MODULATION OF BIOMARKERS IN MINIMAL BREAST CARCINOMA. A MODEL FOR HUMAN BREAST CANCER PROGRESSION

BACKGROUND: The widespread use of diagnostic breast imaging has yielded an increase in the detection of in situ, microinvasive, and small invasive carcinomas and has provided opportunities to study the earliest stages of breast carcinoma development. The authors of this report analyzed the pathobiologic features of 577 minimal breast carcinomas (MBCs), including in situ carcinomas and invasive carcinomas < or =1 cm, according to the definition given by Hartmann in Cancer (1984;53:681-4). METHODS: Estrogen and progesterone receptors (ER and PR), proliferation index (PI), and p53 and neu expression were studied by immunohistochemical technique and measured by quantitative image analysis in 99 pure in situ carcinomas (ISCp); in 105 mixed invasive/in situ carcinomas, with a separate analysis of in situ (ISCm) and invasive (ICm) components; and in 373 invasive carcinomas < or =1 cm (IC). Follow-up data were available for 164 invasive carcinomas. RESULTS: A progressive increase in the levels of hormone steroid receptors, from the lowest in ISCm to the highest in IC, was observed (ER, P< 0.001; PR, P=0.005). Levels of PI and p53 expression were higher in ISCm than in the other categories (PI, P=0.007; p53, P=0.046). Overexpression of neu was greater in ICm than in IC (P=0.013). Younger women (< or =40 years) with invasive carcinoma had worse biologic profiles, with lower ER (P < 0.001) and higher PI (P=0.021), neu (P=0.008), and p53 (P=0.040). It was demonstrated clinically that PI and neu were the biologic markers with the highest predictive prognostic values in univariate analysis (PI for recurrence, P < 0.015; neu for recurrence and overall survival, P < 0.001 and P < 0.007, respectively) and in multivariate analysis (neu for recurrence and overall survival, P < 0.007 and P < 0.017, respectively). CONCLUSIONS: Biologic phenotypes of MBC can be interpreted as reflecting a dimension of neoplastic progression capacity that is independent of tumor size. This study suggests that biologic markers can be integrated with traditional pathologic indicators for accurate staging of patients