Time-dependent relationship between urinary biomarkers of nucleic acid oxidation and colorectal cancer risk.

[BACKGROUND] Experimental models indicate that oxidative stress may shift from promoting to suppressing tumor development as cancer progresses.

[METHODS] We conducted a nested case-control study within 2 Shanghai cohorts for primary analysis and 1 US cohort for replication. Over a median follow-up of 15.1 years in the Shanghai cohorts, 1938 incident colorectal cancer (CRC) cases were identified and matched to 1 control each. In the US cohort, 251 incident CRC cases were matched to 2 controls each. Systemic oxidative stress was assessed using urinary markers of DNA oxidation (8-oxo-7,8-dihydro-2'-deoxyguanosine [8-oxo-dG]) and RNA oxidation (7,8-dihydro-8-oxo-guanosine [8-oxo-Guo]) via ultra-performance liquid chromatography tandem mass spectrometry. Multivariable-adjusted odds ratios for CRC risk were calculated.

[RESULTS] After adjusting for selected covariates, an inversion association between oxidative stress markers and CRC risk was observed in the Shanghai cohorts and independently replicated in the US cohort. This inverse association was time-dependent, manifesting only for CRC cases diagnosed within 5 years following enrollment. Odds ratios for CRC at the 10th and 90th percentiles of 8-oxo-dG levels, relative to the median, were 1.87 (95% confidence interval [CI] = 1.39 to 2.53) and 0.48 (95% CI = 0.37 to 0.63), respectively, demonstrating a threefold difference in risk, with a P value for overall association less than .001. A similar pattern was observed for 8-oxo-Guo. No statistically significant associations were found for CRC diagnosed beyond 5 years of enrollment.

[CONCLUSION] This novel finding of a time-dependent inverse relationship between systemic oxidative stress and CRC risk, if further confirmed, may prompt a reevaluation of redox-based chemoprevention strategies.