D. Ross Camidge, M.D., Ph.D.
We started screening lung cancer patients for various molecular markers that could predict for benefit from different targeted drugs back in 2008. At that point in time we were probably one of only a handful of sites in the world pursuing that approach. Since then, I am pleased to say it has become the standard of care. But the list of what we test people for is gradually getting longer and longer.
For a period of time I believed if we could only sequence more of the genome we could fill in all of the unknown slices of the pie chart that began to represent lung cancer as not one, but a series of different molecular diseases. But now I know that looking for the classical DNA sequence abnormalities (mutations and gene rearrangements) will only take us so far.
As a physician, everything we learn always comes back to patients. And two patients stand out in my memory in relation to this trial.
The first represents an example of how sequence abnormalities will not be the whole story and, unfortunately, how no one is ever truly protected from this disease. P. was a young man who was a lawyer and lived in the UK. He was a never smoker and only 29 years old when he developed a cough and was eventually diagnosed with advanced stage lung cancer. His family also happened to be extremely wealthy. After initial treatments in London he sought a second opinion from both the University of Colorado and Memorial Sloan Kettering in New York. Together, we analyzed his cancer in every way known to man, next generation sequencing, nanostring technology, multiple FISH tests, etc but we could not find any mutation or gene rearrangement to target. Although we were successful in holding off his disease with various out-of-the-box approaches using standard therapies, eventually this delightful, humble and witty young man succumbed to his disease.
The second represents the hope of developing a clinical trial to test some of the hypotheses we have developed relating to widltype FGFR1 signaling in lung cancer. S. was a good-looking older man, with a twinkle in his eye, despite being well into his 80s. He was one of the volunteers who made coffee and looked after the patients getting chemotherapy in our cancer center. He had smoked 40 or so years beforehand, but so had most of his generation. When he developed advanced lung cancer, I viewed him as just as much of a VIP as the young man described above and we did everything we could to find out what was driving his cancer. The only difference was that we had just developed the preliminary biomarker assays for FGFR1 within our center's Lung Cancer Specialized Program of Research Excellence (SPORE) and used S.'s sample as one of the test cases. S's tumor really lit up when we used the test to see if his cancer was making FGFR1 message. I wish this story had a happy ending but I am sorry to say that it doesn't. Although we had the test up and running, clinical trials take much longer to activate and there was no treatment opportunity at the time. Unfortunately, S. passed on, and is no doubt charming all the young angels he has gone on to meet with the same twinkle in his eye. But his story tells me that there is hope - his cancer was making that message for a reason - I believe it may just be telling us that his cancer, and other cancers like it, if we can find them, are actually dependent on this pathway and a specific drug targeted to these people might make a difference. With the clinical trial described here about to go live, we hope to be able to answer that question over the next few years.
Molecular preselection of lung cancer patients to define biomarkers of ponatinib sensitivity:
Lung cancer represents one of the major global health problems. Fortunately, many significant breakthroughs in the treatment of advanced lung cancer have happened in the last few years. All of these breakthroughs have been based on idenitfying subsets of the disease with different mutations or gene rearrangements in the cancer's DNA that drive sensitivity to specific targeted drugs. These abnormalities are called 'predictive biomarkers'.
Although new mutations/rearrangements continue to be discovered, not all lung cancers will have these simple ‘DNA sequence’ abnormalities, i.e. differences you can detect just by reading the cancer's DNA sequence from left to right. Other kinds of predictive biomarkers will be needed to identify the full spectrum of disease subtypes that are all hiding under the same 'lung cancer' umbrella.
Our preclinical work, looking at cancer cells growing in the laboratory, has shown that a significant proportion of lung cancer is driven by increased signaling from a non-mutated, non-rearranged molecule called the fibroblast growth factor receptor 1 (FGFR1). This increased signaling occurs through the cancer producing more downstream 'message' (mRNA) from the FGFR1 gene, telling the cell to make more of the FGFR1 protein. This can either happen because there are more copies of the FGFR1 gene than normal (FGFR1 gene copy number gain or gene amplification), or because of other changes that increase the amount of message being produced independent of the number of copies of the gene.
