Ethnic variation in kallikrein expression in nipple aspirate fluid

✍ Scribed by Edward R. Sauter; Tamara Welch; Angeliki Magklara; Gary Klein; Eleftherios P. Diamandis

• Publisher: John Wiley and Sons
• Year: 2002
• Tongue: French
• Weight: 293 KB
• Volume: 100
• Category: Article
• ISSN: 0020-7136
• DOI: 10.1002/ijc.10556

✦ Synopsis

Socioeconomic factors cannot entirely explain why black women have an earlier age of breast cancer onset and higher mortality rates, stage for stage, than whites. We and others have shown that prostate-specific antigen [PSA, also known as human kallikrein (hK) 3] is a marker of breast as well as prostate cancer, that hK2 and hK3 are highly homologous at the DNA and protein level and that the level of progesterone, which appears to upregulate hK3, is influenced by ethnicity. We hypothesized that nipple aspiration fluid (NAF) hK2 and hK3 levels are (i) lower in black than white women; (ii) independently associated with breast cancer; (iii) influenced by menopausal status; and (iv) in combination are more informative about whether a woman has breast cancer than either marker alone. NAF was assayed for hK2 and hK3, and the results were stratified by ethnicity, presence or absence of cancer and menopausal status. Statistical analysis was then performed. When stratified by ethnicity, hK2 (p = 0.003) and hK3 (p = 0.027) levels in blacks were lower than in whites. hK2 was lower in premenopausal black than in white subjects, regardless of cancer status. Overall, hK2, hK3 and the ratio hK2/hK3 were lower in subjects with breast cancer than in normal subjects. hK3 was lower in postmenopausal women with breast cancer, regardless of ethnicity. hK2 and hK3 levels were higher in pre- than in postmenopausal whites. Using logistic regression and considering hK2, hK3, hK2/hK3 and ethnicity, hK3 was significantly associated with breast cancer in both pre- (p < 0.001) and postmenopausal women (p = 0.023). In conclusion, whereas hK2, hK3, hK2/hK3 and ethnicity are each significantly associated with breast cancer bivariately, after entering the strongest predictor, hK3, into a logistic regression model, no other variable accounted for additional variation, although this observation is preliminary due to the limited number of black subjects in the study.