Lung cancer is the leading cause of death from cancer worldwide, and the care rate remains less than 15% despite improvements in surgery, radiotherapy and chemotherapy [1]. Insulin-like growth factor-I (IGF-I) and IGF-binding protein-3 (IGFBP-3) have been widely accepted that they may have key role on the genesis and development of many types of tumor including lung cancer [2–7].

Growth hormone stimulates production of IGF-I in the liver and peripheral tissues. IGF-I is also released locally in response to damage, either directly or through the action of other factors associated with tissue responses to damage, including epidermal growth factor, fibroblast growth factor, and platelet-derived growth factor [8]. IGFBP-3 is the dominant circulating binding partner for both IGFs, accounting for 70 to 80% of their blood levels [8, 9]. Multiple lines of evidence suggest involvement of the IGF pathway across a range of malignancies, including both non-small cell lung cancer (NSCLC) and small cell lung cancer [5, 10, 11]. Elevated plasma levels of IGF-I have been associated with an increased risk of lung cancer, and high plasma levels of IGFBP-3 associated with a reduced risk [5]. Similarly, IGFBP-3 promoter methylation in tumor cells has been linked to decreased survival in stage I NSCLC patients. These suggest that IGF-I may promote tumor cell growth, while IGFBP-3 acts as a tumor suppressor gene [12, 13]. At the same time, different results were obtained from other studies. Recently, many large-scale clinical prospective case-control studies on association between circulating levels of IGF-I, IGFBP-3 and the risk of lung cancer were performed [14–19]. However, the results of these studies still remain inconclusive, partially because of the possible relatively small sample size in each of the published studies.

Here we performed a systematic meta-analysis of all studies published to date to determine and assess the strength of the association between circulating levels of IGF- I and IGFBP-3 and lung cancer. It may be helpful in the diagnosis and treatment of lung cancer.

Lung cancer is the leading cause of malignancy-related mortality. The mechanism of carcinogenesis is very complex, which involves many factors, such as IGF-I and IGFBP-3. Conventional studies coordinately think that IGF-I and IGFBP-3 may promote and inhibit tumor growth, respectively. In recent years, there are many epidemiological studies have different results. In this meta-analysis, our data suggests that IGF-I low in the lung cancer population, though we could not demonstrate statistical significance. With regard to the association between IGFBP-3 and lung caner, the data suggests IGFBP-3 acts as a tumor suppressor and has a inverse correlation with the risk of lung cancer, and it does have statistical significance.

The IGF family is supposed to play a pivotal role in regulating cell proliferation, apoptosis and transformation [24]. Most circulating IGFs are produced by hepatocytes in response to growth hormone stimulation [25–27]. Circulating IGFBP-3 is produced by hepatic endothelium and Kupffer cells [26, 27]. A number of in vitro and in vivo studies have demonstrated that IGF-I is an effective mitogen in normal epithelial cells and has strong antiapoptotic effects on lung cancer cells [5, 10, 11]. However, the effect of IGF-I may be modulated by IGFBP-3 in circulation because most of the IGF-I is bound to IGFBP-3 and once bound it is not in its active form. The results of this meta-analiysis indicate that there are no statistically significant association between IGF-I and lung cancer, while the associaton between IGFBP-3 and lung cancer is very significant. High serum levels of IGFBP-3 associated with a reduced lung cancer risk.

Lung cancer is a multifactorial disease that results from complex interactions between many genetic and environmental factors. This means that there will not be single gene or single environmental factor that has large effects on lung cancer susceptibility. Due to limited studies and published data in original report, we could not make subgroup analysis by sex, smoking status, histological types and other variables. Possible limitations of our meta-analysis includes relatively small number of studies, different heterogeneous matching factors, different countries and ethnicities, possible publication bias, as well as possible interaction with other biologic and environmental factors. It is well documented that ethnic factor contributes to the lung cancer incidence. In our study, we included 2 U.S., 1 Chinese, 1 Japanese, 1 Finnish and 1 British studies. Therefore, heterogeneity by ethnicity needs to be taken into account when interpreting our data. Heterogeneous matching factors and differential adjustment for confounding factors are other sources of bias. The above limitations might have contributed to the low statistical power of our meta-analysis.

Despite some limitations, our results based on nested case-control studies which represent of best study design. In addition, we obtained the results from dichotomous and continuous variable respectively, which made the results more reliable. What's more, heterogeneity and publication bias of the studies were not significant. Thus, the data of our study are reliability.

In summary, we found that association between circulating levels of IGF-I, IGFBP-3 and the risk of lung cancer are marginally and statistically significant, respectively. So it may be helpful in the diagnosis and treatment of lung cancer. Since circulating IGF-I and IGFBP-3 remain important factors in lung cancer, more studies need to be conducted to discern this association. And uniform adjustment of confounding factors across the studies will help in terms of interpretability and comparability.