Whenever we publish anything about imaging core labs, it seems, a good Samaritan rushes forward to point out all of the room for improvement in our previous articles. It's moderately discouraging on one level, but always nice to hear from interested readers.

This is our third recent article on the topic. If you're running an imaging core lab and you haven't participated in our journalistic-education campaign, here's .(JavaScript must be enabled to view this email address)

The latest help is from BioClinica's Mark Weinstein, who is president and CEO. He's amiable about setting the record straight. But he clearly believes a recent story overemphasized the role of technology in imaging core labs.

For starters, Weinstein says, the Metrics Champions Consortium (MCC) benchmarks (mentioned in our earlier coverage) remain timely. They are not obsolete. His firm supports the MCC effort. The company believes that its numbers will become more challenging targets over time. And BioClinica will make the MCC imaging benchmarks continuously viewable to customers in trial dashboards.

PM Acclaim

Weinstein's main critique is that our coverage hasn't stressed the service side of central imaging. For sponsor firms, regulators and his imaging core lab competitors, Weinstein says, there are more important issues than technology. He acknowledges that his company's overall efficiency is dependent on his technology. But he doesn't put it on a pedestal. "There is a huge service component to this business," he says.

Specifically, imaging core labs must include virtuoso project managers (PM). The sponsor and the PM at an imaging core lab may spend three years in weekly or daily contact. If the PM's relationship with a sponsor works, it will lead to more projects. If not...

BioClinica appears to be doing something right. Weinstein says it typically has 200 imaging projects running at any given time, with 1,000 core lab projects completed over the company's existence since the early 1990s. "People don't give me projects because of my technology," says Weinstein.

No Rush

Weinstein also intimated that our earlier article could have used some history. Core labs didn't come into being simply to move images. Their raison d'être was to provide independent analysis and scientific consistency. "Imaging core labs were created by the FDA, which said, 'we're going to accept surrogate end points. But we are concerned about 1,000 reads by 1,000 radiologists,' " Weinstein says.

Another quibble with our coverage is an unstated inference that imaging core labs are somehow involved in ongoing patient care. Wrong, says Weinstein. The two worlds are distinct. "We do not find patients or guide their care. We don't do patient management," he says. "There is nothing we do where we are determining how a patient should be treated." So turnaround times for scans for industry research, he says, seldom need be immediate.

Fitting In

Beyond that, Weinstein says, clinical trial systems and processes must be (for lack of softer language) subordinated and coordinated with those for patient care. He has no illusions that clinical trial systems are going to take precedence over, say, looking at Mrs. Smith's broken hip or Mr. Jones' brain tumor.

The technology for clinical trials, in brief, must route a small subset of images toward a small number of the best radiologists. Thus any imaging core lab must be able to anticipate the needs of a thought leader at a university medical center like Johns Hopkins or the Cleveland Clinic. "The guy who has seen 10,000 scans of the disease is going to be better than the guy who has seen 20," he notes.

Conforming to both community practice and academic settings without disrupting either is important to Weinstein's customers. And that, in the end, is why (by his estimate) 90 percent of all industry research-related images are burned to CDs or DVDs and shipped by overnight courier. That's what works, Weinstein says.

Overnight Shipping

"Reading for clinical research is different than clinical practice," says Weinstein. He can rattle off the names of a dozen top key opinion leaders in radiology. He says the sponsor community knows such individuals are invaluable when reviewers at regulatory agencies have questions about a trial. In any high-level imaging debate with a regulator, Weinstein explains, a perfectly competent young radiologist at a community hospital will struggle. A genius consultant from Mass General will hold her own.

The limitations of the current approach are becoming a bit more visible. Weinstein says that recently sponsors have started to be aware of the cost—$50 to $100—to send a CD of images or actual film around the world. Some trials probably spend hundreds of thousands of dollars on such shipping costs. Tighter budgets are putting that under scrutiny. Images can be moved electronically for perhaps $18-20 each, Weinstein suggests.

Scan Subtleties

Indeed, Weinstein is enough of a technology guy to confess what sounds like a milliliter of chagrin that so many images are couriered around in ways that would have been recognizable to Charles Dickens. But he says he is a pragmatist. When a sponsor is informed that a keenly sought radiologist has declined to use some fancy new system, the sponsor typically relents and (to keep the radiologist on board) allows FedEx or DHL to handle the images. At every step, that disk has to be futzed around with as its metadata or images are inspected and matched to a case report form or larger work flow.

For now, Weinstein is skeptical that technology will make much difference. Yes, he says, the pharmaceutical industry will slowly move toward more electronic processing of images. But what really counts is knowing how to get exactly the same inflection angle on every knee scan. Or which details of the protocol must be comprehensively specified, and which can be allowed to drop inside a range of values. Or how to measure brain shrinkage in Alzheimer's disease.

CNS Wins

Getting those details right across dozens of sites, with hundreds of patients, is something they don't teach in radiology residency training. Weinstein is especially proud of the number of trials BioClinica is running in the central nervous system (CNS) therapeutic area, and he doubts new arrivals to the imaging core lab arena will be able to make inroads on that terrain. BioClinica acquired a small French outfit specializing in CNS imaging, and is relying on that group in Lyon as part of its ongoing R&D effort.

For Weinstein, there is little doubt that central imaging projects are tougher to do well than, say, electronic data capture (EDC) databases. There may be 50 firms doing EDC, and just five or six major imaging core labs. In the imaging world, Weinstein says, setting up a scan incorrectly could in some cases forfeit the patient to the trial (if she was destined to be exposed to too much radiation).

Newly Selective

Like any other global firm, BioClinica has international reach. An office in Leiden, the Netherlands, has a staff of 100 and is especially involved in imaging projects. BioClinica has worked with 14,000 research-oriented imaging centers in 80 countries over the years. Weinstein sounds impressed at the energy and sophistication of up-and-coming radiologists in developing markets. They do not regard industry-related work with the stifled yawns of some U.S. and European researchers who are sufficiently well compensated by their day jobs. "There are a lot of good radiologists all round the world that are well published," he says. "They don't have to be in the U.S. and Europe."

BioClinica's best guess about the size of the imaging core lab market is that $400 million of such projects unfold annually. (The company says there is no independent research on the niche, but believes it can gauge the market size from requests for proposals.) Synarc, BioClinica and a handful of top 10 contract research organizations, including Icon and Parexel, may split 50 or 60 percent of the work between them. Academic groups remain an important competitive element, Weinstein allows.

One new dynamic: sponsors are confining their choices to companies on their preferred provider lists. That may lock out firms that have not been vetted. Back in the day, some sponsors quietly used plenty of firms that were not on their preferred provider lists. But that is starting to change. "It is a very difficult market if you are starting from scratch," Weinstein says. "Which is why we have not seen any entrants to speak of in a couple of years."

Editor's note: Our second story concerned Merge's central imaging effort; our first detailed that of Perceptive Informatics.