Expression of insulin-like growth factor I and II in various regionally important solid tumors.

ABSTRACT

Insulin-like growth factor I (IGF-I) may thwart a T cell response to several cancers, since suppression of IGF-I expression using an episome transcribing an antisense RNA into rat glioma and murine teratocarcinoma, which are known to actively express IGF-I, resulted in tumor regression and CD8+ T cell infiltration.1,2,3,4 This finding led to the gene therapy protocol for human glioblastoma multiforme, a known IGF-I over-expressing tumor, using the same strategy in 1993. The hypothesis that cancers which over-express IGF-I can be cured by this episome has led us to search for other cancers which actively produce IGF-I. Many cancers were reported to express IGF-I4, but most of the geographically important tumors in Thailand had not been studied for the expression of IGF-I. Our immediate goal was to survey the leading human cancers in the Tumor Registry of Siriraj Hospital (i.e., non-small cell lung cancer, cholangiocarcinoma, hepatocellular carcinoma, papillary thyroid carcinoma, nasopharyngeal carcinoma) including all stages of astrocytomas for the expression of IGF-I and IGF-II (a homologue of IGF-I). Tumor tissue collected from paraffin blocks deposited at the Department of Pathology, Siriraj Hospital was screened for the presence of the predominant form of IGF's (IGF-I or -II). Tumor tissues that express any specific form of IGF's more than adjacent wild-type cells are potential candidates for an antisense gene therapy against the expression of that specific form of IGF. Paraffin sections containing these cancer cells and adjacent wild-type cells were examined by immunohistochemical technique using mouse IgG-1 antihuman IGF-I, or antihuman IGF-II as primary antibodies. Normal liver tissue, known as the largest producer of both forms of IGF's, was used as a positive control. The staining pattern of both IGF-I and IGF-II in the liver was highest in the cytoplasm of normal hepatocytes. The majority of the tumors except for papillary thyroid carcinoma and nasopharyngeal carcinoma strongly expressed IGF-I. Only a few samples of the tissues studied expressed a low level of IGF-II, if any. This raises a further working proposal that the blockade of IGF-I synthesis in these tumors may help to delay the tumor progression so that an effective immune response can be established and destroy the tumor.