A Pilot Study of Celecoxib in Patients With Grade 2 or 3 Uterine Cancers

Endometrial cancer is the most common gynecologic cancer in the United States. The number of deaths from endometrial cancer has risen 128% since 1987. In 2001, an estimated 38,300 women will develop endometrial cancer (ACS Facts and Figures) and an estimated 6,600 women will die from endometrial cancer. Preinvasive and well-differentiated endometrial cancers are hormonally driven and often cured with surgery alone. Higher-grade tumors are usually not hormonally driven and proliferate via unknown mechanisms. These tumors are largely responsible for the rising death rate. Responses to toxic treatment protocols for recurrent endometrial cancer are dismal. Unfortunately, these post-menopausal women also often have comorbidities, which limit their eligibility for current chemotherapy and radiotherapy treatments.

Expression of COX-II has been identified in many human cancers including: colon cancer, gastric cancer, esophogeal cancer, bladder cancer, head and neck cancer, liver cancer, pancreatic cancer, prostate cancer and breast cancer. COX-II expression is also strongly expressed in the primary tumor and metastasic site in human cervical cancer. COX-II may influence cell cycle control by upregulating the proliferative capacity of neoplastic endometrial cells. Furthermore, COX-II inhibitors inhibit tumor proliferation even in cells that do not express COX. This suggests an alternative mechanism of action not yet defined that may play a role in inhibiting the growth of cancer tissue.

The enhanced expression of COX-II has led investigators to use COX-II inhibitors in the prevention and/or treatment of colon and prostate cancers both in vivo and in vitro. Celecoxib is now FDA approved for chemoprevention of colon cancer in familial adenomatous polyposis patients. If it can be shown that COX-II is downregulated by COX-II inhibitors in endometrial cancer, they may offer similar chemopreventative or chemotherapeutic potentials that have already been proven in colon cancer.

COX-II enzyme activity may not always correlate with end organ gene expression. Multiple genes have been implicated in apoptotic pathways and are affected by COX-II inhibitors. NS-398, a selective COX-II inhibitor causes elevations in APC expression and downregulation of c-myc. Prostate apoptosis response 4 (Par-4) levels are increased in cells treated with COX inhibitors. PTEN and hMLH1 are genes which are implicated in malignant transformation of endometrial tissue. 5-Lipooxygenase (5-LOX) is often correlated with COX-I and COX-II. Thus, in addition to COX-I and COX-II, these are good candidate genes to study the effects of COX-II inhibitors on uterine cancers.

Preliminary Data Since COX-II expression is seen in the endometrium and in other hormonally-dependent tumors, we have investigated the expression of COX-II in endometrial cancer. Our preliminary studies on 41 fixed samples of benign and neoplastic endometrium revealed that COX-II was not expressed in benign endometrial tissue, stains minimally (~1% of tumor cells) in well-differentiated endometrial carcinomas, and stains most strongly in poorly-differentiated carcinomas (~12% of tumor cells, most staining strongly). COX-II is expressed in all poorly differentiated uterine cancers. Our study also demonstrated that COX-II was also strongly expressed in uterine papillary serous carcinomas (UPSC) as well as clear cell carcinomas of the uterus. These findings were confirmed by Ferrandina, et al. A small percent of our patients as well as the patients in the Ferrandina study have only 1+ or 1-5% staining. These 'low-expressers' only made up 1/13 (7.7%) of our patients.

COX-II expression in endometrial carcinoma has a slight inverse correlation with apoptosis (r=-0.534). However, COX-II expression in endometrial carcinoma correlated with lymphovascular invasion (r=0.69) and depth of invasion (r=0.68). There was no correlation between COX-II expression and ER (r=0.03) or PR (r=-0.02). The presence of a poorly-differentiated tumor may imply a hormonally-independent pathway resulting in de-differentiation. In summary, our preliminary data reveals that COX-II expression is high in grade 2 and 3 endometrioid-type endometrial cancers, as well as UPSC and clear cell subtypes and is correlated with known clinical prognostic factors.

Expression of COX-II has been identified in many types of human cancers. Uterine cancer is the most common gynecologic cancer in the US and there has been an increase in uterine cancer deaths over the past decade mainly due to the difficulty in treating recurrences in the more aggressive histologic types. The study co-investigators have also identified COX-II expression in grade 2 and 3 endometrioid-type, clear cell, and papillary serous types of uterine cancers. Upregulation of COX-II may control the cell cycle by regulating the proliferative capacity of neoplastic endometrial cells. This is a Phase II pre-post intervention comparison study in eligible patients looking at the effects of a COX-II inhibitor on uterine cancer. The patients whose endometrial biopsy shows grade 2 or 3 endometrioid-type, clear cell, and papillary serous types of uterine cancers will be put on a selective COX-II inhibitor, Celebrex (Celecoxib) until the day of their surgery. The expression of COX-II and p21 will be quantified after treatment with Celecoxib in eligible patients. This expression will be evaluated by performing immunohistochemical staining on the endometrial biopsy (pre-intervention) and the hysterectomy specimen (post-intervention). Apoptosis, evaluated by the terminal deoxynucleotidyl transferase (TdT)-mediated dUTP- digoxigenin nick end-labeling (TUNEL) assay, will also be evaluated and compared to COX-II expression in endometrial cancer in the two specimens, endometrial biopsy (pre-intervention) and uterus (post-intervention). In addition to IHC analysis and apoptosis, gene expression of COX-related genes in the post-intervention uterine specimens will. This gene expression will be compared to matched controls who were not treated with a COX-II inhibitor. COX-II expression will be correlated with established clinical prognostic factors including lymphovascular invasion, depth of myometrial invasion and lymph node involvement. We hypothesize that Celecoxib will downregulate the expression of COX-II in these tumor types as it does in other similar tumors. We also hypothesize that apoptosis, as measured with the TUNEL assay, will be increased in areas with less COX-II expression and should be inversely proportional to cellular p21 expression. Additionally, COX-II inhibitors affect apoptotic pathways even in cells that do not express COX-II. For low expressing cells, COX-II inhibitor activity may be better documented with apoptosis. We hypothesize COX-related gene expression will be altered thus suggesting an up- or down-regulation of these genes in the end-organ tissue. Documenting downregulation of COX-II enzyme and altered gene expression in endometrial carcinoma after treatment with Celecoxib may result in further prospective studies using selective COX-II inhibitors as effective, well-tolerated chemotherapeutic agents in these uterine cancers that are resistant to many current therapies.