells of a tumor enter the blood stream, the individual bearing the tumor is doomed because they are bound to metastasize. Nothing could be further from the truth. An experiment that has been repeated over and over again (in mice) is one in which a breast cancer which does not ordinarily metastasize is massaged; the animal is killed and the lungs of the animal are looked at under the microscope. Invariably, there are large numbers of cancer cells in the lung. In spite of this, other animals that are treated in the same way with the same tumor and are allowed to live, do not develop metastasis. With some other lines of the same type of tumor, lung metastasis develops even if the tumor is not massaged. The body apparently has some mechanism for destroying cells that are not where they are supposed to be.
An interesting experiment was done by Paul Weiss and Gert Andres, in which they injected cells from one chick embryo into another chick embryo. They used pigmented cells from a pigmented chicken, which were injected into an albino (no pigment). They found that the pigment cells did not stay where they were not supposed to. Some of the injected cells did settle down and grow where they were supposed to; in the skin. In other words, either the cells knew where they were going, or the body knew where they belonged, or both.
The work of Cameron Wallace on the metastasis of mouse breast cancer showed that some tumors metastasize readily while others do not. My own experience with mouse breast cancer confirms this. There is no way of telling by looking at the tumor under the microscope, what it is going to do. If we take several lines of transplantable breast cancer, which are microscopically indistinguishable, and implant them into ten animals each, one transplanted line may show extensive metastasis in ten out of the ten animals, while other lines will show an occasional metastasis. The implication of this is that if a tumor is of the metastasizing kind, it will probably do so early in the game; and if it is the non-metastasizing kind, it takes a lot of tumor cells to get a metastasis started. Even a "non-metastasizing" tumor will metastasize if given sufficient opportunity. It would be a mistake to draw extensive conclusions on the basis of this very meager evidence. There is a crying need for more research on what makes one tumor spread, and another stay in the same place. Until more evidence is forthcoming, it is necessary to use the surgeon's law of "When in doubt, cut it out," and add "cut it out while it is still small."

What is it then that makes tumors metastasize? There is some indication that the ability to metastasize is in some way related to the ability of the tumor to grow in an extraordinary site. This is also related to how much of the tumor gets there. If you put enough tumor cells into a site, you will eventually get growth, even though the tumor would not ordinarily grow there. But this is not the ordinary state of affairs. What happens more often is that some tumors have the extraordinary ability to grow almost anywhere, even if small numbers of cells get there. There is also some evidence that certain tumors grow better in some organs than they do in others, and that it is not simply the trapping of cells that causes metastasis to appear.

The probability of a surgical cure for cancer is related to the ability of the tumor to metastasize. There are a few tumor types that are known to metastasize very readily, while some types do not. With most tumors, there is no sure way of telling by looking at them. This is certainly true with regard to breast tumors in mice. The only way that a scientist can find out if it will metastasize or not is to inoculate it into mice of the same strain, and see what happens. Your cancer specialist should be able to give you a better indication than I can of the odds that any particular type of human tumor will metastasize.
By the time a tumor is detected, some of the cells will probably have already spread. If the tumor is of the metastasizing kind, and only a few cells are able to produce a metastasis, then a surgical cure is unlikely. If, on the other hand, it is not of the readily metastasizing kind, then the surgical removal of the tumor will probably result in a complete cure. There is no way of telling by looking at the tumor under the microscope whether the tumor is of the metastasizing kind or not. Some odds can be given for different kinds of tumors, but they are only odds and tell you little about what will actually happen to an individual. The odds of being cured by an operation can be supplied by the physician; but they are only odds.
Suppose the tumor does not metastasize readily, what then? Again, the odds of the cells having left the tumor are pretty high, but the tumor cells have probably already been destroyed by the body. In this case, surgical removal of the tumor will probably have cured the patient. Suppose that the victim does not go to the doctor, and allows the tumor to grow to a large size? In this case, there is a very good possibility that the tumor (that does not readily metastasize) may metastasize because cells are being literally poured into the body. In this case, the probability of the cure will have been changed from a1most certainty to something considerably less than that. Most tumors fall between these two extremes, so that the patient having the tumor can change the odds of a cure considerably by having the tumor removed at an early stage in its development. Considerable inroads in mortality due to cancer of the uterine cervix have been made by detecting abnormal growths before they have a chance to grow appreciably, or spread.
I would like to point out that no one has devised a way of determining, by looking at the tumor, whether it has spread or not. If tumors are detected at distant sites, then we know that it has spread. If they are not detected, then the only recourse is to wait and see if they appear. If they do not appear, they have not spread; and if they do appear then they have; it's as simple as that.