SAN DIEGO -- Treatment with the anti-inflammatory drug celecoxib (Celebrex) failed to induce regression of grade 3 cervical intraepithelial neoplasia (CIN3) any better than placebo did, a small randomized trial showed.
After 14 to 18 weeks of treatment, complete or partial regression of CIN3 occurred in 40% of celecoxib-treated patients and 34% of the placebo group. However, a small subgroup of patients with elevated vascular endothelial growth factor (VEGF) at baseline did substantially better with celecoxib, leaving the door open to future investigations, Janet Rader, MD, of the Medical College of Wisconsin in Milwaukee, reported here at the Society of Gynecologic Oncology meeting.
"Celecoxib for 14 to 18 weeks did not show a significant decrease in the severity of CIN3 compared to placebo," said Rader. "We observed decreased CIN3 in women with initial high serum VEGF levels. A decline in cervical COX [cyclo-oxygenase]-2 expression occurred from pretreatment to the end of treatment for the celecoxib group. A decrease in serum VEGF was seen in patients with high pretreatment cervical COX-2."
The enzymes COX-1 and COX-2 have a central role in arachidonic acid metabolism and resulting production of prostaglandins. The enzymes have counterbalancing effects on the inflammatory and thrombotic processes that arise from arachidonic acid metabolic pathway.
Investigators hypothesized treatment with a selective COX-2 inhibitor might have beneficial effects on CIN lesions, which arise in inflammatory environment driven by human papillomavirus infection. Rader reported findings from a randomized, placebo-controlled Gynecologic Oncology Group 0207 trial of celecoxib in women with CIN3. Specifically, investigators sought to determine whether 3 to 4 months of treatment with celecoxib could induce complete regression (remission) of CIN3 or partial regression to CIN1. They also wanted to assess the safety of celecoxib in that setting.
Patients with CIN3 that persisted 2 to 8 weeks after biopsy were randomized to celecoxib 400 mg daily or placebo. Follow-up assessments by colposcopy occurred after 8 weeks of treatment and again after 14 to 18 weeks. Patients with persistent cervical lesions after the second follow-up colposcopy underwent large loop excision of the transformation zone, and those without evidence of persistent lesion had a cervical biopsy for confirmation.
The trial also had a number of translational objectives, centering primarily on associations between histologic response and HPV type, lesion size, tissue concentration of COX-2 (H-score), serum VEGF levels, serum celecoxib levels, and cytology karyometry.
Of the 130 patients randomized, 91 were included in the final analysis: 50 treated with celecoxib and 41 assigned to placebo. Overall, 11 patients in the celecoxib arm achieved complete remission and nine others had partial remissions, resulting in an overall response rate of 40%. That compared with eight complete remissions and six had partial remissions in the placebo group, leading to an overall response rate of 34%. The 6% absolute difference did not achieve statistical significance (HR 1.28, 90% CI 0.57-2.89).
Data on serum VEGF levels were available for 81 of the 91 patients, and Rader and colleagues found that VEGF levels remained stable throughout the study period in the celecoxib group but increased substantially in the placebo group. Patients with the highest cervical concentrations of celecoxib had the largest decreases in both COX-2 and VEGF.
A subgroup analysis provided evidence of a greater effect of celecoxib in patients with elevated VEGF levels at baseline. Of 41 patients with elevated VEGF, nine of 19 in the celecoxib arm had objective responses in CIN3 lesions (47.4%) compared with three of 22 in the placebo group (13.6%). In the remaining 40 patients, response rates were 33.3% with celecoxib and 43.8% with placebo.
Interest in the effect of celecoxib on VEGF related to existing evidence that celecoxib inhibits VEGF gene expression and reduces VEGF levels in various murine models of carcinogenesis. Laboratory and clinical studies of COX-2 inhibition have shown significant reductions in VEGF in association with high levels of COX-2. Previous studies also have shown increased levels of VEGF and angiogenesis along the basement membrane of pre-invasive lesions in human cervical tissue and in a transgenic mouse model of cervical cancer.
"Our findings are very consistent with previous work," said Rader. "It will be interesting to validate some of these associations and to better understand the VEGF-COX-2 relationship in CIN and whether this has to do with VEGF possibly being a marker of patients who are less likely to have spontaneous regression or it may have something to do with an actual treatment effect of celecoxib."
The study was supported by Pfizer, NRG/GOG, and the National Cancer Institute.
Rader disclosed no relevant relationships with industry.
Aucun commentaire:
Enregistrer un commentaire