In January of this year, the International Committee of Medical Journal Editors (ICMJE) published a proposal for the sharing of clinical trial data. This groundbreaking proposition, if accepted, could upend the industry’s historic tendency toward data hoarding (i.e., sharing just enough clinical data in order to gain regulatory approval). This is because one of the ICMJE conditions not only requires researchers to proactively agree to share deidentified clinical data (as a condition of manuscript consideration by any of the ICMJE member journals), but to do so within six months of article publication.
While many biopharmaceutical insiders would likely agree that such data sharing could have a significant benefit on public health, the pros of such transparency have traditionally been outweighed by the realities of business. For if a biopharmaceutical company is to be sustainable, it requires revenues being generated from innovative and well-protected intellectual property. But as we have seen with a variety of industry initiatives (e.g., TransCelerate), companies are realizing, albeit slowly, that there is much more to be gained from being more open than closed. And though challenges of patient privacy and intellectual property protection for some remain immovable clinical data-sharing obstacles, for others, such as J&J, these are mere hindrances capable of being hurdled. This is why J&J proactively adopted a data-sharing approach long before the ICMJE proposal. “J&J started working on clinical data sharing as far back as 2012,” says the company’s chief medical officer, Joanne Waldstreicher, M.D. So how does such an immense organization develop a feasible data-sharing plan while so many others remain in the early debate phase? Here is the behind-the-scenes look at how one person’s leadership changed the culture about data sharing within her large company — and across the industry.
Creating The Necessary Data- Sharing Infrastructure
A Harvard-trained physician and endocrinologist, Waldstreicher has been in the industry for over 20 years. She says it was four years ago when she saw an opportunity for J&J to “really enhance its work of serving patients and consumers.” “The CEO has these small groups where senior leaders get to spend several days with the CEO,” she explains. “One of the assignments for participants is to write up an idea you want to propose to the CEO.” Being a scientist and a physician, Waldstreicher had an idea that had little to do with the commercial business. “I believed J&J could have a significant impact on public health and better serve patients and consumers by setting up a global science and ethics-based group within the company that was completely independent of R&D, quality, regulatory, commercial, etc.” Though independent, Waldstreicher envisioned the group being responsible for important company efforts, such as the safety of J&J products or bioethical policies across the entire corporation. About a year later, Alex Gorsky, who had taken over as CEO, asked Waldstreicher to develop the chief medical officer role, a new position for J&J. “Having this role laid the groundwork for the culture, mission, and alignment necessary for J&J to pursue a data-sharing initiative,” she asserts.
Once the chief medical officer infrastructure was in place, one of the first stumbling blocks Waldstreicher encountered in pursuit of a data-sharing initiative was internal. “Most people tend to view data sharing as just that, being all about the data, but that’s not a very patient-centered perspective,” Waldstreicher states. “You need to first step back and think about the impact patient data could have when other researchers around the world can access it. Only then do you realize that it’s not just about data, but about advancing knowledge, science, and public health.” In other words, if you want to advance science through data, you first need to have the proper perspective.
When the data-sharing project began at J&J, it consisted of two to three people discussing ideas with Yale School of Medicine (which we will elaborate on in a moment). “Once we realized that we had a real partner in Yale, we began to build a much larger team,” Waldstreicher says. “I don’t know the exact number of people we have, but in building such an initiative we included senior executives from legal, clinical R&D, and government affairs — even our CEO was involved.”
The How And Why Behind J&J's Partnership With Yale
Waldstreicher says J&J had previous experience sharing data (e.g., posting studies to clinicaltrials.gov and sharing via publications). “We have a policy to publish all of our clinical trial patient data. But we realized that if we started sharing summary reports, as well as individual participant-level data, we could have a much bigger impact on public health.” Waldstreicher also knew that, despite J&J’s size, to have the greatest industry influence would require the help of other highly credible organizations. “We decided to work closely with Harlan Krumholz, M.D., a professor at the Yale School of Medicine,” she says.
