Diagnostics: Genomics-based Tests Can Help Diagnose Uncertain Cancers

Oncology Live®, July 2008, Volume 9, Issue 7

New genomics-based tests are helping oncologists answer one of the most basic, yet often challenging, questions in cancer care: What kind of cancer does the patient have?

There are an estimated 200,000 cases per year in the United States in which the patient’s primary tumor site is uncertain. Diagnosing such hard-to-identify tumors can be time-consuming, expensive, and frustrating for both physicians and patients. Accurately identifying the site of origin is necessary in order to begin cancer-specifi c treatment per National Comprehensive Cancer Network (NCCN) guidelines. This is especially important given the increasing availability of new, targeted therapies. These difficult primary cases often occur in the following situations: 1) the cancer was found in an unexpected location; 2) the tumor cells are poorly diff erentiated or undiff erentiated, making them diffi cult to interpret; and 3) the cancer was found in multiple locations demonstrating metastatic disease, with no apparent primary site. Traditional diagnostic approaches for such cases are subjective, time-consuming, and often do not produce a defi nitive diagnosis. Even with the use of immunohistochemistry (IHC), electron microscopy, and advanced imaging procedures, the primary site is identified in only about 20-25% of living patients with metastatic tumors for which the primary site is not apparent after the initial workup. Patients in whom the primary cancer is diagnosed have been shown in a prospective study to have a longer survival compared to patients in whom the tissue of origin remains unknown.

Current diagnostic techniques

A number of different methods are available to aid physicians in evaluating hard-to-identify tumors. These include such imaging tests as computerized tomography, MRI, and PET scanning to locate the primary tumor mass. These imaging techniques may pinpoint the anatomic location of the tumor, but cannot confi rm the origin of the tumor. Further tests on the biopsy sample—such as IHC stains—have proven to be highly useful tools for narrowing the diff erential diagnosis of candidate primary sites. However, they have limited specifi city and sensitivity and do not address the full range of potential tumor types. The most commonly recommended staining phenotypes (CK7 and CK20) produce enough false positives and false negatives to make a definitive diagnosis difficult.

New advances in genomics testing

Gene expression tests that use microarray and real-time polymerase chain reaction (RT-PCR) technology are helping to fi ll the diagnostic gap in cases of uncertain tumors. Microarrays simultaneously measure expression levels of large numbers of genes, enabling multiple diagnostic questions to be answered at once. Probes representing a short section of a specifi c gene’s DNA sequence are arranged on a chip. RNA is extracted from the tissue of interest, labeled with a fl uorescent marker, and spread over the microarray. The RNA from the tissue of interest binds to complementary genespecifi c probes on the array in a process known as hybridization. Th e relative fl uorescence intensity is then measured, which refl ects the hybridization level and thus the expression of a particular gene. Once confi ned to the research lab, microarrays are increasingly being used in clinical diagnostics.

The Pathwork® Tissue of Origin Test, for example, uses more than 1,500 genes to compare an uncertain tumor’s gene expression profi le to those of 15 known tumor types, representing more than 60 morphologies, in order to provide an objective, probability-based score for each potential tissue. The test demonstrated a sensitivity of 89% and a specifi city of 99% in a validation study of 487 metastatic and poorly diff erentiated and undiff erentiated tumors, in which the primary had already been identifi ed using current methodologies. The test is currently off ered through Pathwork Diagnostics’ CLIA-certifi ed laboratory. With RT-PCR, the expression of a small number of pre-selected genes (generally <100) can be measured by amplifying the gene segments of interest using appropriate primer pairs. MD Anderson Cancer Center, for example, is studying a 10-gene RT-PCR assay to identify certain types of unknown tumors, which may respond to targeted treatment. RT-PCR is a cost-eff ective and accurate method for measuring the expression of a small number of genes but cannot provide as much information as a microarray. The ability to analyze large numbers of genes is especially important for diagnosing uncertain primary tumors because numerous potential primary tumor sites must be evaluated at once.

New genomics-based tests are increasingly being used to help oncologists identify uncertain primary tumors. This can benefit patients by enabling cancer-specifi c treatment to begin sooner and by allowing patients to potentially access appropriate clinical trials. Ultimately, such genomics tests should improve patient outcomes, while reducing costs in the healthcare system.

Dr. Henner is the Chief Medical Offi cer for Pathwork Diagnostics, Inc.