GRACE :: Cancer Basics
Dr West

Cancer Ouija Boards, Umbrellas, and Baskets: The Evolution of Genomic Oncology

Share
download as a pdf file Download PDF of this page

Cancer treatment is in the midst of a transformation in real time.  Genomic testing of a tumor– looking for a wide range of dozens to potentially hundreds of markers at a time — is moving quickly from bleeding edge to mass adoption, at least in the US. This change is partly driven by ever-changing data and ever-changing clinical experience, partly driven by the general promise felt by patients and clinicians alike that new information will lead to vast improvements in our understanding and therapeutic options, and (lest we be naïve) partly driven by marketing from institutions and diagnostics companies who stand to gain by promoting this work.

That there are potential gains is undeniable – regardless of what the future may bring, even today it is a tangible gain to avoid missing the immediately actionable findings such as an EGFR mutation (for someone with  non-small cell lung cancer (NSCLC), for instance), but it can find many less common but clearly “actionable” mutations ranging from HER-2/neu to BRAF or a few others that are now mentioned in the guidelines developed by the National Comprehensive Cancer Network (NCCN) that typically lead to insurer coverage of the treatments recognized as effective for these rare mutations, which range from <1% to 3-4% of the lung cancer population.

But these tests are not going to offer only unmitigated positive opportunities. Aside from the cost of several thousand dollars per tumor profile performed, the results of these profiling tests most often reveal not a clearly actionable mutation, but one or more rare mutations that are accompanied by a synopsis of lab-based suggestions for unapproved and clinically untested options in that particular tumor type from the testing company. While a patient and their oncologist may say that they will ignore treatment options that are poorly studied and essentially just wildly speculative (there is a rather weak correlation between cancer treatments that work in the lab and those that are safe and clearly active in human cancer patients), that’s easier said than done. Instead, the molecular results often lead oncologists to be tempted to practice the black art of using the profile as a “medical Ouija board” to cobble together a treatment plan with no good clinical evidence to support it, all too often bypassing the treatments that are well established as helping improve treatment options in thousands of cancer patients with that tumor type. 

Ouija Board

This isn’t a paranoid fear. This happens every single day, and the potential harm includes the cost of these unproven therapies (which now commonly exceed $8000-10,000/month), the side effects of these treatments, and the opportunity cost of being led down a rabbit hole and bypassing the treatment that much stronger evidence should have supported but which was ignored in favor of the siren’s song of the rather fanciful collection of vaguely suggested treatment options for the wide range of poorly understood mutations.

(Apologies for mixing metaphors with abandon. It’s my Achilles heel. <- irony)

But even I, who might well be considered a curmudgeon about health care costs and policy, feel optimistic that there’s a path out of this muck. There are ways to methodically collect the data and learn from our experience.  More and more cancer centers are developing databases that will store the data on the molecular markers found, along with clinical characteristics of the patient and their cancer’s response to various treatments. Data mining software should realistically be able to help us detect patterns in the vast abyss of data. And in the shorter term, we’re also a seeing a few new formats of clinical trials have emerged that are incredibly well-suited for the era of molecular profiling.

The first of these is an “umbrella trial”, which takes patients with the same tumor type, such as squamous NSCLC, has them all undergo molecular marker testing, and then assigns patients to one arm or another based on the presence of one potential target or another, with an arm as well for the patients with no identifiable marker. A leading example of this is the “Master Protocol” for second line treatment of patients with squamous NSCLC, in which patients all undergo broad genomic sequencing and are assigned to one treatment arm or another, including the largest arm of all for those patients with no matching marker. In every arm, patients are randomized to standard treatment with Taxotere (chemotherapy) vs. an investigational therapy for the target in question (or an immune checkpoint inhibitor as the investigational therapy for the “no marker” arm):

 LungMap schema

One advantage of this protocol design is that the arms can be designed to move efficiently toward FDA approval, and each arm can open or close independently, making it easy to modify over time without having to hold up the trial for ongoing regulatory changes.

An alternative approach is a “basket trial” in which patients with a specific marker are enrolled, regardless of the type of cancer that they have, to receive a new therapy or perhaps be randomized between standard treatment or novel agent. For instance, a trial with a novel HER2 inhibitor might be opened for patient with a HER2 abnormality whether they have breast cancer, gastric cancer, lung cancer, or any other type of cancer.

To me, these types of trials have many advantages. Most importantly, they provide an answer to the question of “what can we do for someone when the results come back, in the likely event there isn’t a currently actionable mutation?”.  These trials are intended to be broadly available to patients in a broad range of cancer centers, making it possible to bring the most novel treatments, even for rarer cancers, to the patients where they live, rather than requiring patients to travel to a few isolated centers that offer a trial for a rare mutation.  This means that these trials are likely to be completed faster, potentially leading to very efficient (and hopefully less expensive) drug development processes (i.e., getting drugs approved and commercially available for the right patients sooner).  Finally, I also think it’s very likely that if we capture the data to correlate clinical characteristics and patient outcomes with an ever growing array of rare mutations, we’re likely to identify serendipitous findings much more readily than we do now.

The stakes are large with the “big data” that comes from our transition to broad molecular testing of cancers.  If we do it wrong, primarily treating patients outside of trials based on subtle suggestions of possible utility of expensive and potentially toxic treatments based on inferences from animal or test-tube studies, then not collecting data in any systematic fashion, we may well end up just spending more money and treating patients less effectively from this “buckshot” approach. But if we do more novel trials, especially these umbrella and basket trials developed for new era of genomic oncology, we can potentially test new treatments even for relatively rare subgroups close to where patients live and develop new therapies far more efficiently. 


Leave a Reply

Ask Us, Q&A
Cancer Basics Expert Content

Archives

Share
download as a pdf file Download PDF of this page

GRACE Cancer Video Library - Lung Cancer Videos

 

2015_Immunotherapy_Forum_Videos

 

2015 Acquired Resistance in Lung Cancer Patient Forum Videos

Share
download as a pdf file Download PDF of this page

Join the GRACE Faculty

Lung/Thoracic Cancer Blog
Breast Cancer Blog
Pancreatic Cancer Blog
Bladder Cancer Blog
Head/Neck Cancer Blog
Kidney Cancer Blog
Share
download as a pdf file Download PDF of this page

Subscribe to the GRACEcast Podcast on iTunes

Share
download as a pdf file Download PDF of this page

Email Newsletter icon, E-mail Newsletter icon, Email List icon, E-mail List icon

Subscribe to
GRACE Notes
   (Free Newsletter)

Other Resources

Share
download as a pdf file Download PDF of this page

ClinicalTrials.gov


Biomedical Learning Institute

peerview_institute_logo_243