Only a few years ago, oncologists saw lung cancer as divisible into small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC), with very little relevance to any division beyond that point. We knew that patients with NSCLC could be categorized into having adenocarcinoma (including the subset of that larger group, bronchioloalveolar carcinoma (BAC), squamous, large cell and large cell neuroendocrine carcinoma, and into those with and without a significant smoking history, and other demographic variables, but they weren't a primary focus when we didn't perceive that these differences had clinical significance in our treatment recommendations. And at that time, most people received remarkably similar treatments, with first line trials in advanced NSCLC incredibly commonly designed as carbo/Taxol (paclitaxel) +/- new drug, and second line trials of Taxotere (docetaxel) +/- new drug, or a direct comparison to placebo in previously treated patients.

Things have changed dramatically in the lung cancer field over the past few years, and in most ways definitely for the better. We now see more patients being cured with earlier stage disease and living years with advanced NSCLC, even if longer term results we might call cures remain elusive. Many of these improvements have come in the setting of a new era in which we categorize NSCLC patients by histology, sometimes smoking status and patient race, and increasingly by presence or absence of certain molecular features (we still aren't sophisticated enough to have learned how to subcategorize groups within SCLC). We now have different treatment pathways for squamous vs. non-squamous NSCLC, Avastin (bevacizumab) eligible vs. ineligible (for a variety of reasons), and now pathways directed by the presence or absence of molecular markers like EGFR mutations, ALK rearrangements, and potentially K-RAS and other biomarkers like ERCC-1, RRM-1, etc. And by individualizing to deliver the best treatments for more narrowed populations, we've incrementally improved survival for many of these subsets.

But there's been a challenging by-product of this process, which is that it's becoming harder to do clinical trials in which the target population is much smaller than it used to be when drugs were invariably approved for all patients with advanced NSCLC. Pfizer's trials with crizotinib need to screen thousands and thousands of patients to identify a few hundred with an ALK rearrangement. Meanwhile, progress on more limited groups of patients, such as those with BAC or large cell neuroendocrine carcinomas, are extremely understudied because it's infeasible to reach such a limited population of patients who are geographically dispersed. We're now identifying new molecularly defined small populations, such as those with a K-RAS mutation and those with an EGFR mutation who develop an acquired mutation after a great response.

Unfortunately, pharma/biotech companies can't and won't run trials for small target populations (markets) and that require a decade to enroll trials. And research centers can't afford to run trials in which each center only sees 1-2 patients per year who would be eligible.

But the internet provides a remarkable way for people who are geographically dispersed and share an infrequent but common interest to connect with each other and work collaboratively. There may not be enough die-hard fans of Buffy the Vampire Slayer to have groups meet together in person, but there is a vibrant online community. More importantly, people with rarer medical issues can galvanize forces and help neutralize their decentralized status, sharing information and pooling their efforts to move research forward to help themselves.

GRACE member linda22 has BAC and is part of a self-aggregating group with BAC who have found each other online. They're now working together to create banks of tissue of BAC tumors and, hopefully, to engage in clinical trials together. I think that if people with EGFR and KRAS mutations, ALK rearrangements, large cell neuroendocrine tumors, perhaps even squamous NSCLC patients, or any other clinically relevant subsets can find each other and become an easily identified group for clinical trials, investigators could better study these groups we've always regrettable said were "too rare to study";pharmaceutical companies could take a chance on studying a drug in a clinical indication with far more confidence that they could complete these trials in a timely way. And patients and family members could play an active role in accelerating the rate of clinical progress in cancer, so that we no longer need to say that a cancer subgroup is too small to study.

I'm starting a new discussion forum thread in the lung cancer section on lung cancer subgroups (start a new group by clicking this link), hoping to facilitate that self-aggregation of people into groups that could facilitate more investigators investing effort and pharmaceutical companies investing money in new avenues of research. Patients could play a more active role than ever before in overcoming the bottlenecks that occur as we define more sub-specialized clinical populations.