Plasma concentrations of IGF-I and IGFBP3 and brain cancer risk in EPIC

Background: Insulin-like growth factor-1 (IGF-I) is important in normal brain development, but in the adult brain IGF-I overexpression is a signal for neoplastic processes. It is unclear whether circulating levels of IGF-I and its main binding protein (BP), IGFBP-3, are related to subsequent risk of brain cancer. Methods: We examined the association between plasma concentrations of IGF-I and IGFBP3 and its molar ratio (IGF-I/IGFBP3) in relation to risk of high-grade gliomas (n=206), low-grade gliomas (n=74), meningiomas (n=174) and acoustic neuromas (n=49) using a case-control design (2 controls per case) nested in the European Prospective Investigation into Cancer and Nutrition. IGF-I and IGFBP3 were measured by enzyme-linked immunosorbent assays.Conditional logistic regression models were used to compute odds ratios (OR) and corresponding 95% confidence intervals (CI). Results: Circulating levels of IGF-I and IBFBP-3 alone were not associated with the risk of high-grade or low-grade gliomas. However, after mutual adjustment, the risk of low-grade gliomas was positively associated with IGF-I (OR=3.60, 95% CI 1.11–11.7; top vs. bottom quartile) and inversely with IGFBP-3 concentrations (OR=0.28, 95% CI 0.09–0.84). A doubling in the molar IGF-I/IGFBP-3 ratio yielded an OR=2.80 (95% CI 0.98, 7.98). No association was observed for meningiomas. Both high IGF-I and IGFBP-3 concentrations were associated with an increased risk of acoustic neuromas (IGF-I: OR=6.63, 95% CI 2.27–19.4, top vs. bottom tertile; IGFBP-3: OR=7.07, 95% CI 2.32–21.6). Conclusion: Our results suggest that high concentrations of IGF-I, either alone or after adjustment for IGFBP3, might be related to increased risk of low-grade gliomas and acoustic neuromas. While there is a logical rationale for an inverse association of IGFBP-3 with low-grade gliomas, the increasing risk associated with IGGBP-3 in relation to acoustic neuromas is difficult to interpret. These findings are novel but are based on small numbers and require confirmation in other studies.