Impact of Genomics on Medicine  
By Jingfang Ju, Ph.D., Head, Cancer Genomics Laboratory Mitchell Cancer Institute, University of South Alabama

The development of genomics will make personalized medicine a reality in the new millennium. The impact of genomics on drug target discovery, drug development, and biomarker-based therapy tailored to individual’s unique genetic background will be enormous and everyone will benefit from it.

Human disease is quite complex. In the case of cancer, it is not a single genetic disease, but rather, hundreds of diseases consisting of various combinations of genetic alterations. Many types of genetic alterations contribute to neoplastic transformation. This requires a comprehensive and systematic approach to identify key genetic events that are associated with cancer, in particular, disease causing genes and pathway identification, biomarkers associated with disease progression, metastasis, clinical treatment response prediction and prognosis. With the development of information technology and biotechnology in the past decade, genomics has emerged as a powerful integrated approach for study cancer and other human diseases. Genomics integrate the state-of-art genomic technology such as sequencing, expression analysis, array cGH, SNP analysis with sophisticated bioinformatics tools to help us understand the complexity of cancer⁽¹⁾. Genomics will make personalized medicine a reality in the new century. Patients will get the best treatment options based on their individual molecular gene signatures to maximize the efficacy, minimize the toxicity. Pharmaceutical companies will benefit by improving the success rate of their drug development process using efficacy biomarkers and toxicity biomarkers identified by toxicogenomics and pharmacogenomics approaches^{(2, 3)}.

Impact of Oncogenomics on Cancer
Breast cancer is the most commonly diagnosed cancer in women, accounting for 23% of all cases⁽⁴⁾. Despite the improvements in detection and treatment of breast cancer, about 30-40% of patients still succumb to the disease due to tumor metastasis and resistance to therapy⁽⁴⁾. Accurate approach to identify responders with the right therapy is critical for breast cancer treatment. An ideal tumor marker should be expressed both unique to and universally in all breast cancer cells. It should be easily detectable, with little variance and bear clinical relevance.

One such example is the drug Herceptin and its biomarker HER2/neu. HER2/neu is one of the most important oncogenes in invasive breast cancer. Amplification of HER2/neu occurs in 30% of early-stage breast cancers, and a significant correlation between HER2/neu over expression and reduced survival of breast cancer patients has been found⁽⁵⁾. HER2/neu over expression correlates with a lack of response to endocrine therapy and chemotherapeutic agents⁽⁶⁾.

Recently, oncogenomics has provided for the successful development of a multi-gene assay that is capable of quantifying the likelihood of distant breast cancer recurrence to assist in treatment planning, named Oncotype DX (Genomic Health, Redwood City, California). This assay utilizes a 21-gene panel derived from different ontological cellular functions, such as proliferation (surviving, cyclin B1), estrogen activity (ER, PR, Bcl2), invasion (cathepsin L2),and HER2/neu status (GRB7, Her2). Furthermore, this is a good example of the clinical impact that oncogenomics can have on the prognosis of breast cancer patients, capable of predicting a patient’s risk of distant breast cancer recurrence in the future^{(48, 49)}.

Oncogenomics also impact the filed of colorectal cancer. Rational approaches have been used to identify response markers for colorectal cancer chemotherapy in the early 90s. Thymidylate synthase (TS) is a folate-dependent enzyme that catalyzes the reductive methylation of dUMP by 5,10-methylenetetrahydrofolate to form dTMP and dihydrofolate(35, 36). Because the TS-catalyzed enzymatic reaction provides the sole intracellular de novo source of thymidylate, an essential precursor for DNA biosynthesis, this enzyme has been an important target for cancer chemotherapy for the past 50 years(37-39). TS is also one of the most extensively investigated biomarkers in recent years, which attributed to its ability of predicting CRC patient responses to 5-FU based therapy^(40-44). Elevated TS expression levels have been shown to be associated with 5-FU resistance. The regulation of TS expression is quite complex at both transcriptional and post-transcriptional levels. In addition, TS can also down regulate p53 gene expression at the translational level⁽⁴⁵⁾ so that TS might be involved in coordinating the regulation of expression and/or function of cellular growth and proliferation. In addition to TS, thymidine phosphorylase (TP), and dihydropyriminide dehydrogenase (DPD) were also associated with chemoresensitivity to 5-FU based therapy⁽⁴⁴⁾.

