The Merits of Data Sharing in Tackling Rare Diseases

By Shraddha Chakradhar 

May 7, 2012 | BOSTON—Alex Sherman, director of systems in neurology at Massachusetts General Hospital (MGH) presented promising new initiatives in the battle against rare diseases at the Bio-IT World Conference.   

Sherman proposed a model for dealing with the many obstacles that confront research on rare diseases. Sherman’s specific research is with ALS (Amyotrophic Lateral Sclerosis, or Lou Gehrig’s Disease).   

“According to the US government,” Sherman explained, “a rare disease is a disease that less than 200,000 people, or 1 in 1500, have.” By that definition, there are more than 7,000 officially registered rare diseases in the United States.   

The problem for big pharma, as Sherman put it, is “Why bother if there are so many rare diseases with so [few] patients with each one? Pharmaceuticals are not going to bother with them. And no institution has enough patients [with a rare disease] to make any significant discoveries.” Given these confounding factors, Sherman and his team at MGH suggest working collaboratively to get enough data to make significant discoveries.    

“These 7,000 rare diseases compete against each other, compete for the attention of pharmaceutical companies. Instead, however, of encouraging more competition, Sherman suggested collaboration.  When considering a patient, according to Sherman, it’s important to consider everything we might know about the patient, namely, “EMRs (electronic medical records), clinical trials, patient registries, tissue samples, biofluid repositories and so on. All this information is being collected, but not necessarily shared.”      

In order to share and collect information, it is first important to know who owns what piece of information about a patient with a rare disease. EMRs, for example, are owned by institutions, whether hospitals or the institution where a particular trial is being conducted. Similarly, information from clinical trials is owned by the institution conducting them but also by the funding sponsor. This trail of data is important to keep in mind when building a network that encourages collaboration, according to Sherman.   

Sherman’s team runs a consortium of 103 academic institutions involved in ALS research. During the past 12 years, Sherman’s team accumulated data from 1,500 subjects who participated in academic clinical trials. Within a year of starting a new project that appealed to the pharma industry to donate information—the Pooled Resource Open-access ALS Clinical Trials (PRO-ACT)—the consortium gained access to 6,700 more records. While this was encouraging, the biggest and most unexpected challenge Sherman encountered was that of data quality.   

“[Since] pharmaceutical companies are often involved with mergers and acquisitions, many trials were conducted under different companies and [as a result], no one remembers anything about the trial,” he said. “The challenge was [extracting] information from trials when we know nothing about the trial.” Seemingly simple data like gender was often not recorded, compelling Sherman’s team to search for other suggestive data like evidence of ovarian cysts or prostate cancer to point to a specific gender.   

The ALS Consortium with Sherman's help has also set up TREAT ALS (Translational Research Advancing Therapy for ALS), a new platform for clinical research. This is a complex network allowing all ALS partner institutions to be connected, involving everything from EMRs and clinical trial data to biorepositories. The data owned by any one institution are not necessarily freely available to any other institution, but the network offers an opportunity for data to be shared.   

Data sharing required certain standards like common nomenclature and barcode labels to be agreed upon so the issue of data quality, for example, wouldn’t come up again. “It’s like a virtual biobank,” said Sherman. “We created this platform to manage this network. It’s not just data. It’s not just samples. These are institutions; there are people who work for these institutions… We manage the institutions and their members who are involved in the research.”  

“Pharmaceutical companies will only pay attention to you if you’re organized,” he said. He advised that everyone, especially people working in low-profile industries, start thinking along the same lines as ALS researchers. “There are a lot of challenges, but these things work, I can tell you this.”   

As an aside, Sherman also mentioned that having a popular celebrity attached to a particular rare disease didn’t hurt either. “Lou Gehrig did a lot for ALS.  Shostakovich also had ALS, but he wasn’t popular. ” Sherman concluded by saying that when it comes to organization of data, “We are really enablers rather than decision-makers. It’s difficult to be humble, but we should [be].”