Microarray Test Improves Cancer Diagnosis and Treatment

American Cancer Society

New genetic tools may help doctors make distinctions between types of cancer more easily in the future. DNA microarray profiling is one such tool.

Researchers are using DNA microarrays not only to distinguish between cancers, but to predict which cancers might respond to certain treatments as well.

Several research groups have reported successes using the technique in the past few years on cancers ranging from leukemia to breast cancer. But because the technology is complex, it has been limited to research labs and is not yet commonly used in medical practice.

Gavin J. Gordon, PhD, and colleagues from Brigham and Women's Hospital in Boston, used a simplified version of the DNA microarray technique to distinguish between cancers that affect the lining of the lung. They reported their results in the journal Cancer Research (Vol. 62: 4963-4967).

Cancers In The Linings Of The Lungs Hard To Tell Apart
Malignant pleural mesothelioma (MPM) is a rare cancer of the lining of the lung (known as the pleura). Often, one of the first symptoms of mesothelioma is difficulty breathing. This is caused by the accumulation of fluid in the space between the lungs and the pleura, a condition known as a pleural effusion.

But mesothelioma is not the only cancer that can cause a pleural effusion. Other cancers called adenocarcinomas, which start in the lungs or other parts of the body, can also invade the pleura and lead to an effusion.

Distinguishing between these cancers is important because doctors treat them differently. Usually, doctors will remove some of the excess fluid with a needle to look at the cancer cells under the microscope.

But this only provides an answer in about a third of cases, said the study authors. The remaining patients often must undergo surgery to get a tumor sample. And even this doesn't always provide a correct diagnosis.

Microarray Testing Promising But Often Impractical
Researchers using DNA microarray technology take advantage of the fact that cancer cells have defective genes. Genes are pieces of DNA in each cell that act as blueprints to make RNA. RNA, in turn, is used to make the proteins the cell needs to survive.

In cancer cells, faulty genes often lead to either faulty proteins or to too much or too little of a specific protein.

DNA microarray profiling uses a small, flat chip that has thousands of genes (made up of DNA) - or pieces of genes - embedded on its surface. Researchers take pieces of RNA from a tumor sample and apply them to the chip, where they stick to similar pieces of DNA.

By looking at where the RNA sticks on the chip, researchers can determine which genes are being actively used ("expressed") in the tumor cells.

Researchers use this technique to observe different patterns of gene expression between different tumors. But most researchers have had to look at many genes to find a pattern, which has thus far kept the use of microarrays limited to research labs.

Study: Technique Distinguishes Between Cancer Types
Gordon and colleagues wanted to see if they could get around this problem of needing to look at so many genes. They found that they only had to look at a few genes to get a correct diagnosis if they compared the ratios of how the genes were expressed.

The researchers first looked at the gene patterns of tumors from 16 patients with mesotheliomas and 16 patients with adenocarcinomas. From these, they found eight genes that were expressed at very different levels between the two types of cancer. They set up 15 different ratios between pairs of these eight genes.

The scientists then tried to see if these ratios would be useful on a second set of tumors (15 mesotheliomas and 134 adenocarcinomas). They found that any one of the ratios was more than 90% accurate in determining which type of tumor it was. When two ratios were looked at, the accuracy went up to 95%, and with three ratios it was 99%.

Microarray Expression Ratios A Promising Advance
In a second experiment, the authors looked at data from 60 patients with medulloblastoma, a type of brain tumor. Using six genes combined in nine ratios, they were able to accurately predict which patients would respond to treatment.

While noting that the technique is not perfect, the authors pointed out that the ratio method has several advantages over other microarray techniques, including possibly needing smaller amounts of tumor for testing.

The authors concluded by emphasizing that the ratio-based method using carefully chosen genes can be a simple, inexpensive, and highly accurate means to distinguish MPM from adenocarcinoma. Because of this method's advantages, it is "more likely to find immediate use in clinical settings," they said.

< Back