August 2010
Feature Story

Karen Titus

Lung cancer testing is a busy intersection these days. Molecular targeted therapies have made it de rigueur for pathologists to differentiate tumor types, including identi­fying EGFR status in patients with non-small-cell lung carcinoma. Phy­sicians are again looking at lung cancer screening, this time using high-resolution CT. And researchers are rattling the conventional wisdom of how lung tumors evolve.

Actually, the scene is more akin to a traffic circle—say, the chaotic Place de L’Étoile, where Paris drivers arriving by cars, trucks, and les motocyclettes along a dozen avenues try to squeeze around one another and the Arc de Triomphe.

Merging onto the scene now: a new serum assay for detecting lung cancer. The test, EarlyCDT-Lung (Oncimmune), measures autoantibodies to six cancer-associated antigens: p53, NY-ESO-1, CAGE, GBU4-5, Annexin 1, and SOX2.

As the name implies, the test, at least initially, is being touted as a way to detect tumors in their early stages, though it detects later stage tumors as well. Its developers are a restless bunch, however, with their eyes on other potential uses: expanding the panel of autoantibodies (it already has two likely candidates for lung cancer); detecting other solid tumors, including breast; monitoring patients with serial serum measurements; and even developing a general screening test to assess risk for multiple cancer types. “Where we’ve started is quite literally the start,” says John Robertson, MD, chief scientific officer, Oncimmune, and professor of surgery, University of Nottingham, United Kingdom.

But for now, the goal is early detection of lung cancer in a high-risk population. For a “lucky minority,” says William C. Wood, MD, the test might detect lung tumors two or three years before they’re visible on a CT scan. “It’s not perfect. But it’s the first apparent breakthrough we’ve seen,” says Dr. Wood, Distinguished Joseph Brown Whitehead professor of surgery, Emory University, Atlanta, and a member of Oncimmune’s scientific advisory board.

Lung cancer is filled with failures. Everyone paints the same ugly picture: It is the leading cause of cancer deaths in the United States and in the world. In this country, lung cancer deaths outpace the next three biggest cancer killers (colon, breast, prostate) combined. Five-year survival among patients with lung cancer, even with the advent of molecular therapies, is poor. In the United Kingdom, Dr. Robertson estimates, 50 percent of patients die within six months of receiving a diagnosis.

The best results still hinge on early intervention. “We’re really quite good at treating tiny lung cancers surgically,” says Dr. Wood.

“That’s the best hope at the moment,” agrees Philip Cagle, MD, director of pulmonary pathology at The Methodist Hospital, Houston, and professor of pathology, Weill Medical College of Cornell University, New York City.

Several other story lines are waiting to be played out, however. New therapies are almost inevitable, Dr. Cagle notes, following on the heels of EGFR inhibitors such as Tarceva and Gefitinib.

CT also appears to be adept at picking up early lung tumors in a high-risk population. The problem, says Dr. Wood, is CT reveals all sorts of lung damage, which affects most smokers and ex-smokers. “You find these little things, and then all you can do is study them again and see if any of them are growing, or if they’re all just scars and granulomas and things like that,” Dr. Wood says.

Because the studies looking at HRCT are ongoing, it’s hard to say with certainty how, or if, it will be a useful screening tool. It appears that HRCT can pick up early tumors, but for now, the impact of doing so is uncertain. “We’re not actually reducing the number of more advanced cancers by this method—as far as we can tell so far,” says Dr. Cagle, who reiterates the data are only preliminary. “I’m sure there’s more information to come.”

The Oncimmune test has taken a somewhat circuitous route into clinical practice. The company introduced it in the United States last spring in an early release program, in part to share with clinicians data that had not yet been published in peer-reviewed journals. (A paper on the technical validation of the assay, with A. Murray as lead author, was published online in the Annals of Oncology, Feb. 2, 2010. The company presented three abstracts on the test at the American Society of Clinical Oncology meeting in June. And a paper on the clinical validation of the test was published online in the Annals of Oncology, July 30, 2010.) This also allowed Oncimmune to work out the logistics of the test, which is run at a CLIA lab the company set up in De Soto, Kan. “It’s not an easy test to perform at this point,” says Dr. Rob­ertson, citing, among other things, its eight-hour run time.

Using autoantibodies as markers to detect cancer is not a completely fresh idea. Dr. Wood notes that decades ago, while at the National Institutes of Health, he developed a microcytotoxicity test that demonstrated that patients had produced antibodies against cancer-associated antigens. “It taught us about biology, but it wasn’t of clinical utility,” he recalls.

For Dr. Robertson, the test represents advances both big and small. He traces his efforts back to 1995. “So it’s taken a long time to develop,” he concedes. On the other hand, this approach is “trying to move things forward with a leap rather than a small step,” he says. “What we’re aiming for, and what I think we’re going to achieve, is worth that long development process.”

The test was targeted at breast cancer initially. But lung cancer emerged as a viable target when Oncimmune joined a European Union collaborative effort to obtain a grant for lung cancer research. The EU colleagues lacked expertise in measuring autoantibody signal and asked Oncimmune for help. “It became clear to us, as we did that, that the signal was actually bigger in lung cancer, easier to detect, than it was in breast cancer,” says Dr. Robertson.