We have developed tests that can be done on pre-existing biopsies of patients' tumors to tell if their cancer might be driven by these different ways of activating the FGFR1 pathway. Then, by treating these patients with ponatinib, a drug that interferes with FGFR function, in a clinical trial, we hope to show that we can define new subtypes of lung cancer you can test for, who don't have the classical DNA sequence abnormalities, but who will still respond to specific targeted drugs.
Why is this important?
1. If we can prove our hypothesis that wildtype (non-mutated, non-rearranged) FGFR1, activated through diverse means, is a driver of some lung cancers, by getting patients preselected for evidence of exaggerated wildtype FGFR1 signaling to respond to ponatinib, an FGFR1 inhibitor, we will establish a new field of 'signaling loop' driven cancers that patients can started to be screened for.
2. FGFR1-driven non-small cell lung cancers (NSCLC) may be fairly common (approximately 40% of squamous cancers and 10% of adenocarcinomas) and therefore the potential clinical impact of this approach may be considerable.
3. FGFR1 driven NSCLCs are presumed to be mutually exclusive with other common molecular subtypes of NSCLC, such as EGFR mutant or ALK rearranged disease, but may overlap with a proportion of KRAS mutant cancers – another common molecular subtype of lung cancer. Therefore we may be able to define a subset of KRAS mutant cancers that could benefit from a specific therapy. Given the lack of progress in KRAS mutant lung cancer to date, the idea that subsets of KRAS need to be identified and exploited separately would be a major breakthough.
4. FGFR1 driven cancers may also define a subset of FGFR inhibitor sensitive small cell lung cancer (approximately 5% of small cell lung cancer, perhaps) – a histological subtype of lung cancer that has yet to benefit from a targeted therapy.
5. Finally, independent of the underlying science, the novel clinical trial design involved – whereby patients from across the country can be screened remotely, with their tumor sample coming to the site ahead of them for testing and then they would only need to travel to the center to be treated if they were found to be marker positive – is based on some of the ideas I developed with Jack West from the GRACE wesbite (www.cancergrace.org) in an opinion piece from 2012 entitled, ‘Have Mutation, Will Travel.’ If this approach works, this will also change the future for clinical research in lung and many other cancers, opening up biomarker selected trials to all patients, and not just to those who are lucky enough to have a major cancer center in their backyard. Due to some changes in the protocol language relating to some late side effects of ponatinib that had to be addressed, the start of our trial was delayed. However, as of August 2014, we are open and now have a web-based presence to faciliate the kind of nationwide awareness that is set to change how biomarker-selected trials can be conducted in the future. Check out our trials website - please pass it on to anyone you know who has advanced lung cancer as it may help them.
Who will benefit?
Most immediately, it will be those people who have one of the molecular subtypes of lung cancer that we believe may be targets for ponatinib.
This includes those patients with FGFR1 driven lung cancer (which we estimate to be about 10% of adenocarcinoma of the lung, 40% of squamous cancer of the lung and 5% of small cell lung cancer).
Although FGFR1 driven cancers are presumed to be mutually exclusive with most of the other molecular subtypes of lung cancer that have been described in the last few years, the one excpetion is KRAS mutant lung cancer. This is actually one of the most common mutations in lung cancer but attempts to develop targeted drugs in this population have met with very limited success. We hypothesize that there are in fact relevant subtypes of KRAS mutant lung cancer, for example, FGFR1-driven KRAS mutant lung cancer, that are likely to be sensitive to specific drugs and this trial will be one way of finding them.
In addition, poinatinib is also a potent inhibitor of a molecule called RET, which we will also be screening for and is known to drive about 1% of lung cancers. This clinical trial will therefore offer the chance of benefiting several different molecular subtypes of lung cancer at the same time.