Urine PGE-M as a marker of intratumoral cyclooxygenase-2 activity in non-small cell lung cancer (NSCLC)

Abstract

9527 Background: Prostaglandin E2 (PGE2) is implicated in the pathogenesis of NSCLC. PGE2 levels are elevated in NSCLC often in association with overexpression of cyclooxygenase-2 (COX2). The most accurate method to assess PG production in vivo in humans is via quantification of urinary metabolites. We hypothesized that urine levels of 11α-hydroxy-9,15-dioxo-2,3,4,5-tetranor-prostane-1,20-dioic acid (PGE-M), the major metabolite of PGE2, could be a useful marker of tumor COX2 activity & PGE2 production. Methods: PGE-M levels were determined in urine specimens derived from a previously reported phase II trial employing celecoxib & docetaxel in recurrent NSCLC (Proc. ASCO 22:640, 2003). We employed liquid chromatography electrospray ionization MS using selected reaction monitoring. PGE-M was chemically synthesized & converted to a [2H6] methyloxime isomer for use as an internal standard. After purification, endogenous PGE-M was analyzed as a methyloxime derivative. Assay precision is ± 5%; accuracy is 92%. Results: Pts=57; RR=10.5%; MST=228 days. Urinary PGE-M levels = 11.3 ± 5.7 (mean±2 SD, n=10) & 6.4 ± 3.8 ng/mg creatinine (mean±2 SD, n=10) in normal men & women. In men & women with recurrent NSCLC, PGE-M levels were elevated 2-fold (p<0.01) & 3.1-fold above normal mean (p<0.005). PGE-M levels fell by a mean of 51% (p<0.001) & 43% (P<0.001) respectively in men & women following 5–7 days of celecoxib (400 mg p.o. bid). These data suggest that a large proportion of the PGE2 overproduction is COX2 derived. In a multivariate model accounting for sex, race, smoking history & histology we correlated PGE-M changes with clinical parameters & found a strong association with survival. Pts experiencing a reduction in urine PGE-M levels after celecoxib administration were much less likely to die than those who did not (P<0.005). Conclusion: Changes in urinary PGE-M levels following short course celecoxib may be a useful predictor of intratumoral COX2 inhibition. (Grant support: CA68485; P5090949). Author Disclosure Employment or Leadership Consultant or Advisory Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Aventis