Abstract 2528: Age and body mass index are predictive of cancer diagnosis among women with BIRADS 4 mammograms.

Abstract

Listen

Translate article

Abstract Background: Women with BIRADS 4 mammograms are referred for biopsy and have a 15-30% one year risk of breast cancer. Identifying pre-test predictors of cancer among women with BIRADS 4 mammograms may be useful in reducing false-positives. Methods: Women with BIRADS 4 mammograms were recruited from the University of Pennsylvania Health System from 2009-2012. Women <40 years with personal history of breast or ovarian cancer, mantle radiation, or BRCA 1/2 mutation were excluded. Women completed a breast cancer risk factor questionnaire and were followed prospectively for definitive biopsy outcome. Differences in baseline characteristics of women diagnosed with DCIS or invasive breast cancer compared to women with benign biopsy findings were assessed using Fisher's exact tests. Logistic regression was used to estimate the odds of cancer diagnosis by breast cancer risk factors. Results: Among 269 participants with BIRADS 4 mammogram, 41 (15%) were diagnosed with cancer (DCIS N=16, Invasive N=25). Women diagnosed with cancer were significantly older than women with benign findings (p<0.01). A higher proportion of obese women were found to have cancer than non-obese women, though this was not statistical significance (42% vs. 29%, p=0.07). After multivariate adjustment, age remained strongly associated with cancer outcome (OR=1.05, 95% CI 1.02-1.10, p<0.01). Women with BMI≥30kg/m2 had three times the odds of cancer diagnosis compared to women with BMI<25kg/m2 (OR=3.33, 95% CI 1.18-4.34, p=0.02). Family history (OR=1.42, 95% CI 0.64-3.16, p=0.39) and HRT (OR=1.44, 95% CI 0.60-3.42, 0.41) were associated with elevated odds of cancer, but were not statistically significant. Conclusions: Age and BMI were predictors of cancer diagnosis among women with BIRADS 4 mammograms. Recruitment for this cohort is ongoing, and analysis is currently being expanded to include breast cancer associated SNPs, area-based and volumetric measures of breast density, and novel measures of breast complexity. Logistic Regression - Odds of In-situ or Invasive Breast Cancer in BIRADS 4 . Non-Cancer (N = 228) DCIS/Invasive Cancer (N = 41) p-value MultivariateOR Logistic Regression N(%) N(%) 95% CI p-value Age 40–49 78 (34) 9 (22) 0.001 1.05 (Age continuous) 1.02–1.10 0.005 50–59 97 (43) 10 (24) 60+ 53 (23) 22 (54) BMI <25 91 (40) 8 (20) 0.073 1.00 Reference 25–29.9 49 (22) 11 (27) 2.87 0.96–8.51 0.058 30+ 67 (29) 17 (42) 3.33 1.18–9.38 0.023 Race White 117(51) 19 (46) 0.793 1.00 Reference . African Amer. 91 (40) 18 (44) 1.10 0.49–2.48 0.817 Other 20 (9) 4 (10) 1.17 0.31–4.34 0.819 Previous Biopsy No 119(52) 22 (54) 1.000 1.00 Reference . Yes 109(48) 19 (46) 0.87 0.41–1.85 0.716 Age at Menarche > = 12 186(82) 34 (83) 1.000 1.00 Reference . <12 42 (18) 7 (17) 0.82 0.32–2.12 0.682 HRT Never 189(83) 30 (73) 0.188 1.00 Reference . Ever 39 (17) 11 (27) 1.44 0.60–3.42 0.411 Family History No 168(74) 28 (68) 0.453 1.00 Reference . Yes 60 (26) 13 (32) 1.42 0.64–3.16 0.386 Age First Birth None 49 (22) 10 (24) 0.919 1.43 0.54–3.79 0.466 <30 134(59) 23 (56) 1.00 Reference . > = 30 45 (20) 2 (20) 1.42 0.57–3.57 0.454 Breast Density Scattered Fibroglangular 110(48) 23 (56) 0.354 1.16(Extreme/Hetero vs. Fibro/Fat) 0.50–2.68 0.726 Entirely Fat 10 (4) 2 (5) . . Extremely Dense 8 (4) 0 (0) . . Heterogeneously Dense 99 (43) 15 (37) . . Citation Format: Anne Marie McCarthy, Susan M. Domchek, Despina Kontos, Emily F. Conant, Mirar Bristol Demeter, Marisa Jones, Leigh Boghossian, Erin McGuire, Katrina Armstrong. Age and body mass index are predictive of cancer diagnosis among women with BIRADS 4 mammograms. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 2528. doi:10.1158/1538-7445.AM2013-2528