Refining Precision Prevention for Benign Breast Disease

Future breast cancer risk following a diagnosis of benign breast disease (BBD) has primarily been quantified for BBD diagnoses that fall within known high-risk categories, such as atypical ductal and lobular hyperplasia and lobular carcinoma in situ. Many questions remain, however, for accurate risk classification associated with the lower-risk categories of nonproliferative lesions and proliferative changes (or lesions) without atypia (PWoA), such as papilloma or papillomatosis, usual ductal hyperplasia, radial scar, columnar cell hyperplasia, and hyperplasia not otherwise specified (NOS). The 4 categories are part of the Breast Cancer Surveillance Consortium (BCSC) Breast Cancer Risk Calculator Version 2.

Additional questions remain on potential associations between BBD diagnoses and breast density.

Olivia Sattayapiwat, MS, MPH

Image Credit: Breast Cancer Surveillance Consortium

As part of her research conducted for her dissertation, Olivia Sattayapiwat, MS, MPH, University of California at Davis, graduate group in epidemiology, focused on refining the prediction of 5-year risk of invasive breast cancer risk among women with BBD whose disease fell into the PWoA risk category, as well as evaluating if breast density or presence of calcifications on biopsy influenced these associations. She presented her findings on day 2 of the San Antonio Breast Cancer Symposium in the session, "Advancing Evidence of the Associations Between Specific Benign Breast Diagnoses and Future Breast Cancer Risk."

Her patient population was 1.3 million women from the BCSC who did not have a history of breast cancer. They represented 5.3 million mammograms between 1996 and 2019 and were aged 40 to 79 years.

For BI-RADS density levels (A, almost entirely fatty to D, extremely dense), using no BBD as a reference, women with a PWoA diagnosis who did and did not have calcifications had elevated risks across the board. The higher risks were as follows:

• With calcifications:
  • Almost entirely fat: 139% (HR, 2.39; 95% CI, 1.23-4.63; P = .0098)
  • Scattered fibroglandular densities: 72% (HR, 1.72; 95% CI, 1.38-2.14; P < .001)
  • Heterogeneously dense: 102% (HR, 2.02; 95% CI, 1.72-2.38; P < .001)
  • Extremely dense: 55% (HR, 1.55, 95% CI, 1.06-2.27; P = .0224)
• Without calcifications:
  • Almost entirely fat: 33% (HR, 1.33; 95% CI, 0.68-2.6; P = .4124)
  • Scattered fibroglandular densities: 69% (HR, 1.69; 95% CI, 1.41-20.2; P < .001)
  • Heterogeneously dense: 52% (HR, 1.52; 95% CI, 1.3-1.79; P < .001)
  • Extremely dense: 44% (HR, 1.44; 95% CI, 1.01-2.06; P = .0426)

The 5-year risk per 100 was further classified as low (HR < 1.0%), average (HR, 1.0%-1.66%), intermediate (HR, 1.67%-2.49%), high (HR, 2.5%-3.99%), and very high (HR, > 4.0%), and Sattayapiwat and her team used classification trees to group combinations of PWoA diagnoses, breast density, age, and presence of calcifications with similar magnitudes of associations with 5-year risk of invasive disease. Results were then stratified by dense vs nondense breast tissue.

Among women aged 60 to 79 years with or without calcifications who had a BI-RADS breast density of C or D and whose BBD diagnosis was PWoA, having papillomas/radial scar was associated with a very high 5-year risk of invasive cancer of 4.52%, usual ductal hyperplasia had a high risk of 3.92%, columnar cell hyperplasia/PWoA NOS had a 3.01% intermediate risk, and papillomatosis with no prior biopsy (97.5% of the study cohort), 1.83%. For the women aged 40 to 59 years, the trend was similar, with the highest risks seen for PWoA among women with calcifications (2.58%), PWoA NOS (without calcifications) (2.08%), and usual ductal hyperplasia (without calcifications) (1.67%). Overall, categories of risk for the older women ranged from intermediate to very high, and for the younger women, average to high.

Older and younger women with a BI-RADS breast density of A or B, with or without calcifications, and whose BBD diagnosis, too, was PWoA, had elevated 5-year risks of invasive breast cancer that followed similar patterns as their counterparts with more dense breast tissue. The highest risks of 3.64% and 3.13% were seen for papillomas and usual ductal hyperplasia, respectively, among the older women, while among the younger cohort, papillomas again carried a higher risk of 2.04% and usual ductal hyperplasia/PWoA NOS (with calcifications), a 1.83% elevated risk. There were 3 overall categories of risk for each age group in this stratification, but those for the older women were higher overall (from average to high risk) vs the younger women (low to intermediate).

Overall, for women with dense breast tissue, calcifications were shown to modify the 5-year risk of breast cancer for those in 2 categories:

• Among the women aged 40 to 59 years who had a PWoA diagnosis of papillomas, usual ductal hyperplasia, columnar cell hyperplasia, and PWoA NOS—although they represented just 0.5% of the total study cohort—the 2.58% elevated risk places them in the high-risk category
• Among the women aged 40 to 59 years who had a PWoA diagnosis of usual ductal hyperplasia or PWoA NOS—0.3% of the overall study cohort—the 1.83% elevated risk places them in the intermediate-risk category vs those without calcifications

In addition, for those with nondense breast tissue (or fatty breasts), the women aged 40 to 59 years—12.6% of the overall study cohort—had a low risk of breast cancer regardless of BBD diagnosis, and the women aged 60 to 79 years—17.1% of the overall study cohort—had an immediate 5-year risk if their PWoA diagnosis was papillomas, usual ductal hyperplasia, or PWoA NOS.

“Women with a PWA BBD diagnoses and calcifications had elevated risk for breast cancer in all levels of breast density,” Sattayapiwat concluded. “Specific BBD diagnoses and the presence of calcifications can change a woman's predicted 5-year breast cancer risk compared with broad BBD categories alone.”

Anne Marie McCarthy, PhD

Image Credit: University of Pennsylvania

Acknowledging Sattayapiwat’s determination that these data could be incorporated into BBD risk prediction models to improve model accuracy for precision prevention, Anne Marie McCarthy, PhD, cancer epidemiologist and assistant professor at the University of Pennsylvania, who discussed her findings said that such approaches in breast cancer prevention have lagged behind.

She noted the confusion many women feel when it’s time to initiate breast cancer screening, due to all of the variables involved. Their questions often include when to start screening, how often to screen, is mammography enough or will supplemental screening be necessary, and thinking about chemoprevention. At present, there are no guidelines on follow-up from the National Comprehensive Cancer Network for patients who have PWoA, except among those with a lifetime risk of 20% or more.

“The findings by Sattayapiwat and colleagues really give some evidence to be able to better define risk and do precision prevention for patients with BBD,” she declared, especially with this being the largest study to date of potential future breast cancer risk among patients who have PWoA.

The integrative fashion of this analysis, with its use of classification trees, facilitates the identification of subgroups of patients, she added; echoes what clinicians already do when they see patients in the clinic; and highlights both challenges to and opportunities for progress.

“Differentiating risk for specific types of PWoA may enable precision prevention,” she said. “Better tracking of PWoA diagnoses and clear follow-up guidelines are needed to clinically implement risk-based management for these patients.”