Krumholz, labeled by some as the most powerful doctor you have probably never heard of, has spent decades shining the light on hospital outcomes research, long before it was fashionable. “We were both at an Institute of Medicine conference where he was a speaker on a panel,” Waldstreicher recalls. “After I heard him speak, I realized we were both working toward the same goal.” Waldstreicher shares that she and Krumholz actually attended medical school together, but until the conference had probably not seen each other for about 30 years. While many view Krumholz as being one of biopharma’s biggest critics (and perhaps someone best to be avoided), Waldstreicher saw the potential of partnering with Krumholz as being the perfect opportunity to work with someone sharing her data-sharing vision. Besides, Yale already had a data-sharing platform in place, the Yale Open Data Access project, or YODA for short (see sidebar).
After the conference, Krumholz and Waldstreicher met to discuss each of their data-sharing visions. From those conversations came their shared guiding principles. Yale and J&J agreed on the following core principles for advancing open science:
“As we began our discussions, we realized that what we thought were big challenges on our independent sides were actually not that difficult to overcome,” she states. For example, one challenge for J&J was how to ensure that only scientifically sound data-request proposals would be approved. “We didn’t feel it would be right for just anyone to be able to look at individual participant data,” Waldstreicher explains. “We felt requesters should have some sort of predefined hypothesis that they were hoping to answer by looking at our clinical data.” Krumholz viewed this as a reasonable concern with a relatively simple solution. He proposed that he and his group would provide the scientific review expertise to ensure all data request proposals were scientifically sound. One of Yale’s concerns was an assumption that J&J would want to have the final say on all data-sharing decisions. “Krumholz felt that for a data-sharing approval mechanism to be effective, it needed to be objective and completely independent from J&J,” Waldstreicher states. “This was a big issue for him.” But J&J was fine with Yale approving the data-sharing requests as long as the review process ensured approval for only scientifically sound data requests.
In 2014, J&J and Yale announced the joint agreement to work together on the YODA data-sharing project. “The first agreement involved data sharing of all J&J’s approved pharmaceutical products,” she states. “A year later, we announced the addition of data from approved medical devices as well.”
It is interesting that prior to the ICMJE proposal the Institute of Medicine published a report in 2015 recommending data sharing and explained how to maximize the benefits and minimize risk. “One of the recommendations of that report was that journals should require authors to commit to data sharing,” Waldstreicher says. As such, not only is the 2016 ICMJE proposal very similar to what the Institute of Medicine produced, but not much different from what J&J has been doing with Yale since 2014.
How Much Does Having A Data-Sharing Platform Cost?
Although she couldn’t assign a specific dollar figure to the cost of this project, Waldstreicher says “it has been costly because we were the first to work with an external independent academic group.” Moving forward, J&J hopes to spread data sharing across the industry and get more companies involved on a broader scale so that the costs come down and become routine.
One of the costs associated with this project pertained to revising processes and forms. “We are changing the template of our consent forms to inform patients of our plans to share clinical trial data in a deidentified manner,” she shares. “While our goal would be to only include patients in our clinical trials who agree to share their deidentified data, there may be times when exceptions need to be made, such as clinical trials involving patients with rare conditions or children.” Dr. Waldstreicher says that if J&J deems the risk of reidentification as being too high, the company reserves the right to not share certain data. “We do not think this will be a common occurrence,” she states. “In fact, we have yet to be faced with such a situation. But as we don’t know what the future will bring; we can’t make a blanket statement that says we will only include patients in our trials who agree to data sharing.” Removing patient identifiers before sharing the data is another cost associated with the project. “We now obtain a signed confidentiality agreement from researchers and scientists who want to access J&J data. The agreement requires commitment to maintain data confidentiality, and to not attempt to reidentify study participants,” she explains. Another cost involves maintaining a secure data-sharing platform. Through YODA, approved external researchers are given access to the data via a platform where the data is housed. This also serves as a means of protecting patients. “External researchers can’t download the data, but have to conduct all of their analyses within the website.”
