Neoplasms escape selective COX-2 inhibition in an animal model of breast cancer.

BACKGROUND: Cyclo-oxygenase-2 (COX-2) is up-regulated in malignant tumours rendering it an attractive target for cancer therapeutics. However, whether long-term antagonism maintains its initial efficacy on established tumours is unclear.
METHODS: 4T1 cells were injected into the mammary fat pad of BALB/c mice (n = 8). Once tumour deposits were established, animals were randomized into two equal groups to receive either a selective COX-2 inhibitor (SC-236) or a drug vehicle. Further animals similarly treated (n = 7) were studied in diuresis cages allowing urine capture and analysis by mass spectrometry to determine Prostaglandin F-1 levels (PGF-1). In addition, both wild-type receiving SC-236 and COX-2 knockout mice receiving either SC 236 or vehicle were subjected to the same studies to determine whether tumour-derived or host-derived (stromal) COX-2 was the critical element. Finally, BALB/c mice with 4T1 tumours (n = 7) were treated with a combination of COX-2 and lipoxygenase (LOX) inhibition to attenuate this escape phenomenon.
RESULTS: While selective COX-2 inhibition initially retarded tumour growth, a rapid increase in tumour growth rate occurred later (day 9). This escape phenomenon correlated with an increase in urinary PGF-1 levels. An identical trend was also observed whether COX-2 knockout mice received SC-236 or not, suggesting that this effect is due to increased tumour-derived COX-2 production rather than recovery of host COX-2 functional capacity. Finally, dual inhibition of COX and LOX pathways attenuated this escape process.
CONCLUSION: The anti-neoplastic effects of selective COX-2 inhibition may not be sustained as tumours demonstrate an escape capacity. However, this phenomenon maybe attenuated by a combination of COX/LOX inhibitors.