American Association for Clinical Chemistry
Better health through laboratory medicine
October 12, 2006

In this Issue...

Predicting the Return of Early Stage Lung Cancer
by Julie McDowell


Patients in the earliest stage of non-small-cell lung cancer (NSCLC) undergo surgery to remove the tumor, but are not treated with chemotherapy. For many of these patients, the cancer will return, and they could benefit from chemotherapy treatment. But most health insurance plans—including Medicare and Medicaid—don’t cover chemotherapy for these patients. This month's Strategies looks at a diagnostic tool developed by Duke University researchers that assesses the recurrence of early stage lung cancer, therefore identifying ideal candidates for chemotherapy, and how the researchers are working to get both the test and the treatment covered by insurers.

Among different kinds of cancer, lung cancer is the leading cause of death among men and women in the U.S. NSCLC cancer accounts for 80% of these deaths. Patients with NSCLC are staged according to their tumor size into I, IB, IIA, IIB, or IIIA. Under current treatment standards and guidelines, those with the smallest tumors, stage I, have their tumor removed, but 30%–35% will have a relapse, making them good candidates for chemotherapy.

Researchers at Duke University recently developed a test that predicts stage I NSCLC patients will have a recurrence [New England Journal of Medicine (2006;355:570-80)]. “The current staging for early-stage lung cancer is very crude,” said Anil Potti, MD, a lead author of the study and Assistant Professor of Molecular Genetics at the Institute for Genome Sciences and Policy at Duke University in Durham, N.C. “It’s largely dependent on tumor size. If a patient’s tumor is less than 3 cm, they have surgery. But if the tumor is greater than 3 cm, the patient would have surgery followed by chemotherapy. It’s obvious that we need to have something more to look at than tumor size, but unfortunately, there hasn’t been any pathologic feature that would significantly differentiate patients with early stage lung cancer.”

In the study funded by the National Institutes of Health, Potti and his colleagues looked at gene expression patterns of 198 tumor samples. These samples came from patients in three cohorts: the Duke Lung Cancer Prognostic Laboratory, the American College of Surgeons Oncology Group (ACSOG), and the prospective Cancer and Leukemia Group B (CALGB) trial. Total RNA was extracted from the tumors, and then analyzed using the GeneChip, manufactured by Affymetrix (Santa Clara, Calif.). The researchers then generated a recurrence risk number for each patient through statistical analysis of gene expression patterns, as well as tumor diameter, age, sex, smoking history, and cancer subtype.

The findings from the Duke study showed that the test was able to predict recurrence better than clinical prognostic factors and more consistently across all early stages of NSCLC. The statistical model had a predictive accuracy of 72% for the ACOSOG trial, and 79% from the CALGB trial. The test also predicted with up to 90% accuracy which stage I NSCLC patients would have a lung cancer recurrence. “Those patients would be good candidates for chemotherapy,” said Potti.

New Regulatory Hurdles

But if the Duke team wants to put their research findings into clinical practice, they must first submit the appropriate regulatory paperwork to the Food and Drug Administration (FDA). This is a laboratory-developed, genomic test that analyzes multiple markers, and the FDA issued draft guidance in early September that regulates such tests in a new category called in vitro multivariate index assays (MIAs). In addition to analyzing multiple markers, these complex test systems may also use software-driven algorithms to provide diagnostic information or to guide treatment decisions. Although the new test is a home brew, it is now subject to tighter regulation by the FDA compared to other laboratory-developed tests according to recent draft guidance.

In fact, the Duke researchers have already met with the FDA about their lung cancer diagnostic test and have filed preliminary paperwork to begin the regulatory process. If the test is approved by the FDA, Potti hopes that reimbursement for the test will follow, as well coverage for chemotherapy for the stage I NSCLC patients identified at high risk for recurrence.

“The right way to do this is to get FDA approval upfront,” said Potti. The researchers have already been meeting with CMS and private health insurance providers about what data they need to cover both the test and the chemotherapy. “The first step is to get them to cover the tests, and then once the test is covered, I don’t think getting the chemotherapy covered will be a big deal," he added.

With these approval and coverage goals in mind, the researchers are designing their upcoming phase 3 clinical trial study around what data the FDA, CMS, and private insurance companies want to see. The researchers are currently recruiting 1,200 patients for this trial, which is scheduled to begin in Spring 2007. The trial is funded by the National Cancer Institute, and will take place at 50–60 sites in the U.S. and Canada.

“At this point, what I think the insurance providers are looking for is data validated in multiple settings,” said Potti. “More importantly, data that is validated in cohort groups and prospective settings."

For more information on the FDA Guidance:

  • A full copy of the recent FDA draft guidance, “Draft Guidance for Industry, Clinical Laboratories, and FDA Staff: In Vitro Diagnostic Multivariate Index Assays”, is available online at www.fda.gov/cdrh/oivd/guidance/1610.pdf.
  • The November 2006 issue of Clinical Laboratory News will have a feature on this recent draft guidance for multivariate index assays.