NCCN GUIDELINESS FOR CUP

March 19, 2010 — The addition of new markers that add "a lot more breadth and depth" to immunohistochemical microscopy for the attempted diagnosis of cancer of an unknown primary origin (CUP) is the major change in the National Comprehensive Cancer Network (NCCN)'s guideline for these occult tumors, according to a presenter here at the NCCN 15th Annual Conference.

Immunohistochemical microscopy is "rapidly changing" because of the introduction of new markers, said Charles Handorf, MD, PhD, from the University of Tennessee Cancer Institute in Chattanooga.

"It's hard to keep up with the changes," he admitted, saying that an annual update of the CUP guideline is not enough to adequately chronicle the evolving science. "Every 3 months might not be enough," he said.

Despite the many changes, "the tests only give a probability of what a cancer would be," noted copresenter David Ettinger, MD, from the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins in Baltimore, Maryland.

Molecular profiling for CUPs, which are all metastatic cancers, is also rapidly growing, "maybe even more so" than immunohistochemical testing, said Dr. Handorf.

Although the guideline currently does not include any molecular-profile testing, Dr. Handorf believes that "it will find its niche."

But Dr. Ettinger is skeptical about molecular-profile tests, saying: "If they were good, they'd be in the guideline."

Both Dr. Handorf and Dr. Ettinger encouraged clinicians to do a "solid workup" with patients in order to avoid overreliance on testing.

Dr. Handorf has created his own "law" to reinforce this point. Handorf's law says that "the ultimate diagnostic usefulness of immunohistochemistry in a given case is inversely proportional to how much it is being relied upon to make the diagnosis."

"I will never be famous," Dr. Handorf self-deprecatingly told the audience, "but I might be remembered for this [law]."

All cancers undergo a routine light-microscopic evaluation, a process that costs only $10 to $100, said Dr. Handorf. "About 75% to 85% of cancer diagnoses are confirmed here," he said.

The cases that are not confirmed and, therefore, are CUP need further workup. According to Dr. Handorf, CUP is, in general, divided into 5 classifications: well-differentiated adenocarcinoma (60%); poorly differentiated carcinoma and adenocarcinoma (29%); squamous cell carcinoma (5%); poorly differentiated malignant neoplasm (5%); and neuroendocrine carcinoma (1%).

With regard to further testing, Dr. Handorf reminds residents in his pathology program that immunohistochemical evaluation costs roughly $100 to $1000 and molecular testing costs about $1000+ per test.

"I like to tell my residents that they might be making an $8,000 to $10,000 decision with the stroke of a pen when ordering multiple tests," he noted.

Dr. Ettinger lamented overtesting, especially among newer clinicians. "Younger people have gotten used to using every test under the sun," he said.

Poorly Differentiated Carcinomas and Adenocarcinomas

Immunohistochemistry studies are useful for attempting to characterize the poorly differentiated carcinomas and adenocarcinomas that make up about one third of all CUPs, according to the NCCN guideline.

Immunohistochemical markers are useful for cell differentiation and pathologic diagnosis of CUP.

However, Dr. Handorf had a note of caution here. "The level confusion with routine light microscopy usually translates into similar confusion with immunohistochemistry and molecular studies," he said.

Nonetheless, the testing is an attempt to refine guesswork and select the best possible treatment for patients.

For instance, analyzing the distribution of the markers cytokeratin and keratin is a "trick we use a lot," said Dr. Handorf.

Low-molecular-weight cytokeratins CK7 and CK20 help define subsets of carcinomas. CK7 is mainly found in tumors of the lung, ovary, endometrium, and breast. CK20 is usually expressed in gastrointestinal, urothelial, and Merkel cell tumors. Different distributions of CK7 and CK20, such as when a tumor is positive for one and negative for another, help to refine subsets, said Dr. Handorf.

Still, Dr. Handorf noted that "it is easy to get lost in the woods" with immunohistochemistry because of the array of markers.

The use of molecular markers, or gene-expression profiling, to identify the tissue of origin of a metastatic CUP is in a relatively early stage, suggested Dr. Ettinger. Nonetheless, the tools are formidable; for instance, the Pathwork Tissue of Origin Test (Pathwork Diagnostics) and CupPrint (Agendia) are 1500- and 1900-gene microassays, respectively, according to the slides presented by Dr. Handorf.

"While gene-expression profiling looks promising, prospective clinical trials are necessary," according to the NCCN guideline. Trials would confirm whether patient prognosis could improve with testing and refine the related treatment choices.

Treatment includes surgery, radiation, and chemotherapy. One of the most common chemotherapy regimens used to treat patients with a CUP is paclitaxel/carboplatin, noted Dr. Ettinger.

With regard to molecular testing, inclusion in the NCCN guideline of these kinds of tests will depend on the "relevance demonstrated" by "clinical effectiveness and efficiency" said Dr. Ettinger.

Dr. Ettinger reports having been on the advisory board or speakers bureau or having acted as a consultant to AstraZeneca, Bristol-Myers Squibb, Genentech, and GlaxoSmithKline.

National Comprehensive Cancer Network (NCCN) 15th Annual Conference. Presented March 12, 2010.