Chasing My Cure

by David Fajgenbaum

  What I quickly found was that my particular subtype of Castleman disease, iMCD, which we now knew accounted for about 50 percent of MCD cases, totaling about one thousand new diagnoses each year in the United States, received little attention or funding compared to even the other subtypes of MCD. In fact, iMCD had received zero dollars in federal funding for research and treatment discovery. And since it occupied a no-man’s-land between autoimmune diseases and cancers, no one knew which of the two categories to place it in. This meant that even private funding sources for cancer and autoimmune research wouldn’t back it. It was the orphan of the orphans.

  Unsurprisingly, no one knew the cause, the immune cell types, or the cellular communication lines involved in iMCD, impeding drug discovery.

  The news got worse.

  Major silos existed between physicians, researchers, and patients with iMCD. Dr. van Rhee had the only lab dedicated to studying iMCD, but there were also researchers in France, Japan, and a few other locations around the United States that would occasionally perform studies of iMCD along with related diseases. However, few of these disparate researchers had ever even met, and none had shared biospecimens or research ideas. This meant that case reports and studies with small patient numbers (therefore inadequately powered to generate meaningful conclusions) were the default. Most blood and tissue biospecimens collected from patients were discarded after testing was performed for patient care. Among the biospecimens that were kept for research, nearly all were sitting in lab freezers in separate locations around the world rather than being studied, let alone studied together. No registries or biobanks existed for storing data or biospecimens in a central place. No large-scale efforts had been made to systematically characterize the clinical and pathological abnormalities of iMCD, something that has been done for nearly all diseases. On top of that, research was completely uncoordinated. There was no agreed-upon framework for how the disease was believed to work to guide research or propose hypothesized mechanisms.

  The only dedicated research effort going on at the time was the development of a so-called panoptic IL-6 blood test. A couple of recent studies had revealed that IL-6 levels were not coming back elevated in many iMCD patients, which led some to believe that the IL-6 tests being used must have been inaccurate. This theory concerned me: Trusting a result only when it’s what you expect and then assuming the test is wrong when you get a result that you do not expect is antithetical to science. Even so, it would be important to determine if these normal IL-6 levels in iMCD patients were accurate. Of course, it could just be that IL-6 wasn’t elevated in those patients and other factors were at play.

  Nevertheless, a few treatments were being used with some anecdotal success, but there were no treatment guidelines and no databases or registries to systematically track and identify the most effective options. No wonder the Duke doctors didn’t know what else to try after their first treatment didn’t immediately work, and even Dr. van Rhee had a limited arsenal. Thanks to the groundbreaking discovery by Kazu Yoshizaki (remember the Japanese doctor who experimented on himself?) of elevated IL-6 in a couple Castleman disease patients in 1989, drugs targeting IL-6 (siltuximab) and the IL-6 receptor (tocilizumab) had been investigated in clinical trials for the treatment of iMCD. They worked in similar ways, so patients who didn’t improve with one were unlikely to improve with the other. Unfortunately, there were no additional drugs or potential targets for treatment being studied other than IL-6 for patients who don’t respond to these drugs.

  It was now also clear why it had been so difficult to diagnose my case. There were no diagnostic criteria for any of the subtypes of Castleman disease. This meant that doctors didn’t have a checklist of tests to run or results to expect to diagnose iMCD. Even worse, many other diseases—like lymphoma, lupus, and mononucleosis—can demonstrate similar signs and symptoms to iMCD. So patients had to just hope that their physicians even knew about iMCD to consider it, hope that their physicians were aware of some of the testing that could be done based on reading medical journal articles, and hope that their physicians knew what to look for in the test results to diagnose it. That’s the wrong kind of hope. That would be like expecting a pilot to successfully fly a plane she had never flown before to a destination she had never traveled to without any map or any instructions. It’s possible she’d figure out takeoff, navigation to the destination, and landing, but she would have no guidelines to increase her chances of getting there safely.

  The more I learned, the more I realized how little the medical community knew about iMCD and how unlikely they were to make any meaningful progress in the near future on the current path. That was intellectually dispiriting—and personally devastating. Not only did Santa Claus not exist, but now it was as though the Santa Claus at the mall that you’d always looked up to was an evil, criminal mastermind. Just not existing would have been enough, thanks.

  And a lack of progress wasn’t for a lack of intelligence or the best of intentions among members of the community. Rather, it became clear to me that we were dealing with a problem of mission and organization. Though the two existing foundations, International Castleman Disease Organization and Castleman’s Awareness & Research Effort, provided lifesaving referrals to a small group of experts and generated important awareness and research funding, neither foundation was playing the quarterback role that is so desperately needed for rare diseases—building the research community, consolidating their knowledge, identifying gaps in understanding, prioritizing the most likely fruitful research, forging collaborations between complementary partners, and ultimately propelling understanding toward a cure. We needed to coalesce into an army and charge forward with a unifying mission. I still believed that it was entirely possible the cure already existed, and just hadn’t been properly marshaled. Perhaps its pieces existed, but in different labs in different parts of the world.

