I'm very proud to have teamed up with Dr. Ross Camidge from the University of Colorado in writing an editorial piece for the current issue of the Journal of Thoracic Oncology called "Have Mutation, Will Travel: Utilizing Online Patient Communities and New Trial Strategies to Optimize Clinical Research in the Era of Molecularly Diverse Oncology".   This commentary was based on our observation that the world of lung cancer, and oncology in general, is changing quickly, that our longstanding practices for research are poorly equipped to meet the demands of a new world of "molecular oncology", and that we can overcome these challenges by having clinical researchers collaborate far more actively with online patient communities and by revising some of our rather sclerotic (if not ossified) methods for conducting clinical research.

Arguably one of the most exciting and beneficial quantum leaps in the field of oncology has been the recognition that "lung cancer" isn't really a large population of people with NSCLC or SCLC, but rather collections of many smaller subgroups (and the same change has been seen in other cancer types).  A few years ago, the lung cancer community came to realize the importance of tumor histology, but that too probably over-generalizes and is only a weak proxy for what matters far more: the pattern of molecular defects that a particular cancer has.   The Lung Cancer Mutation Consortium described in a post from last year has identified potential "driver mutations" in over half of the lung adenocarcinomas they evaluated, slicing one big pie into many, many pieces, some only 1% of the total.

This is a major advance because we're actually seeing a high probability of dramatic and prolonged responses if we treat the right targeted population with the right targeted therapy, rather than seeing more lackluster results from applying our treatments bluntly to everyone, where we see low response rates and may even miss a valuable signal because the dramatic responses are too infrequent.

But this new molecular oncology world also brings many new challenges.  First, it's not clear that companies happy to develop drugs to be used indiscriminately on large populations will want to develop drugs that work for 1-2% of that population.  We have to hope that the fact that the treatment can be given for a long period of time to well-selected patients, and that the agent can justify a high price because of the dramatic benefit it confers in such people, will make pharma companies confident that they can succeed with a new model that isn't always aiming for a blockbuster.   Pfizer has been successful in developing XALKORI (crizotinib) for the narrow subset of patients with an ALK rearrangement, which amounts to 3-4% of NSCLC, in part because of the high price they can command because it's so effective in these patients, and in part because they can expect to treat these patients for a median of nearly a year instead of a median of 2-4 months, as is the case for conventional chemo or targeted therapies administered in an untargeted way.   (It also doesn't hurt that they have a partnership with Abbott Labs on the diagnostic testing, so Pfizer actually makes money whether a patient does or doesn't have an ALK rearrangement -- the house always wins.)

The development of XALKORI also highlighted a new challenge that Dr. Camidge and I focused on in this editorial: how can companies feasibly run trials for 1-5% populations that are scattered in a geographically dispersed way all over?  Trials a decade ago were run at hundreds of sites at once with designs of "standard chemo with or without new drug X", with eligibility open to anyone with advanced NSCLC. Because of that, any cancer center interested in participating in clinical research could open the trial and expect to enroll many of their patients.  But in the new world of small subgroups, almost all local centers will just have one or a few eligible patients in a year, making it infeasible to run trials that cost just as much to open and run through regulatory hurdles as trials that would have broad eligibility.  So such trials are now only open at a few centers across a country or around the world.   And if these trials are being done on people with investigational molecular markers, how will such patients be identified, if local physicians can't realistically be expected to be aware of every new development in every field of cancer?

To answer the second question first, we feel that the way to ensure that clinical research on small populations is to leverage the motivation and reach of online patient communities like GRACE and others.  In February of 2010, Dr. Camidge did a GRACE webinar on molecular testing and ALK (in molecular oncology terms, almost old enough now to be considered a timeless classic), relatively early in the ALK story and the development of crizotinib, which led to dissemination of this information to patients and caregivers throughout the online world.  With this came greater awareness of the importance of this test, almost certainly with many patients and caregivers bringing it to the attention of their oncologists.  And ultimately, it led to patients coming from all over the US, and even some from outside of the US, to Denver, CO, to participate in the trials that drove the FDA approval and commercial availability of XALKORI.  

 Of course, there were many other patients in many other online communities sharing this kind of critical information, but the key point is that motivated patients and caregivers connected through the internet create a terrific network for sharing important, breaking information. 

The requirement of travel remains a significant hurdle, especially in our current system for clinical trials that requires patients to receive nearly all of their work up for trial eligibility, followed by the vast majority of their care visits, at the center doing the research.  This is clearly a financial and practical burden on patients and their families.   By changing the system to permit far more of the initial evaluation and then routine assessment be done by enabling the local medical team to work with the centralized trial team, potentially also with use of telemedicine follow-up that would allow the patient to remain close to home while still "visiting" with the central trial team, we could facilitate far easier participation by people who are a significant distance from what is likely to be a very limited number of trial sites for these small populations. 

Finally, the FDA will also need to adapt and not expect companies to run randomized phase III clinical trials with hundreds of patients for these targeted therapies in targeted populations.   That may be possible when there are tens of thousands of people who would be candidates, but when your treatment is only applicable for a few thousand people, that's not feasible.  Fortunately, the magnitude of the benefit we hope and expect to see with such therapies is so profound that it is often possible to demonstrate remarkably convincing effects in small populations.  We can only hope that the FDA is able to adapt rather than be enslaved by doing things the way it has always done them.

It remains to be seen whether we can overcome the current regulatory barriers and wariness of companies to change their practices here, but Dr. Camidge and I hope that initiating the conversation and proposing potential solutions will provide a vision of how much better things could be if we capitalize on the interest and networking abilities of the online patient communities for cancer and loosen the regulatory restrictions on telemedicine consultations in clinical research.  I'll also have the opportunity to present an educational session at this year's ASCO on the value of physician engagement with patients in online educational efforts, so this should be another mechanism to reach and convert more people in the oncology world to the cause.