Format

Send to:

Choose Destination
See comment in PubMed Commons below
Cancer Epidemiol Biomarkers Prev. 2005 Feb;14(2):324-8.

Nipple aspirate fluid cytology and the Gail model for breast cancer risk assessment in a screening population.

Author information

  • 1Division of General Internal Medicine, Department of Medicine, University of California, San Francisco, 1701 Divisadero Street, Suite 554, San Francisco, CA 94143-1732, USA. jtice@medicine.ucsf.edu

Abstract

BACKGROUND:

Recent guidelines suggest that chemoprevention with tamoxifen may be appropriate for women who have a 5-year risk of breast cancer greater than 1.66% calculated using the Gail model.

OBJECTIVES:

To determine whether nipple aspirate fluid (NAF) cytology combined with the Gail model provides breast cancer risk assessment that is superior to either method alone.

METHODS:

Prospective observational cohort of 6,904 asymptomatic women. Breast cancer cases were identified through follow-up with the women and linkage to cancer registries. We used proportional hazards modeling to recalculate the coefficients for the predictor variables used in the Gail model. NAF cytology was added to create a second model. The two models were compared using the concordance statistic (c-statistic).

RESULTS:

During 14.6 years of follow-up, 400 women were diagnosed with breast cancer. There were 940 (14%) women with hyperplasia and 109 (1.6%) women with atypical hyperplasia found in NAF. Adding NAF cytology results to the Gail model significantly improved the model fit (P < 0.0001). The c-statistic for the Gail model was 0.62, indicating only modest discriminatory accuracy. Adding NAF cytology to the model increased the c-statistic to 0.64. NAF cytology results had the largest effect on discriminatory accuracy among women in the upper third of Gail model risk. The relative incidence for the highest quintile of risk score compared with the lowest quintile was 7.2 for the Gail model and 8.0 for the model including NAF cytology.

CONCLUSION:

NAF cytology has the potential to improve prediction models of breast cancer incidence, particularly for high-risk women.

PMID:
15734953
[PubMed - indexed for MEDLINE]
Free full text
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for HighWire
    Loading ...
    Write to the Help Desk