Microarray-based gene expression analysis has significantly impacted CRC diagnosis and prognosis. It is a powerful tool for the systematic and combinatorial search for biomarkers for cancer classification and prediction. Bertucci et al reported on the prediction of CRC metastasis using gene expression cluster analysis, and separated stag IV patients from other patients. Wang et al defined a set of prognostic markers in Duckes’ B CRC patients using microarray gene expression profiling⁽⁴⁶⁾. A set of gene signatures were found to have predictive utilities in CRC^{(47, 48)}.

Eschrich et al used molecular staging based on 43 core genes from gene expression profiling to predict 36-month overall survival in 78 CRC patients, and concluded that molecular staging can better discriminate poor prognosis patients and has the potential to direct adjuvant therapy⁽⁴⁹⁾. A comprehensive overview of TS and 5-FU co-affected genes was discovered through microarray gene expression profiling by Xi et al⁽³⁴⁾. This study analyzed the expression variations between the steady state and polysome-associated mRNA transcripts regulated by TS and 5-FU. The post-transcriptionally regulated genes by TS and 5-FU were determined and will potentially benefit the predicting therapeutic outcomes of fluoropyrimidine-based cancer chemotherapy^(50-52).

Non-coding microRNAs (miRNAs) have been shown to be involved in colorectal cancer⁽⁵³⁾. The expression of a number of miRNAs, including let-7a, let-7g, has-miR-143, has-miR-145, hsa-miR-15b, hsa-miR-181b, hsa-miR-200c, and has-miR-191, was reported to be up- or down-regulated in CRC from independent in vivo and in vitro studies^(53-59). Bioinformatics analysis indicated that putative p53 binding sites were contained in 46% of the miRNAs putative promoters⁽³⁴⁾. Another study based on clinical analysis indicated that p53 mutation was associated with some miRNAs expression. Most importantly, patients with lower hsa-miR-200c expression tend to survive 12 months longer than patients with higher hsa-miR-200c expression⁽⁵⁵⁾. Let-7g and hsa-miR-181b were reported to be associated with patients’ response to oral 5-FU drug S-1⁽⁵⁴⁾.

One of the major limitations of microarray-based gene expression profiling is the quality of RNA from formal in-fixed paraffin-embedded (FFPE) sample specimens. Since fresh frozen samples are not always available, the application of the FFPE in microarray analysis was concerning to physicians and scientists. Preliminary data from our lab compared the gene expression correlation between paired fresh frozen and FFPE tissue and found only 28% of overall concordance in gene expression. Thus, a key issue related to the reproducibility of the RNA extraction process is whether RNA derived from FFPE tissue has sufficient quality to allow the generation of robust prognostic or predictive signatures(60). In contrast, in the same study, we found 89% of miRNAs from fresh frozen tissue correlated to those from FFPE tissue based on Locked Nucleic Acid (LNA) miRNA array. Due to its important roles mentioned above and high stability in archived FFPE specimens, miRNAs may hold a greater promise for biomarker and novel target discovery for cancer research and other diseases.

Given the continual rise in the number of potential biomarkers of CRC, future studies will increasingly employ genomic and proteomic technologies, which enable the measurement and analysis of numerous potential biomarkers simultaneously. These techniques are able to produce gene or protein ‘profiles’ associated with clinical outcome, the analysis of which may then yield novel biomarkers with prognostic and/or therapeutic potential⁽⁶¹⁾. Therefore, in the future, prediction of the individual risk of CRC with individualized screening and prevention could become reality.

Although the progress of oncogenomics has brought promising achievement on colorectal cancer diagnosis, treatment and prognosis, we should realize that these achievements have not been widely translated into clinical practice. There are many reasons that limit the translation. One is the limited sample size for many studies while another is the lack of prospective clinical study^{(62, 63)}.

