Search

Background. Recent evidence indicates that a soluble fragment of the erbB-2 oncogene product may be released from cell surface and become detectable in the serum of patients with breast cancer. Methods. To study the diagnostic utility of this phenomenon, the authors measured serum erbB-2 levels with a quantitative enzyme-linked immunosorbent assay in 227 preoperative samples from women who underwent breast surgery and in 339 samples from 225 patients with breast cancer during follow-up. Results. Eleven (9%) of 114 preoperative samples from patients with a histologically verified breast cancer and 2 of 113 (1.8%) from patients with benign breast tumors had elevated (greater than 20 U/ml) serum erbB-2 antigen levels. Ten (91%) of the 11 carcinomas and one of the benign tumors from patients with elevated serum erbB-2 levels also showed overexpression of the erbB-2 protein in immunohistochemical analysis of tissue sections. Elevated preoperative serum erbB-2 levels were predominantly found in patients with large tumors, and those with axillary lymph node or distant metastases. Sixty-three of the 339 (19%) follow-up samples had elevated serum erbB-2 antigen levels. Approximately one-third (30.9%) of the samples taken during recurrent disease were serum erbB-2 positive, which is close to the overall overexpression rate of this oncogene. Elevated erbB-2 levels were more common in patients whose disease was not responsive to treatment. Patients with distant metastases had elevate erbB-2 levels more often (40%) than did those with locoregional recurrence (20%). Elevated erbB-2 levels predicted the appearance of metastases within the next 6 months in 10 of 27 (37%) patients. Conclusion. The study's results suggest that assay serum erbB-2 levels may be valuable in the follow-up and monitoring of patients with breast cancer whose primary tumors show erbB-2 overexpression by immunohistochemistry.

Human breast carcinomas (n = 232) were evaluated for estrogen and progesterone receptors (ER, PR) by immunohistochemical study and by cytosol steroid-binding assay (n = 185). The staining was scored (histoscore) by estimates of relative nuclear staining intensity and the percentage of positively stained carcinoma cells. Of the invasive ductal carcinomas 72% were ER-positive and 55% were PR-positive. The invasive lobular, intraductal, tubular, and mucinous carcinomas were the most frequent ER-positive tumor types, whereas comedo and medullary carcinomas only rarely contained ER. Progesterone receptor was most frequently present in intraductal, tubular, and mucinous carcinomas. Better differentiated tumors with lower histologic grade were significantly associated with high prevalence of immunohistochemically determined ER and PR (P less than 0.0001). Proliferative cell fraction, determined by DNA flow cytometric study (n = 63), was inversely related to ER (P = 0.03) and PR (P = 0.05) status. Aneuploidy was independent of ER or PR content.

Growth kinetics of 102 breast carcinomas were studied by immunohistochemical analysis with the monoclonal antibody Ki-67, which reacts with a nuclear antigen in proliferating cells. The results were correlated with ploidy and S-phase fraction (SPF) analyzed by DNA flow cytometric study and with mitotic count analyzed by light microscopic study. The proportion of Ki-67-positive cells (Ki-67 score) in breast carcinomas varied from 0.6% to 80% (median, 6.3%). Ki-67 scores were significantly higher in the DNA aneuploid than in the DNA diploid tumors. Ki-67 scores correlated significantly with tumor SPF in DNA aneuploid tumors. In DNA diploid tumors SPF showed no correlation with Ki-67 score. High Ki-67 scores were associated with high mitotic counts (P less than 0.0001) and histologic grade (P less than 0.0001). Nuclear pleomorphism, tubule formation, or lymph node status were not correlated with Ki-67 score. In conclusion, Ki-67 immunostaining correlates with other measures of cell proliferation (SPF, mitotic count) supporting its clinical significance.

To optimize the prognostic value of DNA flow cytometry in breast cancer the authors calculated several parameters from the DNA histogram, including the DNA index, the size and number of aneuploid peaks as well as S-phase and G2/M-phase cell cycle fractions. Of these, DNA index and S-phase fraction (SPF) proved to be the most valuable prognostic indices. DNA aneuploidy was associated with a three-fold risk of death as compared to DNA diploidy (P less than 0.0001). The highest risk of death was associated with hypertetraploid (greater than 2.20) DNA index, whereas a tetraploid DNA index (1.80-2.20) was associated with a relatively low risk. The SPF had significant additional prognostic value in both DNA diploid (P = 0.0002) and DNA aneuploid (P = 0.02) tumors. By combining DNA index and SPF the authors defined three types of DNA histograms, which were associated with favorable, intermediate, and poor prognosis of the patients. DNA diploidy together with low (less than 7%) SPF (type I DNA histogram) was associated with very favorable prognosis, whereas DNA aneuploidy with high DNA index (greater than 2.20) or high (greater than 12%) SPF (type III DNA histogram) was related to the worst prognosis with approximately eight-fold relative risk of death. In a Cox multivariate regression analysis the type of DNA histogram was an independent and most powerful prognostic indicator in breast cancer. The other independent factors in the Cox analysis were primary tumor size, nodal status, and progesterone receptor status.