Srisairam Achuthan, Ph.D.
Senior Scientific Programmer, Research Informatics Division, Department of Information Sciences, City of Hope, CA, USA
Applying Machine learning for Automated Identification of Patient Cohorts
Srisairam Achuthan, Mike Chang, Ajay Shah, Joyce Niland
Patient cohorts for a clinical study are typically identified based on specific selection criteria. In most cases considerable time and effort are spent in finding the most relevant criteria that could potentially lead to a successful study. For complex diseases, this process can be more difficult and error prone since relevant features may not be easily identifiable. Additionally, the information captured in clinical notes is in non-coded text format. Our goal is to discover patterns within the coded and non-coded fields and thereby reveal complex relationships between clinical characteristics across different patients that would be difficult to accomplish manually. Towards this, we have applied machine learning techniques such as artificial neural networks and decision trees to determine patients sharing similar characteristics from available medical records. For this proof of concept study, we used coded and non-coded (i.e., clinical notes) patient data from a clinical database. Coded clinical information such as diagnoses, labs, medications and demographics recorded within the database were pooled together with non-coded information from clinical notes including, smoking status, life style (active / inactive) status derived from clinical notes. The non-coded textual information was identified and interpreted using a Natural Language Processing (NLP) tool I2E from Linguamatics.
Gokul Das, Ph.D.
Co-Director, Breast Disease Site Research Group, Roswell Park Cancer Institute, Buffalo, NY
Probing Estrogen Receptor−Tumor Suppressor p53 Interaction in Cancer: From Basic Research to Clinical Trial
Tumor suppressor p53 and estrogen receptor have opposite roles in the onset and progression of breast cancer. p53 responds to a variety of cellular of stresses by restricting the proliferation and survival of abnormal cells. Estrogen receptor plays an important role in normal mammary gland development and the preservation of adult mammary gland function; however, when deregulated it becomes abnormally pro-proliferative and greatly contributes to breast tumorigenesis. The biological actions of estrogens are mediated by two genetically distinct estrogen receptors (ERs): ER alpha and ER beta. In addition to its expression in several ER alpha-positive breast cancers and normal mammary cells, ER beta is usually present in ER alpha-negative cancers including triple-negative breast cancer. In spite of genetically being wild type, why p53 is functionally debilitated in breast cancer has remained unclear. Our recent finding that ER alpha binds directly to p53 and inhibits its function has provided a novel mechanism for inactivating genetically wild type p53 in human cancer. Using a combination of proliferation and apoptosis assays, RNAi technology, quantitative chromatin immunoprecipitation (qChIP), and quantitative real-time PCR (qRT-PCR), in situ proximity ligation assay (PLA), and protein expression analysis in patient tissue micro array (TMA), we have demonstrated binding of ER alpha to p53 and have delineated the domains on both the proteins necessary for the interaction. Importantly, ionizing radiation inhibits the ER-p53 interaction in vivo both in human cancer cells and human breast tumor xenografts in mice. In addition, antiestrogenstamoxifen and faslodex/fulvestrant (ICI 182780) disrupt the ER-p53 interaction and counteract the repressive effect of ER alpha on p53, whereas 17β-estradiol (E2) enhances the interaction. Intriguingly, E2 has diametrically opposite effects on corepressor recruitment to a p53-target gene promoter versus a prototypic ERE-containing promoter. Thus, we have uncovered a novel mechanism by which estrogen could be providing a strong proliferative advantage to cells by dual mechanisms: enhancing expression of ERE-containing pro-proliferative genes while at the same time inhibiting transcription of p53-dependent anti-proliferative genes. Consistently, ER alpha enhances cell cycle progression and inhibits apoptosis of breast cancer cells. Correlating with these observations, our retrospective clinical study shows that presence of wild type p53 in ER-positive breast tumors is associated with better response to tamoxifen therapy. These data suggest ER alpha-p53 interaction could be one of the mechanisms underlying resistance to tamoxifen therapy, a major clinical challenge encountered in breast cancer patients. We have launched a prospective clinical trial to analyze ER-p53 interaction in breast cancer patient tumors at Roswell Park Cancer Institute. Our more recent finding that ER beta has opposite functions depending on the mutational status of p53 in breast cancer cells is significant in understanding the hard-to-treat triple-negative breast cancer and in developing novel therapeutic strategies against it. Our integrated approach to analyze ER-p53 interaction at the basic, translational, and clinical research levels has major implications in the diagnosis, prognosis, and treatment of breast cancer.