Insulin-like growth factors (IGFs) and IGF binding proteins in PSA-detected prostate cancer: a large population-based case-control study


Session type:

Mari-Anne Rowlands1,Jeff MP Holly1,David Gunnell1,Jenny Donovan1,J Athene Lane1,Freddie Hamdy2,David E Neal3,Richard M Martin4
1University of Bristol, Bristol, United Kingdom,2University of Oxford, Oxford, United Kingdom,3University of Cambridge, Cambridge, United Kingdom,4MRC Centre for Causal Analysis in Translational Epidemiology, Bristol, United Kingdom


Several studies have investigated associations of insulin-like growth factors (IGFs) and IGF-binding proteins (IGFBPs) with prostate cancer risk. These are generally small, comprising either clinically-detected or a mixture of clinically- and screen-detected cancer; and results have been inconsistent. With more cancers now being detected in asymptomatic men via prostate-specific antigen (PSA) testing, the involvement of the IGF system is of increasing interest.


Cross-sectional case-control study nested within a population-based sample of 111,000 men (aged 50-69) receiving PSA tests in the community-based ProtecT trial of prostate cancer treatments (ProtecT Study).  Of 3,174 with prostate cancer, 2,686 provided serum samples at recruitment, along with a random sample of 2,766 age- and GP-matched controls. Cancers were mainly localized (n=2,355) and low grade (Gleason score<7) (n=1,808). IGF-I, IGF-II, IGFBP-2 and IGFBP-3 assays were performed on these serum samples.


IGF-I was not related to prostate cancer (OR* 0.99, 95% CI: 0.93,1.04) (ptrend=0.62). There was a positive association of IGF-II (OR* 1.16; 1.08,1.24) (ptrend <0.001), IGFBP-2 (OR* 1.18; 1.06,1.31) (ptrend<0.01) and IGFBP-3 (OR* 1.27; 1.19,1.36) (ptrend <0.001) with prostate cancer. IGF-I was inversely related to cancer after adjustment for IGFBP-3 (OR* 0.85; 0.79,0.91) (ptrend <0.001). IGFBP-3 remained positively associated after adjustment for IGF-I (OR* 1.42; 1.32,1.53) (ptrend <0.001). There was some evidence that IGF-I was positively associated with localised cancer (after IGFBP-3 adjustment, OR* 1.12; 0.97,1.31) but IGFBP-2 was positively associated with advanced cancer (OR* 1.10; 0.97,1.26). IGF-II and IGFBP-3 associations did not differ by stage or grade.

*OR per standard deviation increase in IGF/IGFBP


IGF-I was unrelated, whilst IGF-II, IGFBP-2 and IGFBP-3 were positively related, to PSA-detected prostate cancer in this population-based study, the largest in the world.  This contrasts smaller studies, and suggests that the role of the IGF system may differ according to whether cancers are PSA- or clinically-detected.