Prognostic and treatment predictive factors for radio- and chemotherapy resistance in breast cancer patients- a step towards personlized medicine

Detta är en avhandling från Emma Niméus-Malmström, Dept of Oncology

Sammanfattning: Breast cancer is the most common cancer form among women in the Western world. Although treatment has improved during the last decades, there is still a significant proportion of the patients who are not cured. To further improve clinical outcome we need new treatment strategies, new prognostic markers, and new treatment predictive factors for personalized medicine.In this dissertation I have focused on local relapse and distant recurrences, where the former is a tumor recurring in the same breast and the latter is distant spread in the body. I have focused on high through-put techniques, but have also evaluated one single factor, using immunohistochemistry. The most promising result is the gene expression profile for “radioresistance” found in a patient cohort consisting of 100 lymph node negative patients operated with breast conservation surgery and either postoperative radiotherapy or not. The samples were analyzed with oligonucleotide array. A gene expression profile was found that clearly separated patients who developed local recurrences despite radiotherapy (“radioresistance”) from patients without local recurrences (either with or without radiotherapy). The clinical consequence, if these results can be confirmed, would be that patients with a“radioresistant” gene profile should be offered mastectomy instead of breast conservation surgery and radiotherapy. In another patient cohort consisting of 85 node positive breast cancer patients treated with CMF (cyclophosphamide, methotrexate, and 5-fluorouracil) we found a gene expression profile, using cDNA microarray, capable of distinguishing patients who developed distant recurrences from non-recurring patients. This profile was compared to a previously published gene list and to drug-associated genes from the literature. Our results were slightly better. However, our gene profile was not able to exceed the performance of conventional clinical markers. From the same patient cohort, we developed a protocol for protein extraction from the same samples used for RNA. We analyzed the protein expression pattern using 2-DE (two-dimensional electrophoresis) and found several differentially expressed proteins, both when comparing distant recurrences to no recurrences and estrogen receptor positive to estrogen receptor negative tumors. Similarities of regulated genes and proteins were also found when comparing the two studies. Finally, we investigated the prognostic importance of a proliferation marker, cyclin B1,in a case-control study. There were 190 lymph node negative breast cancer patients with no chemotherapy who died from breast cancer and 190 corresponding controls who were alive at the corresponding case’s time of death. Cyclin B1 was an independent prognostic proliferation marker and had a high reproducibility. The marker may be useful instead of histological grade, or as a complement, to identify patients in need of adjuvant chemotherapy. In conclusion, breast cancer is a heterogeneous disease and should be subdivided even more than it is today into different risk groups with the aid of new markers. We used different high through-put techniques to analyze hundreds to thousands of gene expressions and proteins. Our aim was to find new prognostic and predictive gene expressions and protein profiles that may improve individual treatment schemes and help provide personalized medicine. However, so far it is too early to conclude that the use of single markers can be eliminated, because we also found significant prognostic value of the proliferation marker cyclin B1.