Elastosis in human breast cancer. Correlation with sex steroid receptors and comparison with clinical outcome.

Elastosis in human breast cancer has been compared with various prognostic indicators, but few studies have compared elastosis directly with prognosis. We examined 100 primary breast cancers for estrogen receptor (ER) and progesterone receptor (PR) content, tumor grade, and extent of elastosis and compared these data with clinical follow-up in 71 patients. The ER content was greater than 10 femtomoles (fmole) per milligram of protein in 68% of tumors, while PR content was greater than 10 fmole/mg of protein in 48%. Histologic study showed that 22% of the tumors were grade 1, 28% were grade 2, and 50% were grade 3. Grade 2 or grade 3 elastosis was seen in 36% of the tumors and grade 0 or grade 1 elastosis was seen in 64%. There was a negative correlation between tumor grade and elastosis grade and a positive correlation between ER content and elastosis grade. Elastosis grade and PR content did not correlate. A significant positive correlation existed between lymph node positivity and elastosis. No significant direct relationship between elastosis and tumor recurrence could be demonstrated once lymph node status was accounted for. Tumor elastosis correlates with histologic grade, receptor content, and lymph node involvement but does not appear to be a direct marker of prognosis.

Biochemical correlates of morphologic differentiation in human breast cancer.

In 115 breast carcinoma tissues, histologica grade and cell cytosol concentrations of estrogen receptor (ER) and progesterone receptor (PR) and two breast cyst fluid proteins (gross cystic disease fluid protein [GCDFP-15] and nonreceptor progesterone-binding protein [PBP]) were deterMined. Higher levels (expressed as femtomoles per milligram of protein) of ER (128 +/- 28 versus 11 +/- 1, P less than 0.001) and PR (82 +/- 16 versus 3 +/- 1, P less than 0.001) were found in grade 1 (well-differentiated) carcinomas as compared with grade 3 (poorly differentiated) carcinomas. Similarly, higher concentrations (expressed as nanograms per milligram of cytosol protein) of GCDFP-15 (2110 +/- 840 versus 210 +/- 40, p less than 0.001) and PBP (4920 +/- 1200 versus 370 +/- 60, P less than 0.001) were found in grade 1 as compared with grade 3 carcinomas. Tumor cytosols that contained low levels of both cyst proteins (less than 225 ng/mg GCDFP-15 and less than 750 ng/mg PBP) had a high incidence of grade 3 (35 of 46, 78%) or grade 2 (15 of 46, 33%) histologic findings and had a high incidence of receptor-negative specimens (27 of 52, 52%). Based on these cutoff levels, grade 2 lesions were subdivided into a "high" cyst protein group, which had ER and PR levels similar to grade 1 tumors (93.1 +/- 26.7 for ER and 84.7 +/- 32.4 for PR, P greater than 0.3), and a "low" group, which had receptor values similar to grade 3 carcinomas (14.1 +/- 5.3 for ER and 9.1 +/- 5.2 for PR, P less than 0.3). Although the mean cytosol content of carcinoembryonic antigen (CEA) was significantly higher in malignant tissues (125 +/- 27 ng/mg cytosol protein) than in benign tissues (4.8 +/- 1 ng/mg cytosol protein), the CEA content was not significantly different between grades 1 and 3 tumors.