Waldstreicher admits that the impact of the YODA project leads to new insights that had not been considered when research studies were originally conducted. “The research proposals coming in have ideas and analyses we had never thought about [e.g., gender difference with certain products, comparing J&J data to products and studies conducted by other researchers],” she explains. Perhaps one of the biggest achievements was the actualization of the initiative itself. “Many people didn’t think that what we were trying to do would ever be possible,” she concludes.
YODA: A Powerful Data-Sharing Enabler
J&J wasn’t the first company to partner with Yale University on a data-sharing initiative; the YODA group had previously worked with device company Medtronic to share clinical trial data. In fact, the Yale Open Data Access (YODA) project was actually initiated in 2011, as a “trusted intermediary” approach in which an independent partner provides support, accountability, fairness, and transparency. According to Joanne Waldstreicher, J&J’s chief medical officer, the difference between the Medtronic and J&J agreement is that Medtronic’s was only for sharing data for one product, and it had an end date. “Our initiative with YODA is much bigger and broader, and includes all of our approved pharmaceutical products as well as our medical device products approved after 2014,” she states. “We have committed (for approved products where J&J owns the data) to share data via the YODA mechanism.”
The YODA system for sharing data was designed to ensure cooperation with the data owner while promoting secure, responsible access. To promote independence regarding data-sharing decisions, applicants submit proposals directly to the YODA project, which is responsible for all communication with applicants. Once an application has been submitted, YODA project scientists conduct a blinded review of proposals to ensure that a scientific purpose has been clearly described and that the request will be used to create or enhance scientific knowledge. In addition, a scientific advisory committee is available for resolving any challenging issues. However, the appropriate J&J subsidiary is responsible for conducting the necessary due diligence to ensure that J&J is the legal holder of the data, that the product has been approved, and that the data is electronically accessible. That being said, J&J cannot veto any request proposals. Though Waldstreicher mentioned that the data is not released to investigators, there is a provision that allows data to be released. However, there has to be a compelling justification as to why, and the decision to do so is made jointly between the YODA Project, J&J, and the advisory committee. Between the launch of the YODA Project platform and September 2016, J&J has prepared 132 trials and the YODA Project has approved 49 research proposals. “Yale looks at the proposals very carefully to be sure that every one is a good scientific proposal and that the data we have will answer the proposal’s scientific question,” she reiterates. While Waldstreicher admits that the J&J/YODA model is not perfect, she anticipates the two organizations will continue to learn, evolve, and grow. “There are a lot of different data-sharing approaches being considered (e.g., ACCESS CV, Project Data Sphere),” she states. “One thing we should be doing as an industry, regardless of the approach taken, is convening a gathering of these data-sharing groups on an annual basis so we can share lessons learned and experiences in an effort to continue to advance the science surrounding data sharing.”
Changing The Way Compassionate Use Is Reviewed
In May 2015, J&J made headlines when it announced a pilot program to change its approach to compassionate use, opting to have an independent review panel consider requests for the investigational medicine undergoing clinical testing. Referred to as the Compassionate Use Advisory Committee (CompAC), the group is overseen by Art Caplan, director of the division of medical ethics at NYU, and includes internationally recognized medical experts, bioethicists, and even patient representatives. The goal was not only to eliminate situations such as that encountered by the former CEO of Chimerix, Ken Moch, whose company’s initial refusal to provide access to an experimental medication ignited an aggressive social media campaign, including death threats, but to level the playing field between the haves and the have-nots. “There will be none of this, ‘Call the governor, call your rich brother-in-law’ kind of thing,” Caplan stated upon the announcement of the CompAC pilot.
Compassionate use, also known as expanded access, is a subject that has stymied many pharmaceutical companies. This is because the decision of whether or not to make an investigational drug available to patients that may or may not save their life can also cause a variety of adverse effects. Further, what if patients do get the medication, and yet they do not get the results for which they had hoped? There are valid concerns surrounding how a few well-publicized negative outcomes could prevent a company from getting a potential life-saving investigational medication to market. For more on how J&J “Hopes To Change The Paradigm On Compassionate Use Review,” please refer to the article of the same name written by Ed Miseta in the January 2016 issue of Life Science Leader magazine.