By Ben Butkus
Castle Biosciences, a molecular diagnostics firm focused on rare and orphan cancers, has acquired a worldwide, exclusive license from Indiana University Research and Technology Corporation to intellectual property related to a gene expression profiling test for thymoma.
With the license in hand, Castle Bio and IU researchers are developing a multiplex quantitative real-time PCR test to help diagnose and stage thymoma patients. The company expects the assay to be commercially available through its CLIA-certified laboratory by the end of the first quarter, President and CEO Derek Maetzold told PCR Insider this week.
In addition, the company expects to add three or four more molecular diagnostic tests for various orphan cancers by the end of this year, adding to its currently available assays for uveal melanoma, gliomas, and glioblastoma, Maetzold said.
Thymoma is an extremely rare cancer that afflicts just several hundred patients each year, Maetzold estimated. Derived from the thymus gland, thymomas are one of the most common tumors found in the upper chest, but can spread throughout the chest and body. Because of the role of thymus in immune system function, the tumors are also frequently associated with other abnormalities of the immune system such as myasthenia gravis, hypogammaglobulinemia, and anemia.
Current treatment of thymomas consists of surgical resection followed in some cases by radiation or chemotherapy for patients at high risk of metastasis. Unfortunately, histologic assessment of tumor type is of limited value as all types of thymomas can give rise to metastases.
Under the agreement with IURTC, Castle Bio gains access to gene signatures predictive for early-stage thymoma and metastatic risk. A group led by Patrick Loehrer, director of the IU Melvin and Bren Simon Cancer Center in Indianapolis, discovered the signatures.
Loehrer and colleagues first presented data on the signatures at the 2011 American Society of Clinical Oncology meeting. Using microarray analyses, the group identified a 10-gene signature that predicted the absence of metastases in thymoma patients with an error rate of 4 percent; and a set of nine genes that predicted the presence of early stage (stage I or II) disease with an error rate of 9 percent.
According to Maetzold, the IU research group had "already done some lab work to transfer over the high-value targets and discriminating genes, as well as some control candidate genes over to TaqMan-based assays … and began moving it off the discovery platform to the development platform" when they began discussions with Castle Bio to explore commercialization routes.
As a result, the Friendswood, Texas-based company, through its CAP-accredited, CLIA-certified lab in Phoenix, Ariz., began working with the IU team to "go ahead and process all the specimens for the development set, adjust our protocols for the CLIA guidelines, and then go ahead to get a thumbs up for a final test validation set," Maetzold said. "We held off announcing the license because we wanted to make sure we would have something clinically important for these patients, and we're at that point now."
Like Castle Bio's currently available molecular diagnostic assay, the test for thymoma, called DecisionDx-Thymoma, will be based on multiplex qRT-PCR analysis of a certain number of genes. Maetzold declined to disclose exact number of genes on the panel until the final test is launched, but said that it "will end up being similar to [Genomic Health's] OncotypeDx [test for cancer recurrence], probably 12 to 20 [genes] at the end of the day."
Castle will likely offer the test out of its CLIA-certified lab as opposed to making the test available as a kit, because "it's a rare enough cancer [that there] is limited interest for even a low-volume lab to perform it a couple times a month," Maetzold said.
He also noted that DecisionDx-Thymoma will likely be tied to a single qRT-PCR instrument platform, but declined to identify which one due to the fact that Castle Bio researchers are "still working on a couple of variables."
Founded in 2008, Castle Bio has built its business around developing reliable diagnostics for rare and orphan cancers. "My sense is that there are good research dollars in academia toward improving staging, prognosis, recurrence scores, et cetera, in all the big cancers. But there are only a few players that think about [people] that might be afflicted with something not so common," Maetzold said.
Furthermore, in general, "the smaller the cancer is … the less accurate the staging is for prognosis," Maetzold added. "It's really disappointing — you could almost have a coin toss in terms of outcomes, and people think that's effective staging. Telling someone they have a 50 percent chance of living or dying … wouldn't be acceptable in breast cancer. And yet these people who have a rarer cancer are just told it's a 50-50 chance, and that's not right."
Castle Bio already offers DecisionDx-UM for predicting risk of metastasis in uveal melanoma, also known as ocular melanoma; DecisionDx-LGG to determine gradation in gliomas; and DecisionDx-GBM to help determine which glioblastoma patients might best respond to current treatment options.
In October, Castle Bio licensed the BAP1 gene from Washington University in St. Louis, to be used for detecting metastasis in melanoma patients. The company said at the time that it planned to incorporate the gene into its DecisionDx-UM test, and that it has an option to use the gene for other cancer tests.
Maetzold told PCR Insider this week that the company is currently putting the final touches on another licensing agreement with an academic institution that would allow development of an additional cancer diagnostic in the first quarter; and that it is "currently validating two additional tests that we aren't ready to disclose."
"We're also looking at one more from a licensing standpoint from an academic center," Maetzold said. "We've felt from the beginning that if we kept our expenses lean, focused carefully, and did the right job ethically, we would be able to offer a very good service to a small group of doctors and patients to really improve medical care. I think we've been able to do that to varying degrees of success, and this year I think we'll have three or four more tests available by the end of this year, to kind of flesh that out."
Have topics you'd like to see covered in PCR Insider? Contact the editor at bbutkus [at] genomeweb [.] com.