There are three main theories describing mechanisms which may cause endometriosis. These include

• retrograde flow,
• vascular spread and
• metaplasia of the coelomic epithelium.

Each theory has many supporters. It may be that each of these three theories may accurately describe one of several possible mechanisms. Additionally, an alteration in the immune system (often subtle abnormalities are noted in endometriosis patients) may assist the development of endometriosis in the presence of one of these inciting causes.

In the 1920s a physician named John Sampson devoted a great deal of time to the study of endometriosis and is widely credited with the initial suggestion that retrograde flow is the primary path for this tissue to grow in the pelvis. His reasoning was based on the distribution of implants that he observed.

Additional support for Sampson's theory has been collected over the years and includes

1. virtually all women undergoing laparoscopy at the time of menses can be seen to have blood flowing from the fimbriated end of their fallopian tubes (retrograde flow)
   
2. the most common sites for endometriosis at both laparotomy and laparoscopy are in dependent portions of the pelvis and close to the ends of the tubes
   
3. when recovered surgically and grown in culture endometrial fragments within menstrual blood have the potential to survive and grow
   
4. a higher incidence of endometriosis is observed in women who have obstructions to the normal downward flow (out through the vagina), forcing a greater percentage of flow through the tube in a retrograde fashion
5. there is an increased risk of endometriosis if the woman has a shorter intermenstrual interval, longer duration flow or larger volume of flow
   
6. in female monkeys where the uterus has been surgically transposed (turned around) so that the flow is predominantly into the abdomen rather than out the vagina there is a higher incidence of endometriosis
   

One interesting question regarding this information is

"if virtually all women have retrograde flow of endometrial cells into their pelvis during each menses then why don't virtually all women have endometriosis?"

The answer to this question is not known, however, there are several proposed theories. The most widely accepted theory concerns the immune system. Basically, the immune system can be thought of as a housekeeping mechanism that rejects or destroys any tissue that is "foreign." It can also regulate where a tissue will be allowed to grow within the body.

When the immune system is activated an inflammatory response occurs and these inflammatory cells help to reject the abnormally placed tissues. If there is a problem with the immune response then tissue might be allowed to grow in abnormal locations, that is, to form implants such as endometriosis. In theory, if the amount of tissue is very large it may overwhelm the available normally functioning immune response and result in the growth of this tissue in abnormal locations. Therefore, endometriosis may occur either when a defective immune system is presented with a normal amount of retrograde flow or a normal immune system is presented with a large amount of tissue.

The presence of endometriosis in sites distant from the adnexae (ovaries and tubes) has led to alternate theories.

Vascular spread of endometrial tissue (via the blood vessels or lymphatics supplying the uterus) from the uterus to distant sites may allow this tissue to be transported to areas like the lungs or the knee. Certainly, the uterine walls are highly vascularized and endometrial tissue might find entrance through damaged vessels, such as those that are "broken" during the course of menstrual flow. Direct evidence supporting the ability for endometrial implants to be transported through the vascular system to distant sites where they might implant came in 1981 when diced endometrial tissue from a rabbit was experimentally placed into the vein of the animal's ear and endometriosis implants then developed in the animal's lungs.

Metaplasia (change from one normal type of tissue to another normal type of tissue) of the coelomic epithelium (type of tissue lining the pelvic structures where endometriosis is commonly found and from which uterine endometrial cells are normally derived) allows endometrial tissue to replace other types of tissues outside the uterus. This theory is indirectly supported by the finding of endometriosis in some men who have received estrogen treatments, where no endometrial tissue normally exists. Further support that is more direct comes from research in which rabbit endometrial tissue placed in a diffusion chamber (which does not allow the tissue to leave the chamber but does allow chemical substances produced by the endometrium to leave the chamber) was implanted surgically below the peritoneal lining in these rabbits and some time later excised tissue surrounding these implants were found to have endometriosis like changes. This research suggests that something made by these tissues rather than the tissues themselves may be responsible for endometriosis.

Immunologic defects have been found in women with endometriosis. Both cell mediated and humoral immunologic defects have been noted. Cell mediated abnormalities are suggested by a decrease in cell lysis (cell death) during cytotoxicity assays in women with severe endometriosis as compared to control women. Humoral defects are suggested by the finding of autoantibodies (IgG and IgA types) directed against endometrial tissue in the blood and peritoneal fluids of women with endometriosis. Autoantibodies directed against subcellular components (like phospholipids in the cell walls, possibly resulting in an "anti-phospholipid syndrome") have also been identified in women with endometriosis. The antiphospholipid syndrome is associated with recurrent pregnancy loss (see this other separate section for more information on this syndrome)