THYMOMA 9-GENE SIGNATURE

A new test that predicts metastases could prove useful in the management of patients with thymoma, a rare epithelial tumor arising in the thymus gland. However, because fewer than 1000 cases of thymoma are diagnosed in the world each year, a prospective study to validate the test could be a long time coming.

The 9-gene signature DecisionDX-Thymoma test, being marketed by Castle Biosciences, is more accurate than traditional staging methods for predicting 5- and 10-year metastasis-free survival, according to senior author Sunil Badve, MD, director of the translational genomics core and director of research immunohistochemistry at Indiana University School of Medicine in Indianapolis, and colleagues.

Their results were published in the July issue of PLOS ONE.

The test, which was developed at Indiana University and is licensed to Castle Biosciences, might help clinicians direct more aggressive treatment to thymoma patients at greatest risk for metastasis and prevent overtreatment of those at lower risk, the researchers note.

Signature Independent of Other Characteristics

"Objective assessment of thymomas based on molecular characteristics can provide independent and reliable data to predict outcomes in patients with thymomas. The predictor is independent of standard predictors such as tumor size, type, stage, and residual disease," Dr. Badve told Medscape Medical News.

The researchers used a training set of archival tissue samples and medical records from 36 primary thymoma patients to test the 9-gene assay. They found that 10-year metastasis-free survival rates were 77% for predicted low-risk cases and 26% for predicted high-risk cases. Ten-year metastasis-free survival rates were 65% for patients with no evidence of disease and 8% for those with residual disease. In comparison, 10-year metastasis-free survival rates predicted using the Masoka stage criteria were 80% for locally confined disease (stages I and II) and 0% for extensive disease (stages III and IV).

The 9-gene test was then validated using 75 archival tissue samples. The test identified 29 patients as being at low risk for metastasis and 46 as being at high risk. Five-year metastasis-free survival was 97% for the low-risk patients, and no additional events were observed at 10 years. Five-year metastasis-free survival was 30% for high-risk patients.

In comparison, 5-year metastasis-free survival rates for the validation set were 49% for Masaoka stages I and II and 41% for stages III and IV. The 5-year metastasis-free survival rates were 56% for patients with resection and no evidence of disease and 29% for those with residual disease.

Some Unexpected Results

"In our original discovery attempt, we thought that molecular data would be able to classify tumors better than histology," Dr. Badve explained. "In contrast to our expectation, we did not identify a distinct molecular classification to standardize the nomenclature for diverse histologic subtypes. However, the molecular information clearly identified genes that determine the metastatic behavior of thymomas." This led to the development and validation of the 9-gene signature.

Dr. Badve noted that most gene signatures are based on proliferation, but the thymoma signature is independent of proliferation, which is consistent with the indolent yet metastatic behavior of thymomas. "Many of the genes within the signature have well-described roles in cancer. We are directing our efforts to use these for targeted therapeutics. However, funding for rare cancers such as thymomas is very meager, and this has significantly impaired our ability to make significant progress," he said.

Prospective Validation Needed

"This study is an interesting first step and is certainly provocative, but this test needs to be validated in a prospective study to influence patient care, and that will be challenging," said Edward Kim, MD, from Department of Solid Tumor Oncology and Investigational Therapeutics at Levine Cancer Institute in Charlotte, North Carolina.

As the researchers conclude, "this type of gene signature is complementary to the clinical information," said Dr. Kim, who was not involved in the study.

"I like the fact that they were able to pull tissue and do an analysis like this comparing the genetic information with clinical information. That's a novel aspect of this study. The results certainly look promising, but a longer process is required before using anything like this in the routine treatment and management of thymomas," Dr. Kim explained.

He pointed out that "at least half a dozen" gene signature assays have been developed for early-stage lung cancer, but no single gene signature is widely accepted for use in determining whether or not chemotherapy should be used, despite the fact that there are more than 200,000 cases diagnosed each year.

"The second step is the tough part with thymomas, which is to test this prospectively. The problem is that this is a rare disease. It is difficult to accrue patients because there are so few of them. For this to become the centerpiece of clinical decision-making in a disease where we haven't moved the treatment forward very much in several decades will be very challenging," he said.

Dr. Kim noted that achieving complete resection remains a strong predictor of outcome and that complete resection is often not achieved in current practice. "To have a gene signature on top of incomplete standard practice is not ideal. We should only be resecting tumors when we can reasonably expect to achieve complete surgical resection," he said.

His advice for clinicians: "If you identify a thymoma, find a group that is experienced in this disease and refer the patient."

The study was funded by Indiana Clinical and Translational Sciences, the Hochberg Family Foundation, Indiana University Melin and Bren Simon Caner Center, and Castle Biosciences. Some of the study authors are employees of Castle Biosciences. Dr. Kim has disclosed no relevant financial relationships.

PLoS One. 2013;8(7):e66047. Abstract