Insulin-like growth factor physiology and cancer risk M. Pollak * Cancer Prevention Research Unit, Lady Davis Research Institute of the Jewish General Hospital and McGill University, Montreal, Quebec, Canada H3T 1E2 Received 24 March 2000; accepted 14 April 2000 Abstract In the past few years, both laboratory investigations and population studies have provided strong circumstantial evidence that insulin-like growth factor (IGF) physiology in¯uences cancer risk. In contrast to the in¯uence of germ line mutations that are rare but are associated with very high risks, the impact of inter-individual variability in IGF physiology on risk appears to be modest but to eect a relatively high percentage of the population. Although this ®eld of investigation is young, attention is already being given to the possibility that it may be relevant to clinical assessment of risk and/or to the identi®cation of novel prevention strate- gies and intermediate endpoints. This review summarises key results in this ®eld and provides a hypothesis concerning the mechanism by which IGF physiology in¯uences risk of common epithelial cancers including those of breast, prostate, lung and colon. # 2000 Published by Elsevier Science Ltd. All rights reserved. Keywords: Colon cancer; Breast cancer; Prostate cancer; Lung cancer; Prevention; Risk; IGF-I; IGFBP-3 1. Introduction Insulin-like growth factors (IGFs) are potent mito- genic and anti-apoptotic molecules involved in the reg- ulation of cell proliferation in renewing epithelial cell populations of organs including breast, prostate, colon and lung. Unlike many other regulatory peptides, IGFs have characteristics of both classic `endocrine' hor- mones and also characteristics of tissue growth factors. IGF physiology has been the subject of several recent reviews [1,2]. Circulating levels are subject to complex physiological regulation. The vast majority of circulat- ing IGFs are bound to high anity IGF binding pro- teins (IGFBPs), with >90% bound to a high molecular weight complex comprised of IGF binding protein 3 (IGFBP-3) and a separate protein known as the acid- labile subunit. However, IGF bioactivity in tissues is not merely a function of circulating levels: local expres- sion of genes encoding IGFs, IGFBPs and proteases that digest IGFBPs are at least as important as circu- lating levels in determining the IGF bioactivity in a given tissue. Whilst IGFBP-3 functions as a carrier protein in the circulation, this molecule as well as other IGFBPs are also found in the extra-vascular tissue compartment, where they appear to have separate roles, generally as growth inhibitors [2]. It is important to recognise that there is considerable heterogeneity between normal individuals with respect to IGF-I and IGFBP-3 levels. In the past, endocrinolo- gists de®ned the relatively rare disease states of hypopi- tuitarism (associated with IGF-I de®ciency) and acromegaly (associated with IGF-I excess), but did not investigate in depth the possibility of physiological sig- ni®cance of variability between individuals within the broad `normal' range between these extremes. More recent studies have con®rmed the inter-individual var- iation, and have begun to investigate the genetic and non-genetic determinants of IGF-I and IGFBP-3 plasma levels [3±5]. Converging data from recent population studies and laboratory studies (see below) suggest that IGF phy- siology has important in¯uences on cancer biology, cancer risk and carcinogenesis. This convergence of data from dierent disciplines has generated considerable interest. However, it must be emphasised that this is a relatively novel topic in the IGF ®eld, and in many cases additional studies are needed to con®rm initial obser- vations and determine their signi®cance. 0959-8049/00/$ - see front matter # 2000 Published by Elsevier Science Ltd. All rights reserved. PII: S0959-8049(00)00102-7 European Journal of Cancer 36 (2000) 1224±1228 www.ejconline.com * Tel.: +1-514-340-8248; fax: +1-514-340-8302. E-mail address: md49@musica.mcgill.ca (M. Pollak).