GENE SIGNATURE TO IDENTIFY PRIMARY SITE OF CUP

January 27, 2011 — Real-world data have confirmed the effectiveness of a test designed to help identify the primary tumor site in patients with difficult-to-diagnose cancers. After using the novel gene-expression-profile assay, the determination of the primary diagnosis site and treatment management was altered in more than half of the patients in a new study.
After receiving results from the Tissue of Origin Test (Pathworks Diagnostics), the tumor site diagnosis was changed in 54% of patients (P < .0001). In addition, cancer-specific management changed in 68% of patients, and the recommended chemotherapy regimen was altered in 54% (P < .0001 for both).
"In about 3% to 5% of all new cancer cases, it is difficult to determine the primary tumor site, even after imaging and other pathologic tests," said W. David Henner, MD, PhD, chief medical officer at Pathworks Diagnostics. "Pathologists have been able to give a tentative identification in many of these cases, and that can be useful, but until very recently, there weren't any more definitive ways to identify the primary site."
"Technology now allows the primary site to be more definitively identified," he told Medscape Medical News. "It is becoming more and more important to identify the primary site because therapies have . . . become more targeted."
The study results were presented at the 2011 Gastrointestinal Cancers Symposium recently held in San Francisco, California, which was cosponsored by several organizations, including the American Society of Clinical Oncology.
Difficulty in Diagnosis
The Tissue of Origin Test is a microarray-based gene-expression assay that compares the RNA profile of a formalin-fixed, paraffin-embedded specimen to established RNA profiles of 15 known cancer types representing 58 morphologies. The test has been cleared by the US Food and Drug Administration and measures the gene-expression levels in more than 2000 genes.
Diagnosing a primary tumor site in some cases can involve a large number of various tests, including immunohistochemistry (IHC) panels, serum markers, and imaging studies. However, current histologic and imaging techniques often fail to yield a definitive identification for a significant number of patients. Lead author John Hornberger, MD, CEO and president of Cedar Associates LLC, believes that clinicians are interested in having more reliable methods of determining primary tumor site.
"Going through the literature, we found that there were 147 types of IHCs," he said. "With that many options, there is going to be variability among pathologists — what they order and what they don't order."
However, even with all of the current tests, there are going to be a fair number of patients for whom a diagnosis remains elusive, Dr. Hornberger said in an interview.
Previous studies have demonstrated the reliability and high performance of the test. One validation study, for example, showed that the overall agreement with the reference diagnosis was 89% (95% confidence interval, 85% to 91%), and an average of 12 tissues for each specimen could be ruled out with more than 99% probability. In addition, a multisite reproducibility analysis showed 89.3% concordance between laboratories (J Mol Diagn. 2011;13:48-56).
Real-World Results
In this study, the results changed therapy in more than two thirds of cases, said Dr. Hornberger, and it wasn't just chemotherapy regimens. "Some patients underwent additional surgeries, radiation, and referrals."
He also noted that they asked physicians if they found this test clinically useful, and two thirds of them stated that they did. "This is a test that provides information that these doctors thought was changing their diagnoses and their treatment management decisions."
The objective behind this study was to evaluate the clinical utility of the assay and its real-world effect on diagnosis and patient management. Specifically, Dr. Hornberger and colleagues assessed how the results of the Tissue of Origin Test influenced tumor site diagnosis and subsequent management decisions when used as an adjunct to clinicopathological evaluation.
The retrospective observational registry study comprised 66 physicians who used the test between July and December 2009; completed surveys were available for 111 patients.
Most of the patients had undergone a large number of imaging tests and endoscopic procedures (average, 3.2 tests per patient) prior to the Tissue of Origin Test. Tumor biopsies had also been extensively evaluated with IHC (mean, 9.9 stains per patient).
For this cohort, physicians reported a working diagnosis for 56% of patients prior to using the assay; this rose to 86% after they received the test results. In addition, the number of not-yet-determined diagnoses dropped by 2.5-fold (P = .0001), leaving only 15 patients without a primary site diagnosis.
Outcome Data?
"The use of molecular profiling in cancers of uncertain primary site or origin is potentially very useful," said F. Anthony Greco, MD, but added that outcome data are needed.
"We need more direct data regarding the outcome of patients who receive therapy based on the Tissue of Origin Test or other similar commercially available molecular-profile assays," said Dr. Greco, director of Sarah Cannon Cancer Center in Nashville, Tennessee, who was not involved in the study.
He added that "the real-world setting needs documentation of the accuracy of these tests and improved patient outcomes."
However, another expert pointed out that it might be difficult to conduct an outcomes study.
Conducting an outcomes study would be the ideal scenario, but the practicality of it is unclear, explained Federico A. Monzon, MD, director of Research Pathology and Molecular Diagnostics Cores, The Methodist Hospital Research Institute, Houston, Texas.
"We don't have a homogenous population, which means a lot of confounding variables, such as stage, type of tumor, extent of metastasis, and treatment regimen," he told Medscape Medical News.
There would also be ethical considerations, Dr. Monzon pointed out, if a test that could facilitate diagnosis was withheld from a control group.
I think we are going to be left mainly with retrospective studies, he said, "comparing patients who were treated before the era of molecular identification of tissue of origin and those who were treated afterward."
This study focuses on the clinical impact of this test, he noted. "It shows that physicians are finding the information useful and are making clinical decisions based on it. The test is providing important information to physicians that is otherwise not readily available."
It is important to remember that it has been demonstrated by other researchers that if the tissue of origin can be identified, outcomes can often be improved, he explained. Current treatment guidelines, such as those issued by the National Comprehensive Cancer Network, focus on treating each patient on the basis of the most likely site of the tumor of origin.
"We are pretty much convinced that identifying the tumor of origin will result in determining the most appropriate therapy for each patient," he said. "We should expect that this will lead to better outcomes, although that was not shown in this study."
However, there have been studies that have shown that if the tissue is molecularly identified, outcomes are better. "This has been seen particularly in colon cancer," he said. "It is a rough extrapolation to all the tissue types, but there is literature that supports the idea that better identification of the tissue of origin leads to better outcomes."
The study was funded by Pathworks Diagnostics. Dr. Monzon reports receiving honoraria from Pathworks Diagnostics for speaking engagements related to the Tissue of Origin Test.
2011 Gastrointestinal Cancers Symposium (GICS): Abstract 459. Presented January 22, 2011.