Impact of Toxogenomics on Drug Development
We have witnessed the impact of toxogenomics on drug discovery over the past ten years. The concept of eliminating potential toxic drug leads from early drug development pipeline has become a reality with the aid of a toxogenomics approach^(64-68). Toxicogenomics is the study of the structure and output of the genome as it responds to adverse genotoxic exposure. Large-scale transcriptional analysis, made possible through microarray based expression technologies, enables us to understand the complexity of the biological effects of drugs and chemicals, with the ultimate goal of separating wanted effects from adverse effects. Recent advances in genomics include global assessment and classification of genome content, high-throughput biological pathway construction, systematic identification of previously unpredicted genes⁽⁶⁹⁾. These applications of genomics technologies, with conventional drug assessment methodologies, will lead to more tolerable drugs and a better understanding of clinical populations.

Regulatory Issues Related to Cancer Biomarkers
The last decade has seen incredible discoveries in the field of oncogenomics, having an enormous impact with the development of novel therapies and biomarkers. This rapid development has raised new issues and questions regarding scientific and clinical regulatory developments and ethical concerns. Biomarkers capable of predicting a drug's efficacy are particularly important.

Most biomarkers are much more subtle than tumor growth; typically, they detect a protein, genetic variation or other molecules. The American Association for Cancer Research, together with the Food and Drug Administration and National Cancer Institute, have announced the formation of the AACR-FDA-NCI Cancer Biomarkers Collaborative (CBC) to facilitate the use of validated biomarkers in clinical trials and ultimately in evidence-based oncology and cancer medicine⁽⁷⁰⁾.

The Biomarkers Consortium (BC), launched on November 5, 2006, is a public–private partnership whose goal is to identify and qualify new biological markers to accelerate the detection, diagnosis, and treatment of a range of diseases, including cancer(71). The Consortium is made up of a diverse list of partners with this common goal. It includes the Foundation for the National Institutes of Health (FNIH), the National Institutes of Health (NIH), the Food and Drug Administration (FDA), the Centers for Medicare & Medicaid Services (CMS), the Pharmaceutical Research and Manufacturers of America (PhRMA), the Biotechnology Industry Organization (BIO), and representatives of the public, including patient advocacy organizations. The NIH and its partners have long recognized a need for robust and objective measures of disease risk, underlying pathobiological processes, diagnosis and staging of disease, prognosis, treatment response, recurrence and clinical outcomes.

The availability of such biomarkers would not only enable research, but also serve to streamline clinical care and potentially speed the development and availability of new drugs. Although the need is clear, there are several important impediments to developing reliable biomarkers. Biomarker development requires insight into disease risk, natural history, and outcomes. It also requires a sufficiently large number of adequate samples taken from well-characterized patients and handled in a standardized fashion. Analytical platforms that effectively and reproducibly measure the biomarker must also be available and standardized^(72-78). Analytical approaches that assess the utility of biomarkers as signs or predictors of underlying biology or future outcomes must also be developed for the promise of biomarkers to be translated into clinical usefulness.

Once such biomarkers and analytical approaches are developed, they will ultimately prove to be quite useful in promoting discovery science and facilitate translational and clinical research. Useful biomarkers will greatly assist in the identification and stratification of human subjects to be enrolled in clinical trials, will shorten the duration of trials in which biomarkers can stand in as surrogates for more distant clinical endpoints, and may assist in excluding individuals at excessive risk for drug toxicity, thus translating to a more rapid, efficient, economical and safer drug development process. In the industry, as in the FDA and NIH, biomarkers can feed back into the discovery process, promoting the identification of new drug targets and facilitating the development of new drug entities.

The greatest benefit will be to the patients. The major beneficiaries of rapid, efficient and economical drug development are patients and the public at large. They are also the major beneficiaries of new insights into disease risk, characterization, and treatment. This requires a broad partnership of academic, industry and regulatory agencies to accomplish this ambitious goal.

Next Page