Measured and genetically predicted plasma YKL-40 levels and melanoma mortality.

ABSTRACT

PURPOSE: High plasma levels of YKL-40 might be associated with mortality in patients with melanoma, and it is unknown if YKL-40 is causally related to mortality. EXPERIMENTAL DESIGN: We studied two cohorts 2618 patients with melanoma from hospital clinics and 1413 general population patients with melanoma, totalling 4031 patients followed up for mortality end-points for up to 20 years. All were genotyped for CHI3L1 rs4950928, highly predictive of lifelong plasma YKL-40, and plasma YKL-40 levels were measured in 2165 patients. We tested the hypotheses that measured and genetically predicted high plasma YKL-40 are associated with increased mortality in patients with melanoma. RESULTS: For the hospital melanoma cohort, age- and sex-adjusted hazard ratios for death in individuals with measured plasma YKL-40 in the 96-100th percentile versus 1-95th percentile and per 10-percentile increase were 1.52 (95% confidence interval, 1.07-2.16) and 1.07 (1.02-1.11), respectively, most pronounced for patients with localised melanomas. Each C-allele of the CHI3L1 rs4950928 genotype was associated with plasma YKL-40 level increases of 32% in the hospital melanoma cohort (p = 6 × 10-48) and 43% in the general population melanoma cohort (p = 7 × 10-13). Multifactorially adjusted ratios for these increases in the combined cohorts were 1.04 (1.00-1.09) observationally for measured plasma YKL-40 and 0.98 (0.86-1.12) for the genetically predicted plasma YKL-40. CONCLUSION: Measured, but not genetically predicted, increasing plasma YKL-40 was associated with increased mortality in patients with melanoma. Plasma YKL-40 is a marker but less likely to be a cause of increased mortality in patients with melanoma.