Background. Gene arrays have demonstrated different outcomes for breast cancer subtypes highlighting the heterogeneity of breast cancer. The limited availability of gene expression analysis and financial issues have contributed to the development of surrogate markers to identify corresponding subgroups using IHC. 2011 ESMO and St Gallen guidelines suggest the use of an IHC panel consisting of ER, PgR, HER2 and Ki67 cut-off value ≥14 % (Ki6714%) for discriminating luminal A from B.The cut-off value suggested from 2013 St Gallens guidelines was ≥20% (Ki6720%). We wanted to evaluate if the different cut-off values for Ki67 or cyclins A/B1 could reliably separate luminal A from B.
Patients. In a case-control study, we defined 190 women who died from breast cancer as cases and 190 women alive at the time for the corresponding case's death as controls. Inclusion criteria were tumor size ≤50 mm, no lymph node metastases and no adjuvant chemotherapy. Immunohistochemical evaluation of ER, PgR, HER2, Ki 67, cyclin A and cyclin B1 were utilized for subgrouping.
Results. Conditional logistic regression analysis was used to estimate odds ratios (OR) for breast cancer death. Ki6714% did not detect differences in outcome between luminal A and B breast cancer (OR 1.4, 95% CI 0.8-2.26 p-value 0.24). Corresponding values for cyclin A was OR 3.6 (95% CI 1.8-7.0 p-value 0.00), cyclin B1 2.2 (95% CI 1.1-4.5 p-value 0.04) and Ki6720% 2.0 (95% CI 1.1-3.9 p-value 0.04) using luminal A as reference.
Conclusion. In our study, Ki6714% failed to detect any difference in outcome between luminal A and B. In contrast, using cyclin A as a proliferation marker luminal B was found to have an almost 3.5 -fold higher risk of dying from breast cancer. Cyclin B1 and Ki6720%, could also separate luminal A from B but cyclin A separated more effectively these subtypes . We conclude that cyclin A distinctly separates luminal A from B in node negative breast cancer.