Breast complaints and cancer: age stratified predictors of risk from a prospective database.

Abstract

Abstract Abstract #4086 Surgical clinics are presented with a variety of breast complaints. This study looks at major reasons for referral and correlates them to the likelihood of a benign or malignant disease.
 Data comes from a prospective database of new referrals for surgical assessment. Male patients and presentations of recurrent cancer were excluded. Reasons for presentation were sorted into 7 categories defined as breast asymmetry, palpable mass, non-inflammatory skin changes, inflammatory changes, nipple discharge, nipple changes, pain or abnormal imaging. The dominant complaint was applied. Some patients are included twice if they had another problem in the opposite breast, or a second presentation of a new problem. Patient demographics were recorded and all patients were followed to a benign or malignant diagnosis.
 Chi-square testing, odds ratios, and confidence intervals were used for the categorical data. Fisher's exact test was employed for categories with a low cell count.
 Three hundred and ninety of 1050 patients were found to have breast cancer. The most common presentations of malignancy were a palpable mass and abnormal imaging. Less common presentations that also predicted cancer were persistent inflammation, skin changes, and nipple changes. Breast pain (p=0.00001), breast asymmetry (p=0.00001), and nipple discharge (p=0.00001) were significantly correlated with benign disease. The most common diagnosis varied significantly with age. Fibroadenomas were most common in young women, while cysts were frequent in the peri-menopausal group (p=0.0001). A new mass in a woman over 65 years old was malignant 83% of the time (p=0.000001).
 Reason for referral can be significantly correlated with a benign or malignant diagnosis. Breast asymmetry, pain, and nipple discharge were associated with a benign diagnosis, while a new mass in an older woman was more likely to be malignant than benign. This information can be used to help stratify referrals into high and low risk categories, and identify patients for expedited assessment. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 4086.