FYI: BRCAPRO risk estimation with a strong family history

leaf

This will probably not be very helpful for most people with an increased risk of breast cancer, but I was just reading the breast cancer prevention section in the recent Annals of Internal Medicine (a peer reviewed medical journal), and it mentioned this BRCAPRO model is more accurate than the Gail model http://www.cancer.gov/bcrisktool/ for women with a markedly positive family history.

NOTE: This claim of being 'more accurate than the Gail model for women with a markedly positive family history' is NOT a  breathtaking feat.  In real life applications, the Gail model (in women who do NOT have a personal history of breast cancer) has a less-than-breathtaking concordance statistic of about 0.6. For those unfamiliar with the concordance statistic, it takes a random person from the group (in this case) who never had breast cancer, and compares the model prediction with a person who DID end up having breast cancer.

  If the model worked as well as a coin toss, it would have a concordance statistic of 0.5.  If the model worked perfectly, it would have a concordance statistic of 1.  So, for the Gail model with a concordance value of  about 0.6, about 60% of the time, the person with breast cancer would get a higher score than the person without breast cancer.  However, about 40% of the time the woman withOUT breast cancer would have a higher score than the person who ended up having breast cancer.

It may not be very useful for most patients due to several caveats (see below).  NOTE: I have NOT attempted to download the program.

It is free, but it is 'not convenient'.

a)  It requires registration with the website. It does provide spaces for institutional affiliation.  I don't know if they would allow downloads for patients, or those without an institutional affiliation.  It does have categories for 'student' and 'other'.

b) It has software requirements. It was designed for persons running Windows 2000, NT, or XP.  There is a separate link for those who run Unix.  There is no Mac version, though they claim some users have used it on a Mac with PC emulation.

http://astor.som.jhmi.edu/brcapro/

 I have not downloaded it because a) I do not have a strong family history and b) I do not have a PC, nor do I intend to put PC emulation software on my computer. (I have a Mac.)

In addition, there is a link to other conditions on this UT Southwestern website, but at least some of them seem to link to abstracts. http://www8.utsouthwestern.edu/utsw/cda/dept47834/files/68155.html

More than 50% of all breast cancer occurs in women without known familial or reproductive risk factors.

leaf

You may also be able to download the Tyrer Cuzick model (from 2004), but I am not sure.  The risk of breast cancer was taken from 1994 data.  I have not downloaded  this Tyrer Cuzick model either. It is only available for Windows or 'Pocket PC'.

There *IS* a checkbox for LCIS for a woman's personal history.

Typing '?' represents missing data. If nothing is known about parity then the unknown radio button should be pressed. Similarly if nothing is known about the woman’s menopause the no information radio button should be pressed. If the woman has had atypical hyperplasia, LCIS or some other benign breast disease then the appropriate box should be ticked.

<snip>

At this stage, some risk factors have not yet been included. These include use of exogenous hormones such as HRT and ductal carcinoma in situ. The risk of subsequent invasive cancer associated with ductal carcinoma in situ is large, but it is di􏰃cult to get an accurate estimation of risk, especially with the various preventative treatments that are used when this is diagnosed (surgery, radiotherapy, etc.). It is also likely that even if not an obligate precursor of breast cancer, ductal carcinoma in situ represents a state of breast cancer so far advanced towards invasive carcinoma that it shares, rather than adds to, the other risk factors for breast cancer. (emphasis mine)

http://www.ems-trials.org/riskevaluator/  is the website.

I have not downloaded this model either; I do not have a strong family history, nor do I intend to put PC emulation on my Mac.