Designing Better Diagnostic Tests for Diseases
This is the fifth in a series of posts analyzing Rosetta Genomics stock.
Currently most tests for diseases look at protein (or sometimes mRNA) levels in the cells to determine if the cells are diseased. A good example of this is the prostrate specific antigen (PSA) test for prostrate cancer. The test looks at the amount of PSA in the blood stream and then tries to gauge the likelihood of cancer based on the amount of the protein. The problem with this (and many other) diagnostic tests is that they only look at a single marker and there can be a wide variation between normal patients leading to high level of false positives and false negatives.
Rosetta Genomics is attempting to build a better mouse trap. Since a single miRNA often affects the expression of multiple proteins, measuring the miRNA levels in cells may allow for increased specificity in diseases where miRNAs have been shown to have altered expression. Rosetta Genomics is primarily designing tests to detect cancer and/or predict how aggressive the tumor may be. By combining a panel of several miRNAs whose levels are changed in the diseased tissue relative to the normal tissue, they can increase the specificity of the test.
Prostrate, Lung, and Colorectal Cancer
Rosetta Genomics is working on separate tests for the detection of these three types of cancer. So far, they have found four miRNAs whose expression are different in tumor samples vs normal tissue for each of the three tumor types. When combined, they have a p-values in the 0.002-0.006, so there is a very low chance that the difference in the miRNA expressions are due to chance alone. They don’t say what percent of the tumors that they have tested have these particular expression patterns, so it’s difficult to determine how high the false negative rate will be.
Breast Cancer
Research on diagnostic tests for breast cancer are in early stages. Rosetta Genomics is looking for differentially expressed miRNAs in patient samples that had a high recurrence rate vs those with a low recurrence rate to develop a test to predict if a tumor is likely to come back. As you can imagine, predicting the likelihood of recurrence would be helpful for doctors to determine the best treatment options. Like the above tests they are also testing normal vs tumor tissues to develop an early detection test for breast cancer.
Cancer of Unknown Primary site (CUP)
About 70,000 malignacies a year are diagnosed as metastases of unknown primary tumor (CUP). It’s exactly what it sounds like; the doctor discovers a secondary tumor (metastases), but the primary tumor is unknown. There are a variety of expensive tests that can be run to try to find the primary tumor, but in 70-80% of the cases, the primary tumor is never found.
Since miRNAs have different expression profiles for various cell types, it may be possible to determine the cell type (origin of the primary tumor) by measuring miRNAs levels in the cells of the metastatic tumor. Using 6 different miRNAs they can identify the origin of the primary tumor in 80% of the tumor samples with 90% accuracy. This is based on 6 different cell types, so I think they will probably need to add a few more miRNAs in order to increase the ability of the test to detect other cell types and to increase the accuracy of the test.
None of the tests are in the clinical validation stage yet, but fortunately, the time frame for approval is much shorter for diagnostic tests than for drugs. Rosetta Genomics is actually working on some therapeutic drugs as well, and I’ll get to those next time.
Filed under: Rosetta Genomics (ROSG) | 2 Comments »
Does anyone have any idea why a company without any products to sell might be using Google adwords to advertise? I saw the image to the right on my blog the other day.