Correspondence

Mammographic Density of the Breast

N Engl J Med 2003; 348:174-175January 9, 2003

Article

To the Editor:

Boyd et al. (Sept. 19 issue)1 estimated that the heritability of mammographic density was 60 percent among Australian twins and 67 percent among American twins. Although twin studies reveal little about the mode of inheritance, we previously reported a segregation analysis of 258 multigenerational families that provided evidence of a common major gene (6 percent frequency) that was transmitted in a mendelian fashion.2 Stronger evidence of a genetic influence was subsequently published in the form of a linkage analysis involving pairs of siblings, which used information from DNA analysis, mammographic information, and data on covariates to identify possible regions of linkage to mammographic density on chromosomes 6 and 20.3 Ongoing genomic screening of 900 persons from 91 families will help to verify these results and identify other relevant chromosomal regions.

Since obtaining these initial findings, we have collected additional mammograms and estimated mammographic density using the same techniques4 as Boyd et al. in 1227 female members of 328 families. When we used a technique involving components of variance and incorporating the same covariates considered by Boyd et al.,1 our overall estimate of heritability was 26 percent — less than half that estimated by Boyd et al. The difference may be due to the influence of study design, the age of participants, or other factors. For instance, the mean ages at mammography in the samples of twins were 50 years and 55 years, as compared with a mean age of 60 years in our study. When we stratified the group according to menopausal status (premenopausal vs. postmenopausal), we found that heritability was 59 percent among premenopausal women but only 19 percent among postmenopausal women. The estimates of heritability from both groups reflect a significant familial influence on mammographic density and justify further genetic investigations.

Celine M. Vachon, Ph.D.
Thomas A. Sellers, Ph.D.
V. Shane Pankratz, Ph.D.
Mayo Clinic, Rochester, MN 55905

4 References

1
Boyd NF, Dite GS, Stone J, et al. Heritability of mammographic density, a risk factor for breast cancer. N Engl J Med 2002;347:886-894
Full Text | Web of Science | Medline

2
Pankow JS, Vachon CM, Kuni CC, et al. Genetic analysis of mammographic breast density in adult women: evidence of a gene effect. J Natl Cancer Inst 1997;89:549-556
CrossRef | Web of Science | Medline

3
Vachon CM, King RA, Atwood LD, Sellers TA. A preliminary sibpair linkage analysis of percent mammographic density. J Natl Cancer Inst 1999;91:1778-1779
CrossRef | Web of Science | Medline

4
Byng JW, Yaffe MJ, Jong RA, et al. Analysis of mammographic density and breast cancer risk from digitized mammograms. Radiographics 1998;18:1587-1598
Web of Science | Medline

To the Editor:

The Perspective by Thurfjell1 emphasizes the direct relation between breast density on mammography and the risk of breast cancer. Dense parenchyma reduces the conspicuousness of breast cancer on screening mammography. However, dense breasts are hyperechoic on ultrasonography, and against this background, a hypoechoic invasive breast cancer is actually more conspicuous.

A recent study2 involving a large number of women demonstrated these complementary strengths of mammography and breast ultrasonography. The addition of screening breast ultrasonography in women with dense but mammographically normal breasts increased the rate of diagnosis of nonpalpable invasive cancers by 42 percent. These results, published after those of Boyd et al., suggest that Thurfjell should revise his screening recommendations for women with dense parenchyma who are at particularly high risk for breast cancer.

Ferris M. Hall, M.D.
Beth Israel Deaconess Medical Center, Boston, MA 02215
fhall@caregroup.harvard.edu

2 References

1
Thurfjell E. Breast density and the risk of breast cancer. N Engl J Med 2002;347:866-866
Full Text | Web of Science | Medline

2
Kolb TM, Lichy J, Newhouse JH. Comparison of the performance of screening mammography, physical examination, and breast US and evaluation of factors that influence them: an analysis of 27,825 patient evaluations. Radiology 2002;225:165-175
CrossRef | Web of Science | Medline

Author/Editor Response

Vachon et al. support our finding that the heritability of mammographic density (the proportion of mean-adjusted variance attributed to additive genetic factors1) is high for premenopausal women but suggest that it may be substantially lower for postmenopausal women. We found no evidence that the covariate-adjusted variance or the monozygous-pair and dizygous-pair correlations and covariances depended on age. Specifically, correlations (±SE) for monozygous and dizygous pairs were 0.62±0.03 and 0.28±0.06, respectively, for women 50 years of age or younger and 0.64±0.03 and 0.25±0.06, respectively, for women older than 50 years. The adjusted variances were 306±16 and 306±17, respectively. That is, with the classic twin model, we found that the wide genetic variation in mammographic density evident at 40 years of age persists to 70 years of age. Furthermore, our results were strikingly similar in Australia and North America.

The twin design is more powerful for assessing heritability because of its natural control for age and its use of the difference between monozygous and dizygous pairs of twins in the intrapair sharing of genetic effects.2 Estimating the heritability of mammographic density on the basis of nontwin relatives of different ages, especially across generations, is more problematic. There is the potential for underestimating correlations by failing to account appropriately for cohort effects or any age-specific effects of hormone use, reproductive history, or menopausal status. It is also possible that the trait may be influenced by different genes at different ages.3 We look forward to learning how these issues were addressed by Vachon and colleagues when their work is published.

Norman F. Boyd, M.D.
Ontario Cancer Institute, Toronto, ON M5G 2M9, Canada

John L. Hopper, Ph.D.
University of Melbourne, Melbourne, Australia

3 References

1
Elston RC, Olson MJ, Palmer L, eds. Biostatistical genetics and genetic epidemiology. New York: John Wiley, 2002:371-2.

2
Boomsma D, Busjahn A, Peltonen L. Classical twin studies and beyond. Nat Rev Genet 2002;3:872-882
CrossRef | Web of Science | Medline

3
Boyd NF, Stone J, Martin LJ, et al. The association of breast mitogens with mammographic densities. Br J Cancer 2002;87:876-882
CrossRef | Web of Science | Medline

Author/Editor Response

The results concerning ultrasonographic screening of dense breasts reported by Kolb et al.1 are interesting. However, in only 10 percent of the biopsies (33 of 320) prompted by ultrasonographic findings alone were the specimens malignant. The clinical relevance of the breast cancers detected by ultrasonography alone can only be assumed by comparison with mammographically detected breast cancers. Thus, more research is needed — preferably large, randomized trials, as suggested by Kopans2 — before ultrasonographic breast screening can be recommended.

Erik Thurfjell, M.D., Ph.D.
Uppsala University, S-751 85 Uppsala, Sweden
erik.thurfjell@radiol.uu.se

2 References

1
Kolb TM, Lichy J, Newhouse JH. Comparison of the performance of screening mammography, physical examination, and breast US and evaluation of factors that influence them: an analysis of 27,825 patient evaluations. Radiology 2002;225:165-175
CrossRef | Web of Science | Medline