Rapid and prognostically valid quantification of immunohistochemical reactions by immunohistometry of the most positive tumour focus. A prospective follow-up study on breast cancer using antibodies against MIB-1, PCNA, ER, and PR

Abstract

The prognostic significance of immunohistochemical markers for cell proliferation [MIB-1, proliferating cell nuclear antigen (PCNA)] and hormone receptor analysis [oestrogen receptor (ER), progesterone receptor (PR)] was tested by means of immunohistometry in a series of 103 breast cancer patients in comparison with the lymph node status N, the tumour size T, histomorphological grading, and the biochemical ER and PR status. Immunohistochemical reactions were performed on 2 microns sections from paraffin-embedded tissue, using an indirect peroxidase method. The proportion of immunostained tumour cell nuclei was determined using a TV-image analysis system. Measurements were performed using a 20 x objective on 40 viewing fields (1.94 mm2, MIB-1 and PCNA) or 20 viewing fields (0.97 mm2, ER and PR). The mean immunopositivity of all viewing fields and the value of the most immunopositive viewing field (MIB-1max, PCNAmax, PRmax, ERmax) were calculated. The mean values and the maximal values were highly correlated (r = 0.903, P < 0.001). After 1:2:1 quantilization, 84.2 per cent of the 412 single measurements revealed mean and maximal values in the same category (P < 0.0001). For each of the four immunohistochemical markers, the prognostic significance of the maximal values was higher than that of the mean values. The highest prognostic significance was found for MIC-1max (P = 0.0002), followed by PRmax (P = 0.0046), ERmax (P = 0.0154), and PCNAmax (P = 0.0161). From the results of a Cox model, a 'prognostic index (PI)' was developed, ranging from -1 to 8: PI = 2 x N + T + MIB-1max-PRmax. The four groups of patients with PI values of < 2, 2-3, 4-5, and > 5 revealed significantly different 7.5-year survival probabilities (P < 0.0001). The simplicity of the PI makes it a clinically useful, routinely applicable, and understandable parameter in the surgical pathology of breast carcinoma.