There is increasing interest in breast density as a risk factor for breast cancer, in part generated by laws in over 18 states mandating that women be told their breast density when they have a mammogram. The obvious question for the patient and her doctors is, “Does this density increase the risk of breast cancer?” This is not an easy question to answer using just density in isolation. For example:
- Young women have higher density than older women (based on their age) and yet have lower breast cancer risk
- Asians have higher density (based on a lower average BMI) and yet have lower breast cancer risk
The potential role of density based risk models that take into account multiple factors as they interact with density is obvious.
There now appears to be hope, based on the recent publication by Warwick et al, which describes an enhancement of the Tyrer Cuzick model which modifies risk based on “density residual.”
You might ask, what does “density residual” mean? Basically, breast density is strongly dependent on age (On average, the younger the denser) and BMI (On average, the greater the BMI, the lower the density). We have stated in the past that, “Using logistic regression analysis, age and BMI predict density with a 73.4% concordance level.” (Del Carmen MG, Halpern EF, Kopans DB, Moy B, Moore RH, Goss P, Hughes KS: Mammographic breast density and race. American Journal of Roentgenology 2007; 188 (4):1147-50)
In other words, you can predict ACR density 73.4% of the time if you know the age and the BMI. It stands to reason that in patients where the density is commensurate with the age and BMI it is unlikely that density would have an independent effect on risk. Thus the added information from density will likely only occur in the patients where density is greater or lesser than that predicted by age and BMI (i.e., the “density residual”).
Please note that our study used the 4 categories defined by the ACR, while the enhanced Tyrer Cuzick model is based on percentage breast density in increments of 5%. With these finer increments, those falling into the predicted density for their age and BMI will likely be lower than our result of 73.4%, and there should thus be a greater proportion of patients where density will impact risk.
Warwick et al found that by using the “density residual”, the highest breast cancer risk group increased from 14% of controls (75/552) to 26% (145/552) of controls and increased from 21% (15/72) of cases to 39% (28/72) of cases. They also found that women with a high density residual but a low TC risk seem to have a greater chance of developing breast cancer than a similar woman with low density residual but a high TC risk. Thus there was definitely a suggestion in their study that “density residual” would be an important modifying factor in future risk models.
Other risk models that have incorporated density have shown little additional discriminatory power, likely because they did not adjust for BMI and age:
- Tice JA, Cummings SR, Ziv E, Kerlikowske K: Mammographic breast density and the Gail model for breast cancer risk prediction in a screening population.Breast Cancer Res Treat. 2005 Nov;94(2):115-22.
- Tice JA, Cummings SR, Smith-Bindman R, Ichikawa L, Barlow WE, Kerlikowske K: Using clinical factors and mammographic density to estimate breast cancer risk: development and validation of a new predictive model. Ann Intern Med 2008, 148:337–347
- Barlow WE, White E, Ballard-Barbash R, Vacek PM, Titus-Ernstoff L, Carney PA, Tice JA, Buist DS, Geller BM, Rosenberg R, Yankaskas BC, Kerlikowske K: Prospective breast cancer risk prediction model for women undergoing screening mammography. J NatlCancer Inst 2006, 98:1204–1214
- Chen J, Pee D, Ayyagari R, Graubard B, Schairer C, Byrne C, Benichou J, Gail MH: Projecting absolute invasive breast cancer risk in white women with a model that includes mammographic density. J Natl Cancer Inst 2006, 98:1215–1226
The bottom line is that adding some variation of “density residual” (as determined by adjusting density for age and BMI) to a risk model will likely help to identify more women as higher or lower risk. However, adding density to a model without adjusting for age and BMI will likely not be useful.
Warwick et al do warn that their study was developed in high risk women who had participated in the IBIS trial and has not been validated for extrapolation to the general population. Additional validation studies are currently being undertaken. That being said, this approach has a high likelihood of success, and the authors are to be congratulated on a very interesting paper.
We can help you make the models more useful. You can quickly run the various models and compare their results at CRA Health's Risk Express or go to CRAHealth.com to learn more about how our software can make identifying high-risk individuals easy and efficient. We can help you save lives.
About the Author: Kevin S. Hughes, MD, FACS
Kevin S. Hughes, MD, FACS is a co-founder and medical advisor to CRA Health. Dr. Hughes is the Massachusetts General Hospital’s Surgical Director of the Breast Screening Program, Surgical Director of the Breast and Ovarian Cancer Genetics and Risk Assessment Program, and Co-Director of the Avon Comprehensive Breast Evaluation Center, and serves as the Medical Director of the Bermuda Cancer Genetics and Risk Assessment Clinic. He is an Associate Professor of Surgery at Harvard Medical School. Dr. Hughes is actively involved in the establishment of standards and in research regarding the genetics, screening, diagnosis, and treatment of breast cancer. He is the author of numerous papers and book chapters on the subjects of breast cancer, screening, diagnosis and treatment, and risk assessment. More information can be found at: thebreastcancersurgeon.org.