  This effort to cure Castleman disease would be a challenge for a leader as much as for a researcher. I set out to become both. I decided that I wasn’t going to reform any existing structures. Instead, I was going to build something new.

  The first person I wanted to meet with to discuss my plans to take on iMCD was Arthur Rubenstein. He told me I would have his full support, and he’d meet with me as often as was needed. He kept that promise. We’d end up meeting every couple of weeks for the next six years (and counting) to overcome scientific, organizational, and collaborative challenges. Arthur would lean on decades of experience navigating biomedical research and discovery to predict challenges before they arose and advise on preventive measures that we could take. And when unavoidable and seemingly insurmountable challenges did arise, we systematically worked toward solutions and frequently pulled in others for advice. I could never have predicted just how much of an impact Arthur would have, but he has been the most incredible mentor, colleague, and friend.

  I knew from my work on AMF and from Arthur’s sage advice that understanding the lay of the land—the present sad state of iMCD research coordination—was an essential first step. Now, I needed to understand what approaches and steps other disease fields had taken to overcome similar challenges (there are no points for originality in medicine).

  What I discovered wasn’t that much more heartening. Josh Sommer’s work with the Chordoma Foundation was the exception, not the rule. Instead most rare disease research funding organizations raised money, invited researchers to submit research proposals describing how they would use the funds to answer an important research question, and then selected the best proposal for funding. It was random, uncoordinated, and reactive. They hoped that the right researcher with the right skill set and the right biospecimens would apply for funding to conduct the right project at the right time.

  This model works well for the NIH and other large funding bodies, because they get thousands of applications from the best minds in the world, so the top proposed studies are likely to be some of the most im
portant studies for their fields. But most rare diseases don’t have NIH-funded investigators. Instead, foundations focused on a single rare disease may have to select from among only a handful of applicants. And when the number of interested and qualified researchers in an area is limited to just a few, it is unlikely that the applications will propose the most important studies and be submitted by the most qualified researchers. It wasn’t a knock on those researchers; it was just a matter of numbers. This approach was like waiting for the stars to align. When you have tons and tons and tons of stars, they do align every once in a while. But when you have just a few, you’ll be waiting for a long time. Too long, especially if you’re a patient; it felt like a peacetime model with no urgency. And I was at war.

  The other, knock-on effect of this small-pool competition was that each study was often conceived of and developed independently from the others, preventing a coordinated, disease-wide plan from emerging. It also meant that biospecimens and research ideas were often assets for researchers to list in grant applications to distinguish themselves from others rather than shared for collective discovery. If the given grant application wasn’t funded, the biospecimens often remained untouched for the next application. Biospecimens filled with myriad insights to be uncovered often sat on the sidelines, no matter how promising other researchers’ work was. This was just plain wasteful. Not because of anyone’s bad intentions or laziness, but because the major scale model that worked for more common diseases just did not scale for “smaller” diseases. Common diseases don’t necessarily require the same level of cross-institutional collaboration as rare diseases, because single institutions have large enough patient numbers to identify powerful patterns on their own. Not so for rare diseases. Patient samples are inherently scarce. No single researcher can do it alone. There has to be sample sharing to achieve the numbers needed to reach meaningful insights. We needed to rethink the whole operation, the whole field. We didn’t need to build yet another silo.

  I decided that I would take two paths forward: I would conduct laboratory and clinical research into iMCD during my last year of medical school at the University of Pennsylvania and cofound the Castleman Disease Collaborative Network (CDCN) with Dr. van Rhee to accelerate research, diagnostics, and treatment discovery for Castleman disease internationally. Our goal is to identify an effective treatment for every Castleman disease patient. I realize this was quite ambitious. But I did have a secret weapon helping me. Among the three weekly chemotherapies that I was on, two of the drugs were supposed to wipe me out, but one overcame the fatigue and even gave me seemingly endless energy. And a lot of impulsivity. This meant that for twenty-four hours after my weekly chemotherapy day, I didn’t sleep, was hyperalert, and got tons of AMF and CDCN work done, if hastily. Of course, being able to stay awake and alert for that long was no guarantee that I was really operating effectively. As a precaution, Caitlin read drafts of emails I wrote during those bursts of productivity—we wanted to be sure that they weren’t too rambling, too erratic, or uncharacteristically candid.

  When thinking about how the CDCN would accelerate research into how Castleman disease works and how to best diagnose and treat it, I came up with an ambitious but quite simple plan (I thought!): Instead of waiting for the stars to align, we would align the stars.

