For many years, nonspecific and widely acting cytotoxic drugs have been the conventional approach in cancer therapy. The scope of treatments for cancer patients has been expanding and transitioning as scientists continue learning about malignant cells and the genetics that control them. As it has become increasingly clear that individuals who suffer from the same type of cancer often have very similar genetic defects, scientists have begun to focus on gene-expression signatures as molecular diagnostic tests, so that medically relevant therapies can be utilized in the treatment of cancer patients.1 In addition, gene-expression signatures can display a poor prognosis signature in small, newly developing tumors, helping to indicate early on which
…show more content…
With DNA microarray technology, we can acquire the expressions of over 20,000 different genes per experiment.3 A labeled target sample, either RNA, cRNA, or cDNA, is analyzed with probes and a relative level of gene-expression can be obtained. There are numerous approaches and algorithms for interpreting the microarray results, and in doing so, we can see patterns of gene-expression, which we can then use to categorize tumors based on their characteristic expressions.4 These patterns are known as gene-expression signatures. Using gene-expression signatures, we can find both prognostic and predictive biomarkers. Specifically, prognostic biomarkers help tell us whether patients would benefit from additional therapy, and predictive biomarkers can tell us which therapies would be most beneficial to the …show more content…
The data-driven approach for finding patterns in gene-expression is fairly clear-cut. We analyze an entire genome’s expression, and then search for correlation in the genes and the tumor traits. This approach is very useful in finding pertinent therapies for breast cancer, for example. In breast cancer patients, chemotherapy or hormonal therapy will greatly reduce the risk of dangerous metastases, the cancer cells that spread to other parts of the body. However, it has been suggested that up to 75% of patients who receive this treatment do not need it, and are undergoing a harmful therapy with no benefits.5 A data-driven study of patients with breast cancer found a gene-expression signature strongly predictive of a short time interval to distant metastases.2 The findings of this study help provide a prognostic biomarker—a basis for selecting patients who would benefit from additional therapy. However, though this method is very unbiased and impartial, it can also be inefficient to search through entire genomes, and the conclusions we can draw depend greatly on the quality of the samples used. The knowledge-driven approach is completely opposite from the data-driven approach. Only genes thought to be associated with the cancer are analyzed for patterns. If a gene or transcription factor is known to affect a prognosis,
With DNA microarray technology, we can acquire the expressions of over 20,000 different genes per experiment.3 A labeled target sample, either RNA, cRNA, or cDNA, is analyzed with probes and a relative level of gene-expression can be obtained. There are numerous approaches and algorithms for interpreting the microarray results, and in doing so, we can see patterns of gene-expression, which we can then use to categorize tumors based on their characteristic expressions.4 These patterns are known as gene-expression signatures. Using gene-expression signatures, we can find both prognostic and predictive biomarkers. Specifically, prognostic biomarkers help tell us whether patients would benefit from additional therapy, and predictive biomarkers can tell us which therapies would be most beneficial to the …show more content…
The data-driven approach for finding patterns in gene-expression is fairly clear-cut. We analyze an entire genome’s expression, and then search for correlation in the genes and the tumor traits. This approach is very useful in finding pertinent therapies for breast cancer, for example. In breast cancer patients, chemotherapy or hormonal therapy will greatly reduce the risk of dangerous metastases, the cancer cells that spread to other parts of the body. However, it has been suggested that up to 75% of patients who receive this treatment do not need it, and are undergoing a harmful therapy with no benefits.5 A data-driven study of patients with breast cancer found a gene-expression signature strongly predictive of a short time interval to distant metastases.2 The findings of this study help provide a prognostic biomarker—a basis for selecting patients who would benefit from additional therapy. However, though this method is very unbiased and impartial, it can also be inefficient to search through entire genomes, and the conclusions we can draw depend greatly on the quality of the samples used. The knowledge-driven approach is completely opposite from the data-driven approach. Only genes thought to be associated with the cancer are analyzed for patterns. If a gene or transcription factor is known to affect a prognosis,