One of the most serious claims in the class action lawsuits against Dow Corning Corp. during the "silicone crisis" involving breast implants in the late 1980's was that silicone breast implants caused breast cancer and/or delays in diagnosis of breast cancer. Despite there being no evidence for it actually happening, these were reasonable questions to ask. Over the last 20 years, we've been flooded with data that has been reassuring on these issues.
Implants do make conventional mammograms harder to interpret by their "shadow", but the increased ease of doing manual exams by having the implant to push against to feel lesions compensates a great degree. An MRI mammogram can be used to supplement mammograms when needed for better imaging for screening.
One of the more interesting findings in several of the large series of women with implants was the observation of significantly lower (almost 40%) rates of breast cancer in the implant group versus a control population of women without implants. The intuitive reason for this has been that these women with implants were a self-selected (rather then "randomly selected") group who were likely to be healthier and have less breast tissue, which both should lowered their expected rates.
To really sort out a true "expected rate" for breast cancer, you'd have to do some herculean effort of better characterizing the individual risks with a tool like one of the "Gail Model's" of the study participants, which is almost impossible in such large trials. The suggestion that the presence of implants themselves was protective wasn't really taken seriously. There could however, be something that makes us look at this issue a little closer.
I came across a pre-publication in the journal Aesthetic Plastic Surgery entitled, "Breast Implants as a Preventive Factor" describing the differential temperature seen on thermography (a imaging technique that shows temperature) from experimental rats with silicone implants placed and the resultant affect that had on local circulating hormone levels and cellular abnormalities (both of which were decreased in the implant group). Now this was only an animal model mind you, but it immediately occurred to me that maybe part of that effect we were seeing was from this phenomena. Interesting stuff!