Dr. Wood was not involved in developing the Oncimmune test but was a part of the test’s validation, in part because of his longstanding role heading up the Breast Cancer Intergroup of North America, a network of all National Cancer Institute-sponsored clinical trials cooperative groups in breast cancer research. One of the control tumors for the Oncimmune validation was lung cancer. “It became apparent that we were detecting a fair number of these lung cancers,” says Dr. Wood.

It’s not clear why that might be the case. In lung cancers caused by smoking, the chemical carcinogen often has been associated with more aggressive tumors, at least in lab­or­atory models. “Perhaps the body recognizes it as more abnormal and makes immune responses,” Dr. Robertson says. Too, smoking-related lung cancers are often tied to decades of carcinogen exposure, which could elicit a stronger response.

(Though the lung cancer test has been faster out of the blocks, breast cancer has not been abandoned. The company hopes to have a version available for breast cancer in 2011.)

The Oncimmune test may have the potential to help pathologists further characterize tumors, though admittedly this is a dream for a more distant day. Since the antibodies are specific for specific proteins, “What I think we’re seeing is both an amp­lified signal that something is abnormal in the body, and specific biological information,” says Dr. Robertson. He hopes this could lead to early interventions, especially for breast cancer. Good preventive agents are available, such as tamoxifen and aromatase inhibitors, which might be employed even before a tumor appears on an imaging study. Even longer term, he says, protein profiles might be used to target either vaccine treatments or signal transduction inhibitors aimed at a particular pathway.

In lung cancer, preventive drugs aren’t available. So at this point, the main use of the test would be as an aid for early detection.

“This assay would give us a technique for screening people who are chronic smokers or ex-smokers, and saying, ‘You appear to be at significant risk of having an early lung cancer,’” Dr. Wood says. In this scenario, a positive test would lead to followup with an imaging study and, if needed, a biopsy. The test alone should not lead to intervention, Dr. Robertson cautions. Or, the test could be used in high-risk patients whose CTs have revealed worrisome but ambiguous findings. And for those patients whose serum test indicates the presence of cancer that can’t be found on a followup CT? “You worry about them,” says Dr. Wood. “And maybe get another CT in another year.”

Dr. Wood terms the Oncimmune test a narrowing assay. “It’s about 90 percent specific for lung cancer,” he says, “which is not what you’d like for a screening assay, but for a narrowing assay it’s really powerful. And it seems to pick up about 40 percent of lung cancers,” including stages I and II. As a test approaches 50 percent sensitivity, “That’s psychologically, to me, a big marker,” he says.

“Now, you need to understand two things,” he continues. “One, I treat cancer. And therefore I’m an optimist.” In breast cancer, where most patients are cured, it’s not difficult to be an optimist, he says. “In lung cancer, it’s tougher. The majority of people they treat end up dying pretty quickly. So to be able to move that forward a year or three, and save most patients, like we do in breast cancer, that seems absolutely magnificent.”

That’s the optimist speaking. Now for the realist: “I realize that we need more followup data,” Dr. Wood says. Lung cancer patients need to imitate their breast cancer counterparts, he says, and see their physicians for “their five-year, thank-you-very-much meetings, before we can get as excited as I’d like to be,” he says. So far, that’s not happening.

For all the emphasis on early detection of lung cancer, it’s not a given that earlier will always be better. The interest in early screening for lung cancer has echoes in the 1970s, when chest x-rays and sputum cytology were found to be effective at picking up additional early lung cancers, says Dr. Cagle. But finding them didn’t have an impact on the number of patients with advanced lung cancers, nor did it reduce mortality rates. “So screening for lung cancer was aban­doned because the cost-benefit ratio was skewed toward cost and not toward benefit,” he says. And, as he has already pointed out, preliminary data from the ongoing HRCT studies haven’t drawn a clear line between early detection and improved mortality. “Indeed, if you look at the doubling time of the sort of cancers being detected by HRCT, many of them have a much slower growth rate than we would expect with our traditional understanding of lung cancers.”

“This observation suggests that maybe some lung cancers do not progress, at least not very rapidly, to an advanced stage, like we initially thought,” Dr. Cagle says. Conventional wisdom has held that lung cancers had a natural history that was as predetermined as a serf’s destiny: starting out small, becoming larger, and metastasizing to other parts of the body, in an orderly progress from stage I to IV. But pathologists and clinicians have observed that’s not always the case (see box, “Early detection”). Some small tumors appear to be quite aggressive from the start; others seem to be much less so.

If this indeed proves true, it will be impossible to duck some hard questions. Is it always beneficial to pick up tumors sooner? Will every tumor progress to something lethal? The conversation has become quite convoluted for breast and prostate tumors, and early detection of lung tumors could plunge physicians and patients down the same rabbit hole.

This challenge is fine with Dr. Cagle. Even th­­ough there are more questions than answers now, he’s excited that physicians are finally able to start asking them. For too long, he says, lung cancer has resisted the best efforts of medicine. Now, with targeted therapies and a potentially useful screening assay—either in its current version or its future progeny—it may be possible to start moving in the fast lane.

Related Links

• Early detection