1. Bernard, Philip S. MD

Article Content

Nucleic acids serve as important biomarkers in pathology and oncology for the diagnosis, prognosis, and treatment of cancer. Prognostic biomarkers provide information about the probability of survival (either progression free survival or overall outcome) when patients are given the standard of care for their stage of disease.

Philip S. Bernard, M... - Click to enlarge in new windowPhilip S. Bernard, MD. PHILIP S. BERNARD, MD, is Associate Professor in the, Department of Pathology, University of Utah, Salt Lake City.

In early stage disease, prognostic biomarkers can help the clinician decide how aggressive to be with treatment (i.e., whether to give chemotherapy or not). The clinical utility of prognostic assays for early stage, ER-positive breast cancer using multianalyte gene expression tests has been shown extensively (Breast Cancer Res 2015;17:11). In late stage cancer, molecular prognostic tests have little clinical utility since patients are already known to have a poor prognosis and have often failed first-line chemotherapy. Here, the identification of predictive biomarkers becomes important-that is, the detection of specific alterations in the patient's cancer cells or blood that allows targeted therapy.


The identification of translocations in hematologic malignancies set the stage for the evolution of molecular diagnostics in solid tumors. Over 25 years ago, the reciprocal translocation between chromosomes 9 and 22-t(9;22), resulting in the BCR-ABL gene and protein fusion (bcr-abl), became pathognomonic for the diagnosis of chronic myeloid leukemia (CML). Years later, a tyrosine kinase inhibitor (imatinib) targeting the BCR-ABL fusion was validated in clinical trials for prediction of response in CML. Imatinib is a somewhat promiscuous drug and will also bind to the platelet-derived growth factor receptor and c-kit, which are most often expressed in gastrointestinal stromal tumors and occasionally melanoma. The transition had thus begun from using molecular diagnostics to guide drug therapy in hematologic malignancies to now doing this in solid tumors.


A similar course followed in breast cancer with the discovery of the tyrosine kinase receptor HER2/neu, as first a biomarker for poor prognosis, and then a target for predictive therapy with trastuzumab-a monoclonal antibody against the extracellular domain of this transmembrane protein receptor. Within the same family of HER2 is the epidermal growth factor receptor (EGFR; HER1), which also can be mutated and activated mainly in non-small cell lung cancer. Activating mutations in EGFR can be targeted with a monoclonal antibody (erlotinib) or small molecular inhibitor (gefitinib). These biologic drugs are primarily reserved for very aggressive disease and/or late stage disease where other treatment options have been exhausted.


The distinction between prognostic and predictive biomarkers is often blurred due to their potentially overlapping roles indicating disease aggressiveness and their ability to be targeted by biologic therapy. Nevertheless, these distinctions and indications have monetary and regulatory implications for industry and the FDA, particularly for filing a companion diagnostic, as it is the difference between a 510(k)-clearance and a pre-market approval submission.


Currently, only single gene test for specific mutations have been FDA-approved. For example, the THxID BRAF test is intended to detect V600 mutations in melanoma for predicting response to dabrafenib and trametinib, a BRAF and MEK inhibitor respectively. The existence of a FDA-cleared or approved test for a biomarker does not preclude CLIA labs from developing their own laboratory developed test (LDT) for the same analyte. A caveat to developing these LDTs is when the data needs to provide a continuous risk score or when cutpoints are needed to decide whether to provide a particular treatment. These cutpoints need to be made using the same chemistry on the same platform, and using the same bioinformatics algorithm as that which was used in the clinical trial that validated the drug indication.


Next Generation Sequencing

Advancements in nucleic acid-based technologies have provided the ability to interrogate many regions of the genome simultaneously through massively parallel sequencing (e.g., next generation sequencing). Next generation sequencing (NGS) has revealed the complexity and commonality of molecular alterations that occur across different cancers. Compilations of the most frequent (e.g., TP53 mutations) and/or actionable variants have been converted to targeted NGS panels through commercial and academic efforts (J Mol Diagn 2015;17(3):251-64). Overall, the depth and coverage of NGS has improved the accuracy and limits of detection for finding somatic mutations compared to conventional Sanger sequencing or pyrosequencing (J Mol Diagn 2015;17(3):251-64). In addition, it can interrogate hundreds of genes for different types of alterations including single nucleotide variants, small insertions or deletions (indel), copy number variants, and translocations.


Some of the current drawbacks to NGS are the cost of the instrumentation and chemistry, inadequate medical insurance re-imbursement, requirement for sophisticated bioinformatics, and relatively high assay failure from small tissue samplings. Regardless of these limitations, it comes with great opportunity, especially in the area of liquid biopsies. There is now general acceptance in obstetrics to use next generation sequencing as a method of non-invasive prenatal testing (NIPT) to detect fetal chromosomal abnormalities circulating in the mother's blood. By happenstance, NIPT has shown the potential to also detect cancer early in the bloodstream by quantitating the cell free DNA (cfDNA) copy number of oncogenes amplified in the tumor (JAMA Oncol 2015;1(6): 814-9, JAMA 2015;314(2):162-9). In the near future, liquid biopsies could obviate the need for invasive tissue biopsies. There is already evidence that the same alterations found in the primary tumor can be found in the cfDNA extracted from plasma. These tests are finding immediate application in monitoring the progression of solid tumors from the blood. Even more sensitive methods, such as digital droplet PCR (ddPCR), are being used for early detection of emerging resistance to biologic therapy, such as finding the EGFR T790M resistance mutation in the blood of lung cancer patients receiving tyrosine kinase inhibitors (Clin Cancer Res 2016;22(5):1103-10).


Cutting-Edge Oncology Future

The goal of precision medicine is to provide the highest quality health care to the most people and at the least cost. This initiative comes at the same time that the days of "one-drug-fits-all" have past, at least for the treatment of advanced cancer after standard chemotherapy has failed. Truly personalized healthcare comes with a large price tag, especially in oncology, where each cancer is as unique as the individual patient. This requires a continuous pipeline of new and expensive biologic therapies (antibodies and small molecule inhibitors) to be developed, tested in clinical trials, and then used in standard practice. Coinciding with advances in drug development from the pharmaceutical industry, there is parallel progress (and expense) in the discovery of targetable alterations/pathways in oncology and the translation of these biomarkers into clinical care.


It is evident the future of oncology and pathology will use these cutting-edge technologies in molecular diagnostics to monitor solid tumors and guide targeted therapies. As these technologies are applied for early detection and prevention, the cost-effectiveness and gain in quality-adjusted life years should justify their routine use in medicine.