Cancer Biomarkers

The role of early detection in cancer has shown improved survival for certain cancers, including colon cancer, cervical cancer and breast cancer. The possibility for early detection of pancreatic cancer may be realized by an improved understanding of the histology and molecular genetics of precursor lesions and cancerous lesions in pancreatic cancer and the development of sensitive and specific screening tests (both invasive and non-invasive) to detect early pancreatic cancer. The NIH-NCI Early Detection Research Network (EDRN) in Pancreatic Cancer has been focussed on the development and validation of new biomarkers for the detection of early pancreatic cancer. This review will focus on our understanding of the histologic and molecular model of pancreatic carcinogenesis, current strategies and limitations of screening for pancreatic cancer and the development and validation of new biomarkers for the early detection of pancreatic cancer.

We previously demonstrated that human urine contains small, 150 to 250 nucleotide-sized, soluble DNA derived from the circulation, which may be useful in the detection of colorectal cancer. In this report we have determined the stability of DNA in urine and have found that the half-life time interval of this small, fragmented DNA is at least 4 hours post collection. We further compared, in a blinded study, the frequency of detecting mutated K-ras sequence in DNA isolated from plasma and urine derived from individuals who have either a colorectal carcinoma (CRC), or adenomatous polyps that contain a mutation in codon 12 of the K-ras proto-oncogene. There was an 83% concurrence of mutated DNA detected in urine and its corresponding disease tissue from the same individuals, when paired urine and tissue sections from 20 patients with either CRC or adenomatous polyps were analyzed for K-ras mutation. However, only a 56% concurrence was observed when the matched plasma specimens were tested from these 20 patients. These results suggest that urine might be a better resource for detecting K-ras mutation in circulating DNA.

Cervical cancer is a major gynecologic malignancy around the world. However, current diagnostic methods such as Pap smear and human papilloma virus (HPV) testing are insufficient for an early diagnosis of cervical cancer, follow-up on therapy efficacy or to identify the women who might progress to cervical cancer (only about 1–5% of the HPV-positive women will develop cervical cancer). Patients with atypical squamous cells of undetermined significance (ASC-US) clearly need a better screening test. Developing a non-invasive method for early diagnosis of cervical cancer is essential. Our in vitro and translational research data support the hypothesis that:

1. Squamous cell cervical cancer is related to specific upregulation of tissue and serum Insulin-like Growth Factor-II levels (IGF-II; 100% sensitivity and 100% specificity). Serum IGF-I, but not IGF-II levels are elevated in other gynecological, breast, lung and prostate cancers. Serum IGF-II test helps in diagnosing cervical cancer as early as ASC-US or cervical intraepithelial neoplasia (CIN)-I and in monitoring therapy efficacy (p ‹ 0.001 by Student's 't' test and Chi-square analysis).

2. Concomittant to increased serum IGF-II levels, IGF-Binding Protein 3 (IGF-BP3) levels are significantly decreased in persistent CIN and cervical cancer (p ‹ 0.0001). As IGF-BP3 modulates IGF-II biological activity, significantly decreased serum IGF-BP3 levels (levels normalize after therapy; p ‹ 0.001) may indicate a poor prognosis. Similar to serum IGF-II, serum IGF-BP3 levels help monitoring therapy efficacy in cervical cancer and advanced CIN.

Measurement of serum IGF-II levels will help in early diagnosis of cervical cancer and monitoring of therapy outcome. Serum IGF-BP3 in conjunction with IGF-II levels may help in predicting prognosis as well as monitoring therapy efficacy.

Lung cancer remains the leading cause of cancer related mortality, accounting for almost one-third of cancer deaths in men and one-fourth of cancer deaths in women; 160,440 lung cancer deaths are expected in 2004. Survival from lung cancer depends mainly upon the stage at presentation. As localized tumors generally do not cause symptoms, the disease is usually diagnosed in symptomatic patients at advanced stages when the prognosis is poor. As a result, the overall 5-year lung cancer survival rate is only 15%.

It is well known that epigenetic alterations such as DNA methylation of CpG dinucleotides located in CpG islands within the regulatory (promoter) regions of genes are associated with transcriptional silencing in cancer. Promoter hypermethylation of critical pathway genes could identify potential biomarkers for lung cancer risk. Our goal for this study is to identify novel hypermethylated genes in lung cancer. We have investigated the methylation profiles of DNA samples from 14 paired lung tumor and adjacent normal tissues resected from the same individuals using restriction landmark genomic scanning (RLGS). We could assess the DNA methylation status of an average of 2,012 CpG islands for each tumor. We identified 162 differentially methylated loci where CpG islands were hypermethylated in lung tumors but not in adjacent non-cancer tissues. Among 162 sites of differential DNA methylation, detected from at least one tumor/normal pair, 21 hypermethylated genes were identified that were not reported previously as hypermethylated in lung tumor tissue.

DNA repair has a major role in suppressing the rate of accumulation of mutations. Therefore, variations in DNA repair are likely to play an important role in determining cancer risk. While there is compelling evidence that defects in DNA repair cause high predisposition to several hereditary cancers, there is a paucity of data on the role of DNA repair in sporadic cancers. We present our approach of using functional DNA repair tests, rather than gene polymorphism, to study the potential of DNA repair enzymes to serve as biomarkers for lung cancer risk. We have previously developed a functional DNA repair blood test for the enzymatic repair of the oxidative DNA lesion 8-oxoguanine, and found that reduced OGG activity is a risk factor in non-small cell lung cancer. Moreover the combination of smoking and low OGG activity was associated with a greatly increased lung cancer risk (Paz-Elizur et al, JNCI 95 (2003) 1312–1319). The use of OGG activity as a potential biomarker for lung cancer risk is validated in collaboration with the M. D. Anderson Cancer Center, under the sponsorship of the Associate Members Program of the Early Detection Research Network (EDRN, NCI, NIH).