Borderline or Atypical ALK Negative Tumors Deserve Second Look

September 16, 2013

By Megan Brooks

NEW YORK (Reuters Health) Sep 16 - Current criteria for identifying non-small cell lung cancers that might respond to crizotinib might be missing some patients, a new study hints.

The ALK inhibitor crizotinib (Xalkori, Pfizer) has proven highly effective for ALK-positive non-small cell lung cancer (NSCLC).

Now, Dr. Ross Camidge and his colleagues at the University of Colorado Cancer Center in Aurora think borderline and atypical ALK-negative tumors deserve a closer look before automatically deciding against crizotinib therapy.

Currently, the presence of 15% cells with rearrangements identified on break-apart fluorescence in situ hybridization (FISH) classifies NSCLC tumors as anaplastic lymphoma kinase (ALK)-positive.

"When we create any diagnostic test, we have to deal in absolutes - this is positive, this is negative. But biology doesn't deal in absolutes and it occasionally thumbs its nose at that and gives you something that comes right up to the edge of what you are calling negative and now what are you going to do?" Dr. Ross Camidge noted in an interview with Reuters Health.

As reported online September 10 in Cancer, the Colorado team tested 1426 samples of NSCLC clinical specimens, including 174 officially ALK-positive specimens and 1252 officially ALK-negative specimens.

Of the ALK-negative tumors, 121 had greater than 10% ALK-positivity, but were still below the 15% needed to classify the overall tumor as ALK-positive. Overall, 8.5% of NSCLCs were only "borderline" negative, the researchers say.

Dr. Camidge explained that when the diagnostic test for ALK positivity was first developed, "there appeared to be a natural gap between the noise of the assay and a true positive, which gave a cut point of 15%; less than 15% is negative, more than 15% is positive and nobody was actually near that borderline in development. But now we've been testing people for several years, we can look across hundreds of people and we see that about just over 8% of lung cancer patients fall within 5% of that negative cut point."

"I don't know what this means, but it is a sufficiently interesting observation that probably people who are in that borderline negative zone should have a second diagnostic test. We certainly have had a patient who was at 14% and was told he was negative and we retested him and he was 16% positive, went on the drug and responded," Dr. Camidge told Reuters Health.

The researchers have also identified atypical-negative patterns in some tumors, which they think may benefit from a closer re-evaluation of their ALK status. In their sample, for example, increased native ALK copy number (3 or more copies per cell in 40% or more of cells) was detected in 62% of ALK-negative tumors, compared with 19% of ALK-positive tumors.

"In addition to patients who just fail to meet the criteria (the percentage of cells), there are other patients who are nowhere near that borderline, but something else is going on with the ALK gene, perhaps a more complex rearrangement than the classical one," Dr. Camidge said.

"This was the case with a patient in Shanghai. We did a different diagnostic test. It didn't show the classic split signal on the assay, it was more like shuffling the deck rather than cutting it," he explained. The patient was put on crizotinib and is responding well to it, he said.

"The bottom line is, for borderline negative cases or atypical negative cases, you should do another diagnostic test and if positive, give the patient the drug," Dr. Camidge said. The problem now, he said, is that when most assays are done, typically none of the "granular details" are provided.

"It's a bit like the Oscars; you open the envelope and it says you are positive or you are negative, so you don't know if you are 14% or 2%, you're just negative. Physicians need to understand these assays because they are influencing routine care and you need to become best friends with your molecular pathologist and start to understand the results. The molecular pathology reports should probably start to include more granular information, and if a patient is only just negative, you probably should do another test," Dr. Camidge advised.

This research was supported by the University of Colorado Lung Cancer Specialized Programs of Research Excellence (SPORE) and the University of Colorado Cancer Center Shared Resources. Three of the 12 authors have received honoraria, compensation or research grants from Pfizer.


Cancer 2013.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.