By Mark D. Uehling
August 15, 2003 | Could standards work for a secretive, litigious, and cutthroat business like the pharmaceutical industry? Perhaps, but negotiating a consensus among pharmaceutical companies is as easy as putting lipstick on chickens.
Just don't tell the people working on SEBiX, CDISC, XML, and the eCTD. These evolving life science standards are gaining traction. Their partisans have been seeing a lot of red-lipped birds lately, especially at the Drug Information Association's (DIA) annual meeting*, which drew more than 6,000 attendees.
SEBiX and XML seem particularly promising, holding out the prospect of smaller IT staffs and less dependence on custom solutions. SEBiX -- the Secure Electronic Biopharmaceutical Information eXchange -- is a Web-based, vendor-neutral clearinghouse for biomedical information. It can be thought of as a New York Stock Exchange for pharmaceutical data -- a place where frequently traded information would be stored and distributed securely, whether to governmental regulators or to authorized commercial parties.
Pfizer's executive director of worldwide regulatory affairs, William Rosen, is SEBiX's leading evangelist. "We have a need for global research collaboration while research costs are skyrocketing," Rosen said. "SEBiX will provide an opportunity to reduce development costs. It delivers both tactical and strategic benefits. It will reduce dependence on proprietary networks and technology solutions."
In a pilot proof-of-concept project finished just before DIA, Rosen and confederates at several major pharmaceutical companies showed that the FDA’s Form 1572 -- more than 150,000 of which are filed to Washington every year -- could be expeditiously handled by a vendor-neutral clearinghouse. IntraLinks and SAIC helped Rosen deliver the goods, but other vendors will be able to participate going forward. "The FDA is on board," Rosen said. "Pharma has made it a top priority."
GlaxoSmithKline's Michael Brennan was even more effusive about SEBiX. Group director of submissions strategy and new product support, Brennan enthused about the esprit de corps of the volunteers on the Form 1572 project. "This concept is ready for real time. We have to find ways of communicating and exchanging information with each other," he said, noting that 60 percent to 70 percent of drugs are now "in-licensed," or discovered by smaller companies. "Not everything is controlled in our shop."
'Out of the IT Business'?
The foundations of the Henry B. Gonzalez Convention Center then trembled slightly as Brennan said: "We need to get out of the IT business. We need to use the information we’re collecting rather than finding ways of enhancing it. We all need the same tools."
XML is hardly a new idea, but it kept popping up during DIA. XML is a key part of the common technical document, an international structure for regulatory filings; and of CDISC (Clinical Data Interchange Standards Consortium), the standard for clinical trial data that is rapidly being synchronized with HL7 standards. Finally, XML is gaining prominence as an efficiency tool.
A consultant, Walter Sanders of TCA, said that XML could also be a win-win situation for regulatory agencies and companies that interact with them. He noted that after one large drug company began handling adverse event forms with XML, it was able to process 22 percent more forms with 15 percent fewer personnel.
Several sessions concerned clinical trials and electronic data capture (EDC). Should Amnesty International ever need a reliable witness to wanton crimes against clinical data, M. Joseph Mardesich of Quality and Compliance Consulting could testify. One horrific tale of patient databases gone bad ended with the terrifying words: "Everything was lost."
However, most of the news was better. Eli Lilly's Rekha Garg, senior clinical research physician on the product team for Xigris, presented one of the sanest prescriptions for clinical trials in years. Xigris was conditionally approved in November 2001 by the FDA, with Lilly promising to conduct one last trial. That entailed enrolling 11,444 sepsis patients, an increase of more than 570 percent from the company's previous clinical trial. The data began rolling in September 2002 and will be submitted to the government by December 2005. Such an audacious timetable is especially impressive for a trial with more than 10,000 critically ill patients.
But Garg is not worried. Lilly's second trial has been designed to collect only crucial data -- facts that will help Lilly and the FDA identify major issues. Notes Garg: "These patients are in the intensive care unit. They’re extremely sick." This is not to say adverse events won't be tracked, just not in all patients throughout the trial.
During its first sepsis trial, Lilly gathered almost 250 ml of blood from each patient for lab tests, but it needs only 45 ml this time. Fewer lab tests generate fewer lab reports, reducing data-processing costs. Source document verification -- matching the trial data to those in the medical system -- has also been reduced.
On the IT side, the trial also features a minimalist case report form (CRF). "We challenged our data managers," Garg reported, shrinking a 69-page CRF to 25 pages, much to the relief of nurses and clerks at the clinical sites. "They could enter the data in a much easier form than they had in the past," she said.
Garg conceded that the process was difficult to advance within Lilly. "There was a substantial amount of pushback," she said. "But we have to move forward. We can’t continue at this cost." If Lilly had tried to do a full-blown clinical trial in the usual way, she said, it might have cost $100 million. Interestingly, however, her data will not be collected using an EDC system. For reasons of its own, Lilly has chosen to document its second Xigris trial on paper.
*DIA’s 39th annual meeting (Drug Information Association): San Antonio, Texas, June 15-19, 2003.
Sidebar: ‘We Are Far Behind’ and Other Views from Mt. Olympus
The best-attended event at the Drug Information Association meeting was a panel discussion among some titans of R&D, moderated by Murray Lumpkin, principal associate commissioner at the FDA. They discussed a range of topics: pharmacogenomics, trials, productivity, outsourcing.
Mathias Hukkelhoven, senior vice president of global drug regulatory affairs at Novartis, predicted the industry would have to shift not only to more electronic documents, but also to more effective handling of those documents. "As an industry we are far behind in how we handle electronic documents," Hukkelhoven said, ticking off submissions, safety reports, clinical trial data, and other correspondence. "That is where everybody is going. That's a given. We just need to do it. It's a no-brainer."
Per Peterson, chairman of pharmaceuticals research and development at Johnson & Johnson, is not against pharmacogenomics -- the tailoring of medicines to collections of genetic traits. But Peterson does seem to be worried about research costs that remain constant as the target markets for such drugs shrink. "My development program is not going to be cheaper than if I were developing for 100 percent of the patients," Peterson said. "If we cannot lower the development costs, the less interest there is [in pharmacogenomics]."
Peterson described a recent clinical trial in which the company hoped to do gene-expression profiling. "More than 75 percent of the patients refused to participate because they thought we were trying to get too much information," Peterson said. "It's an education issue more than anything else."
In response to a question about the future of contract research organizations, Peterson said. "There will be more efforts in reducing costs in big pharmaceutical companies. It may have the effect that less work will be outsourced."
Peterson also wondered aloud about whether the right information is being collected in clinical trials. "I question whether we generate safety data that are relevant in our trials,” he said. “Are we actually generating the optimal amount of information in our clinical trials to make the decision that we have a safe and effective drug? We see a different pattern of side effects once the drugs are out there. That is something the regulatory authorities could start addressing now."
Merck's David Blois, senior vice president for global regulatory policy, was marginally more optimistic about everything, from products that combine a drug and a device (coated stents to reopen coronary arteries being the most profitable example) to pharmacogenomics. "That could happen in a Phase II trial," he said, "which is really a large screening."
Blois said the pressure on research budgets is inexorable, and predicted that mergers would be responsible for some headcount reductions. "There are a lot of people who are going to be outplaced because of that," he said. -- M.U.