Prognostic markers in breast cancer associated with cell cycle control, proliferation and angiogenesis

This thesis focuses on potential prognostic molecular breast cancer markers which are related to angiogenesis, proliferation or cell cycle control. A common denominator for all of the investigated markers in the thesis is either hypothesized or previously reported association with the tumor suppressor gene TP53. All patients in the present studies derived from a population based cohort of 311 Swedish breast cancer patients with known TP53 gene status. In study nr I expression of the angiogenic cytokine Vascular Endothelial Growth Factor (VEGF) was compared with available TP53 data from cDNA based gene sequencing, a TP53 luminometric immunoassay and TP53 immunohistochemistry (IHC) in 224 patients. High VEGF expression was associated with mutated TP53. The highest VEGF levels were found in breast cancers with insertions, deletions or stop codon point TP53 mutations. These findings support the hypothesis that wildtype TP53 is involved in the regulation of VEGF. Significantly worse outcome for was observed for tamoxifen treated estrogen receptor (ER) positive patients with high VEGF. In study nr II the cell cycle regulatory protein cyclin E was investigated by IHC in breast cancers from 270 patients. TP53 mutations were much more common in patients with high cyclin E expression compared with low cyclin E expression. In TP53 mutated breast cancers, high cyclin E content was associated with insertions, deletions and nonsense point mutations in the TP53 gene, whereas low cyclin E was linked to TP53 missense point mutations. Strong associations were found between a high cyclin E content and presence of several other tumor markers including aneuploidy. High cyclin E content was associated with poor survival. The results implicate that overexpression of cyclin E is associated with an aggressive tumor phenotype and specific types of TP53 mutations. In study nr III the possibility of improving the indirect TP53 mutation screening with TP53 IHC, by co-analysis of two proteins associated with wild-type TP53, p21(waf1/cip1) and MDM-2 (murine double minute-2), was investigated. The expression of p21and MDM-2 was determined by IHC on 276 and 257 primary breast cancers, respectively, and compared with the TP53 mutation status. Expression of p21 and MDM-2 was strongly positively correlated, but no statistically significant correlations between TP53 gene status and expression of p21 or MDM-2 was observed. According to these results is determination of the IHC expression of p21 or MDM-2 not useful for identification of breast cancers with mutated TP53. In study nr IV the proliferation marker Ki67 (MIB-1) was analyzed with IHC in 305 patients, in order to compare and validate Ki67 in relation to S-phase fraction (SPF), other breast cancer markers and patient outcome. A graticula based method for reading the Ki67 sections was also developed. High Ki67 correlated with several other breast cancer prognostic markers, although only moderately to SPF. Using a grid graticule for Ki67 scoring gave comparable results to individual cell counting. Patients with high Ki67 had shorter survival. In summary, VEGF content was associated with TP53 status and gave predictive information for adjuvant tamoxifen therapy. Indirect p53 mutation screening with IHC is not improved by co-analyses of p21 and MDM-2. Among the tumor markers investigated in this thesis cyclin E and KI-67 had the highest prognostic value for patient outcome