Interval Breast Cancers Are Associated with More Aggressive Pathologic Characteristics Compared to Screen-Detected Cancers: A Nested Case Control Study from a Canadian Breast Screening Program
D Rayson, JI Payne, PJ Barnes, RF MacIntosh, M Abdolell, T Foley, AM Burns, T Younis, JS Caines. Queen Elizabeth II Health Sciences Centre and Dalhousie University, Halifax, NS, Canada; Cancer Care Nova Scotia, Halifax, NS, Canada; Dalhousie University, Halifax, NS, Canada; Nova Scotia Breast Screening Program, Halifax, NS, Canada
Background: The evaluation of clinical and pathological differences between screen-detected and interval breast cancers has been limited by the failure to control for factors known to influence breast cancer biology such as age at diagnosis as well as differences in recommended screening intervals between age groups and based on family history.
Design: A case-control study nested within the participants of the population-based Nova Scotia Breast Screening Program diagnosed between ages 40-69 in the period 1991-2004 was performed. Interval cases were selected as having developed after a negative screen and prior to the recommended next screen and were validated by blinded review of the pre-diagnosis screening mammogram by 3 radiologists, 2 of whom had to agree that the screening exam was negative. Screen-detected cases were matched to intervals on a 2:1 basis by 5-year age group and recommended screening interval (40-49 yo, annual; 50-69 yo positive family history, annual; 50-69 yo negative family history, biennial). Pathologic variables were abstracted and slides reviewed for cases with incomplete reports. Logistic regression controlling for matching factors was used to assess differences in pathologic characteristics.
Results: 243 true interval invasive cancers were identified. Compared to 485 age matched screen-detected cancers, interval cancers were significantly more likely to be high grade (grade 3 vs. 1) (OR=4.85; 95% CI 3.00, 7.82), have lymphatic-vascular invasion (OR=4.66; 95% CI 2.97, 7.29), positive lymph nodes (OR=2.30; 95% CI 1.64, 3.21), and have a triple negative phenotype (OR=1.99; 95% CI 1.03, 3.85).
Conclusions: Controlling for age and screening interval, true interval invasive breast cancers have significantly more unfavorable prognostic variables compared to screen-detected cancers suggesting a true difference in tumor biology.
Tuesday, March 10, 2009 1:00 PM
Poster Session IV # 6, Tuesday Afternoon