Can a tumor begin to form without any genetic mutations? I'm specifically interested in a tumor which could later lead to cancer.
This is a very interesting question, and one that has been at the heart of cancer research for a long time. I think it separates into three issues.
Are there non-genetic mechanisms that can induce uncontrolled cell proliferation?
Yes, at least in experimental settings. There are many viruses that induce proliferation in its host cell, and viral genes / proteins have been identified that can disable the host cell's cell cycle regulation machinery. A well-studied example is the SV40 Large-T gene (from Simian Virus 40); this gene can be expressed in cells to cause cell proliferation in a transient, reversible manner, and it does so by interacting with host proteins, without causing genetic changes. See for example this article.
Can such genetically normal cells progress to form tumors?
It depends on what you mean by "tumor". To my knowledge, tumors large enough to be a serious medical problem are always found to be genetically altered in one way or another. There are plenty of viruses that cause tumors in humans, for example Epstein-Barr and papilloma (HPV) virus, but in all cases viral sequences are found integrated into the host genome. This may be because development of large tumors requires many cell division, and any non-genetic change that drives proliferation is likely to be lost on the way.
On the other hand, HPV virus frequently causes warts, which technically are small benign tumors. As far as I know, no genetic modification of the host cell is required for this to happen; rather, the virus replicates from its own separate DNA while simultaneously driving keratinocyte cell proliferation to cause these growths. So this example shows that small "tumors" can be formed without any genetic cause. Perhaps there is a limit on how large tumors can grow without genetic changes.
Can genetically normal cells form cancer, meaning a malignant (invasive) tumors?
No, as far as I know, this has never been observed. Large, malignant tumors without genetic changes are unlikely for the reasons above. There are cases of metastatic cancer where the primary tumor is so small that it cannot be located, but in these cases the metastatic cells are again genetically altered. Experimentally, it is possible to produce invasive behavior in cell lines by expressing oncogenes like ras, but these genes are mutated to begin with. Also, normal cells have mechanisms that detect such alterations, causing senescence, and they must first be inactivated by mutations. Transformation to an invasive phenotype is not usually caused by viruses, probably because it would give no advantage to the virus (unlike cell proliferation).
For more information, Robert Weinberg's Cancer textbook does a good job of discussing these issues. This article also provides a historical perspective of tumor virology with some interesting pointers.
- Given the incidence rate of cancer, it is likely that usually multiple somewhat independent events have to occur, so that cancer forms (see the famous interpretation of the underlying data/thinking by Hanahan et Weinberg 2000 )
- Epigenetic changes seem to be a likely factor contributing to cancer ( summary of an experiment, where they are also causal/inducing )
- People have been suspecting, with good rationale, that the intrinsic variability of (bio)chemical reactions, which often only involve a small number of copies of the involved molecules, could contribute to cancer ( e.g.: Ansel et al. 2008 ). (Note: directly addressing this hypothesis for cancer by experiments would be quite difficult / perhaps-impossible)
- None of the above should be seen as opposing the relevance of genetic mutations
I would go with NO for the simple reason: For a cell to become malignant (cancerous), cells need to break several barriers. One is to be able to activate telomerase, an enzyme that restores telomeres and enables infinite cell divisions.
Another important part is to be able to ignore apoptosis. This is programmed cell death. Specific cell signals lead to systematic chemical changes and cell death. So for cancer cells, it is important to be able to ignore this signal.
These are only two basic things, there are more, for a cell to become cancerous.
Changes to this will necessarily lead to change in DNA. However, I don't have direct proof and I don't work in this area.
Endometriosis fits this definition. In many ways endometriosis is like metastatic cancer. Endometriosis is the movement of fragments of endometrial lining out of the uterus (prsumably via the fallopian tubes) to other sites in the body. There these fragments implant and call in a vascular supply to feed them. Endometriosis fragments usually cause trouble in the abdomen and pelvis but they can go anywhere, including skin, lung and the brain. Metastasis is not an inappropriate name for this.
These fragments grow and stick things together (like intestines), causing the same sort of problems that cancer can cause in these sites. They bleed with the menstrual cycle. Endometriosis in the lung can cause hemoptysis. The masses of endometrial tissue and the blood and fluid they accumulate can become endometriomas: benign tumors which are accumulations of tissues and blood which can be mistaken for cancerous tumors on imaging.
The endometrial tissues causing this are not mutant. They are just in the wrong place. Endometriosis can, however, give rise to a true cancer. Clear cell carcinoma is known to form in deposits of endometriosis.