- Endometrial Cancer (EC): We expanded our EC dataset with clinical and molecular outcome data to include gene expression and proteomic data on over 175 cases obtained from Duke Univ, with miRNA array, SNiPchip6.0, and mutational analysis on over 60 cases. We are identifying biomarkers associated with recurrence free survival and are validating these findings in an independent sample set from the Gynecologic Oncology Group (GOG). The biologic role of selected differentially expressed molecules is under investigation in collaboration with Dr. Risinger from Michigan State Univ and Dr. Annunziata, MOB/CCR. - Racial Disparities in EC: We have continued studies on racial disparities in EC outcome between African American (AA) and Caucasian (C) women. Analysis of GOG Phase III EC clinical trials over the past decade demonstrated worse outcome for AA, even when treated in prescribed clinical trials. These results prompted the hypothesis that there are biological etiologies that contribute to this racial disparity. Differentially expressed molecules were examined by TMA containing ECs from AA and C women, matched by stage and histology. A pending collaboration will analyze associations between 128 African specific mtDNA polymorphisms/haplogroups with clinical and pathological characteristics of EC within and between 250C and 250AA women. This will provide a method of objectively defining AA status based on known mtDNA haplotypes. Finally, in collaboration with UPCI investigators using MS methods to quantitatively evaluate protein expression, the phospho-serine phosphatase PSPH was found to be over-expressed in AA EC. It is an enzyme that catalyzes the last step in the biosynthesis of serine from carbohydrates. A 72 amino acid transcript of a homolog of PSPH (PSPHL) was significantly upregulated in AA EC patients;bacterial expression and purification yielded a 7.8kDa protein, confirmed by MS. - We have examined NF-kB pathway alterations in AA and C EC clinical samples using tissue microarray and cell lines with Dr. Annunziata, TGS, MOB/CCR. Preliminary results suggest that NF-kB is more active in serous EC in AA clinical samples. Inhibition of the pathway in cell lines identified a sample subset that is dependent on IKKb signaling. Further evaluations are underway to identify and refine a NF-kB expression signature in EC cell lines and to extend these findings to clinical samples by gene expression and TMA. - Obesity and EC: The relationship between obesity and EC is more significant than for any other cancer type. Obese women have a 3-5fold increased risk of developing EC and up to a 6fold increased risk of cancer-related mortality. Our data yielded global gene expression profiles that significantly differ between ECs from lean v. obese women. Several of these molecules are associated with inflammatory processes suggesting that anovulation may promote an inflammatory microenvironment conducive to carcinogenesis. Our subsequent work using EC specimens indicated the inflammatory profile is absent when only the carcinomatous component is selectively investigated via laser capture microdissection. We previously identified a biomarker panel associated with EC that was validated in subsequent sample sets to have over 95% sensitivity and specificity. Many of the biomarkers found to be significantly elevated were inflammatory cytokines. Our consortia developed preliminary data demonstrating differences in serum inflammatory cytokines in lean v obese EC patients. This lead us to hypothesize inflammatory biomarkers associated with EC might vary by body habitus among normal women. Recent data collected from patients participating in a prospective weight loss program has confirmed a decrease in inflammatory cytokines among women at various stages of weight loss. 15 estrogen metabolites are under study in a serum sample set from M.D. Anderson EC SPORE. These preliminary data will drive metabolomic analysis of estrogen in tumor/urine/serum/blood correlated with gene expression analysis using tumors in the GOG-210 EC repository. Previous work in collaboration with the CDC has shown that higher higher progestin-potency OCs may be more protective than lower progestin potency OCs among women with a larger body habitus. We have further shown in macaque models in collaboration with Dr. Cline at Wake Forest that these protective effects may be in part due to progestin-mediated induction of apoptosis and TGF-mediated pathways. Recent data from our consortium have shown that a combination of vitamin D and progestin offer a greater protective effect than progestin alone in prevention of endometrial hyperplasia in PTEN+/- mice. A more definitive trial using 240 PTEN+/- mice is underway and we will investigate the effects of placebo, progestin, vitamin D, and combination therapy on phenotype as well as molecular expression. A xenograft study of 64 mice exposed to these chemopreventive therapies has been completed and data are being evaluated in conjunction with Dr. Rodriquez from Evanston Northshore. Recently, we have completed in vitro experiments aimed at elucidating the changes in global gene expression associated with exposure to placebo, progestin, vitamin D, or vitamin D+ progestin. Differentially expressed biomarkers are being validated. Current studies are underway evaluating the gene and protein expression profiles associated the effects of combining progestin and vitamin D on the endometrial lining. - EC Metastasis models: Previous work by our group using the Medicare data base has revealed that optimal referral to a gynecologic oncologist is associated with improved survival for patients with EC and yet only 24% of patients with newly diagnosed disease are referred to a gynecologic oncologist. A decision model constructed by our group has demonstrated the cost effectiveness of pre-operative screening for metastasis among patients with EC. Given the current low rates of full surgical staging, a diagnostic test to detect nodal metastasis for EC has the potential to be clinical useful in more accurately promoting referral to a specialist for treatment. Repeat analysis of serum samples from EC patients has revealed that metastasis can be predicted but only in EC cancers. We are continuing to validate our findings using a sample set from the M.D. Anderson EC SPORE. In collaboration with the GOG, we are finalizing development of an optimal molecular expression prediction model using a set of 140 endometrioid EC from GOG-210 that have been analyzed using gene expression arrays and LC/MS-MS analysis of tissue lysates prepared using laser capture microscopy. Our validated gene expression analysis profile based on transcript expression is improved over traditional pathologic factors. Proteomic based models using tissue lysates offer a more improved level of accuracy in identifying EC patients with metastasis. We are evaluating the accuracy using paraffin tissue samples prior to testing endometrial biopsy from patients with stage I and advanced stage III/IV disease. We are also preparing a set of 190 cases of recurrent and non-recurrent EC from GOG-210. In the setting of adequate surgical staging, stage I patients with EC who recur were most likely cases of unrecognized advanced stage disease. We will develop a model to molecularly examine the hysterectomy specimen to more accurately identify patients with micrometastasis who are likely to recur. Finally, our group is comparing the primary and metastatic lesions from approximately 80 stage III and IV EC in GOG 210 to identify metastasis genes associated with EC that can serve as targets for development of biologic therapeutics.
Showing the most recent 10 out of 13 publications
