Genomics vs Site of Cancer Origin as Basis for Treatment of Cancer Is ‘False Dichotomy’

Get Permission

Our group, like many others, is in the process of developing a genomics-based clinic…But the idea that organ site–specific therapies are going to magically vanish in the genomic era is just wrong.

—George W. Sledge, Jr, MD

In the News focuses on media reports that your patients may have questions about at their next visit. This continuing column will provide summaries of articles in the popular press that may prompt such questions, as well as comments from colleagues in the field.

Despite recent news reports suggesting that the basis of cancer treatment is shifting from the site of cancer origin to the genetic mutations driving cancer growth, genomics is not altering treatment for most current patients with cancer, nor is the site of cancer likely to become inconsequential.

“I think it is a false dichotomy,” George W. Sledge, Jr, MD, stated in an interview with The ASCO Post. A Past President of ASCO, Dr. Sledge is Co-director of the Breast Cancer Program at Indiana University and Distinguished Professor at Indiana University School of Medicine in Indianapolis. “In the breast cancer clinic here, we have breast surgeons, radiation oncologists, and breast pathologists—we call it a multidisciplinary team, and we think that is a good thing.” In addition, “our group, like many others, is in the process of developing a genomics-based clinic,” he noted. “But the idea that organ site–specific therapies are going to magically vanish in the genomic era is just wrong,” he said.

Context-dependent Biology

“It is wrong for another reason, which is that we see a great deal of context-dependent biology, biology that is fairly organ-specific,” he continued. “For instance, if one looks at epidermal growth factor receptors, the resistance mechanisms in one organ are not infrequently different from resistance mechanisms in another organ. In colorectal cancer, KRAS mutations are very important. But in non–small cell lung cancer, mutations in the epidermal growth factor receptor itself appear to be important. So I think it is wrong to suggest that the organ no longer matters.”

The site of cancer also matters in a patient’s choice of physician and initial treatment. A patient with breast cancer is “still going to have a lumpectomy and radiation therapy,” Dr. Sledge said. “If you have non–small cell lung cancer, you still want to have a lung surgeon operating on you, not a genitourinary surgeon. A lot of what we have done in the past is going to continue.” 

Layer Cake Model

One reason for basing cancer treatment on genetic aberrations rather than the site or type of cancer is that some newer drugs that target genetic mutations are effective in more than one type of cancer, Dr. Sledge said. As an example he cited the ALK inhibitor crizotinib (Xalkori), which was approved by the FDA for treating patients who have lung cancer with an ALK mutation. “ALK is anaplastic lymphoma kinase,” Dr. Sledge explained. “It appears to be very active in this particular type of lymphoma but is probably not going to be very important in many other lymphomas. We are going to see a lot of drugs that will be used in more than one disease, just because the biology will direct us to multiple different diseases or disease subtypes,” he said.

“But that it not to say that the old methods are automatically going away,” Dr. Sledge continued. He said that he viewed the use of targeted therapies based on a patient’s genomic testing “more as a new layer on top of the old, rather than something that is just going to immediately replace the old.” He called this “the layer cake model, where you’ve got the base layer that represents locoregional treatments, on which you’ve got more standard systemic treatments, on top of which you may add more specific molecularly targeted treatments.”

‘Splitters vs Lumpers’

In a Reuters report1 from this year’s ASCO Annual Meeting, Dr. Sledge was quoted as saying, “We’ve had this biological revolution that has sliced the pie for these cancers finer and finer as we’ve learned more about the genomics of cancer.”  He used the example of acute lymphoblastic leukemia, for which at least 17 different subtypes have been recognized.

Asked if there are other types of cancer with so many slices of the pie, Dr. Sledge replied, “It depends on whom you talk to. For instance, breast cancer has what are widely recognized as four or five intrinsic subtypes. But Dr. Jennifer Pietenpol and her colleagues at Vanderbilt University have said that one of those subtypes—triple-negative breast cancer—can be further divided into another six subtypes.2 That’s 15% of breast cancer cases with six different subtypes at a molecular level,” according to that study, Dr. Sledge said.

“At some level, the question becomes, how do you define a subtype? In theory, every patient could represent his or her own subtype, because everyone has different mutations. So you can slice it as finely or as nonfinely as you want, depending on whether you are a ‘lumper’ or a ‘splitter,’” Dr. Sledge explained. “But is it important from a treatment standpoint? A lot of the supersubtypes may not be particularly important from a treatment standpoint, in part, because we haven’t yet identified the driver mutations.” Among those involved in genomics research, “the sense is that there are a huge number of passenger mutations and a relatively fewer number of driver mutations,” Dr. Sledge said.

In addition, a mutation that might not initially be considered a driver mutation may become more important once it is treated with a drug. “There is a fascinating recent article by Dr. Charles Perou’s group at the University of North Carolina,3 which said that if you shut down a particular kinase in triple-negative breast cancer, another kinase immediately pops up as being important. So some things that may not be drivers in the absence of therapy may become drivers in the presence of targeted therapy.”

Less-expensive Sequencing

Targeted treatment and personalized medicine require gene sequencing of the individual’s genome, the tumor, or both. Ultimately, the decision to sequence or not will be driven at least partly by price.

“The price of genomics is falling like a rock,” Dr. Sledge said, adding that the prevailing prediction is “this is the year when we get the $1,000 genome. What we don’t get with the $1,000 genome is the $1,000 explain-the-genome kit,” he noted. “Even though we now can identify all these mutations, we don’t yet have the clinical data to tell us which ones are important.” Those data, he said, should be coming within the next 5 years from studies that are now ongoing. ■

Disclosure: Dr. Sledge reported no potential conflicts of interest.


1. Steenhuysen J: Doctors try to make sense of cancer’s genetic jumble. Thomson Reuters, June 5, 2012. Available at Accessed June 6, 2012.

2. Lehman BD, Bauer JA, Chen X, et al: Identification of human triple-negative breast cancer subtypes and preclinical models for selection of targeted therapy. J Clin Invest 121:2750-2767, 2011.

3. Duncan JS, Whittle MC, Nakamura K, et al: Dynamic reprogramming of the kinome in response to targeted MEK inhibition in triple negative breast cancer. Cell 149:307-321, 2012.

Related Articles

SIDEBAR: Expect Questions from Your Patients

For now, genomic sequencing seems to have a greater presence in the news than in the clinic. What can physicians tell their patients who ask about genetic testing to identify mutations driving cancer growth? 

“The real answer is that when you get down to it, genomics is like any other test,”...