M-A Rowlands, K Tilling, JMP Holly, C Metcalfe, D Gunnell, A Lane, M Davis, J Donovan, F Hamdy, DE Neal, RM Martin
Cancer Causes Control
PURPOSE: Active monitoring of prostate cancer requires the selection of low-risk cancers and subsequent identification of disease progression. Our objective was to determine whether serum insulin-like growth factor (IGF)-I, IGF-II, IGF-binding protein (IGFBP)-2 or IGFBP-3 at diagnosis (potential biomarkers of prognosis), and repeated measures of IGFBP-2 (potential biomarker of tumour growth), were associated with annual change in PSA and PSA doubling time (PSADT), proxies for disease progression. METHODS: We investigated associations of circulating IGFs and IGFBPs with PSA measures using multilevel models, in 909 men (recruited between 1999 and 2009) with PSA-detected clinically localized prostate cancer undergoing active monitoring in the United Kingdom. Each man had an average of 14 measurements of PSA during a mean of 4-year follow-up. RESULTS: IGF-I, IGF-II, IGFBP-2, and IGFBP-3 were not associated with baseline PSA. There was weak evidence that IGF-I at diagnosis was positively associated with a rapid post-diagnosis PSADT (≤4 years vs. >4 years): OR 1.34 (95 % CI 0.98, 1.81) per SD increase in IGF-I. IGFBP-2 increased by 2.1 % (95 % CI 1.4, 2.8) per year between 50 and 70 years, with no association between serial IGFBP-2 levels and PSADT. There was no evidence that serum IGF-II, IGFBP-2, or IGFBP-3, or post-diagnosis IGFBP-2, were associated with PSA kinetics in men with PSA-detected localized prostate cancer. CONCLUSIONS: The weak association of IGF-I with PSADT requires replication in larger datasets, and more definitive evidence will be provided on the maturity of long-term active monitoring cohorts with relevant clinical outcomes (metastasis and prostate cancer mortality).