  First, we’d need to identify, connect, and build a global community to identify and prioritize high-impact research projects. Toward this end, we imagined an online crowdsourcing process that allowed any patient, physician, or researcher to suggest lines of inquiry or studies that seemed most important, regardless of that person’s ability to perform the study. This was a novel approach: The norm in medical research dictates that research ideas are pursued if they’re conceived by a researcher who can also garner the necessary funding and do the work. Promising ideas, if conceived by someone who can’t obtain the funding or perform the work, go unharnessed. We couldn’t afford to let that happen. We also proposed the creation of a scientific advisory board to prioritize the crowdsourced ideas based on their likely impact, feasibility, and rational order within an overarching framework (e.g., you shouldn’t perform in-depth investigations of the inner workings of a particular cell type until after you have identified that the particular cell type is in fact important in Castleman disease). Once we had a list of priority studies, we’d recruit the best researchers in the world for those particular studies to conduct them. Of course and importantly, we’d need to reach out to patients and establish partnerships to gather samples and funding. Upon completion of every study, we would earmark time and resources to the search for medications already FDA-approved for other diseases that might be repurposed to target any of the discoveries made (e.g., cell types, molecules involved, etc.). And when medications were repurposed for use in iMCD, we’d systematically track their effectiveness to guide future use and identify promising candidates for clinical trials. Finally, we understood that feeding information back to the community would be the critical component to keep all of this going—sharing findings would promote continued rounds of crowdsourcing, prioritization, expert recruitment, and execution.

  What we were proposing would maximize efficiency and give the best hope for breakthroughs in the shortest amount of time. Put simply, we’d bring all the key stakeholders together and then apply hyperfocus to research. Instead of soliciting grant applications and hoping that the right researcher would apply to conduct the right research, we’d ensure that the right research is done by the right researcher right now. It’s the difference between a high school announcing that they’re holding tryouts to fill the roster for a new football team and the focused and strategic work that the New England Patriots do to evaluate, identify, and recruit the best players in the world to their roster, who they also ensure will fit together as a team.

  To build a community that could be crowdsourced to prioritize research, I needed to find everyone who had ever written a case report or scientific paper on Castleman disease in the last fifty years. In other words, I needed to search Google and PubMed, the NIH’s repository of medical journal articles. A lot. There were about two thousand medical journal articles tagged with the term Castleman disease. I read every paper, extracted the key data for future analyses, and wrote to every author email address I could find. Each email was cosigned “David Fajgenbaum and Frits van Rhee.” We worked really well together as cofounders, and it felt good to begin to transition into being colleagues instead of connected only through a patient-doctor relationship. I understood how ego-free he was to be willing to work alongside and mentor me, a patient and an aspiring doctor. The dual email signature also gave me some street cred when I was emailing physicians around the world: I was presenting myself as an interested medical student; I didn’t feel comfortable sharing that I was also a patient. A few months and hundreds of emails later, we had connected a virtual community of about three hundred physicians and researchers around the world interested in Castleman disease through an online discussion board and invited them to attend an in-person meeting at the American Society of Hematology convention in December 2012 in Atlanta. ASH, as it’s called in the field, is the largest hematology gathering in the world.

  I couldn’t sleep the night before the meeting. It wasn’t because I had just gotten my oxymoronically energy-boosting chemo cocktail; I was excited and nervous. This meeting would help to establish what was currently known, put forth hypotheses for some of the gaps in knowledge, and the attendees would largely be our pool of potential CDCN scientific advisory board members.

  Thirty-one physicians and researchers attended our meeting at the convention, which meant that it was the largest Castleman disease physician and researcher gathering ever, even larger than the only meeting ever held before, in 2005. And though it was no ESPY awards, for me it was just as exciting: I was starstruck meeting physicians whose names I recognized from my reading. Dr. Kazu Yoshizaki, the self-experimenter who had discovered the link
between IL-6 and iMCD, was there with other colleagues from Japan. I even met Dr. Eric Oksenhendler, the lead author of the 1996 study listed on Wikipedia that had led me to tears when I saw the frighteningly low survival rates. I was also struck by the lack of consensus among the community and the paucity of research efforts.

  A scientific advisory board was assembled soon after the meeting. In my excitement and eagerness for progress, I peppered each of these scientific and clinical heavyweights with endless questions. At least at the beginning—I clearly wore out my intellectual welcome with some of them. One board member gently took me aside and tried to explain that Castleman disease was no one physician’s or researcher’s priority and that I needed to lower my expectations for what everyone could contribute. After a talk I gave during a daylong research meeting at another institution some weeks later—where I was sternly instructed to share my patient experience only, but where I hijacked things and gave a presentation about the CDCN approach and research instead—I learned that some thought I was naïve to think that I or the CDCN could actually make a difference. They thought I was crossing into territory where I didn’t belong. Many felt I should raise funds to support the existing research labs and stick to patient advocacy and awareness. A lab manager for one of those labs even kindly redid my PowerPoint slides before a meeting that showed the CDCN’s multiple roles in connecting everyone. Her version had no network building. No research agenda setting. No collaborations. My relative youth and lack of formal hematology/oncology fellowship training were strikes against my mission it seemed, and the CDCN’s unusual approach to research was deemed almost futile.