AstraZeneca will be getting a lot more for its data-collection dollar now that it’s storing clinical images in its own centralized repository. As of January, all new clinical trial projects can utilize the repository to more efficiently manage clinical images and streamline regulatory compliance, says Anastasia Christianson, PhD, director and global skill leader, discovery medicine informatics for the company’s U.S.-based research and development efforts.

The repository is built on Oracle Database 10g, which was selected over other platforms largely because it “allows us to view images, not just store data,” says Christianson. Oracle further supports the DICOM (Digital Imaging and Communications in Medicine) cooperative standard for distributing and viewing any kind of image, regardless of the origin. Familiarity probably also played a role. “Many of our databases are Oracle-based.”

Picture archiving and communication systems were not only missing the image-viewing feature, but also had a lot of unneeded functionalities, says Christianson. For example, AstraZeneca didn’t need clinical images to be stored in one giant database with all other clinical information. It needed linkages between them.

Oracle Database 10g, introduced in January 2004, is now in its second release. The Oracle interMedia features of the database enable efficient management and retrieval of multimedia data, including automated metadata extraction and basic image processing, says Christianson. 

Bringing imaging data in-house creates “extra work,” says Christianson, but is well worth the trouble. “We can go right back to the original data, right here, to re-analyze it or put it in context with other data.” The need to revisit data comes up most frequently in exploratory studies at the “method development” phase.

Reconciling data previously took many times more steps, “to the point where often it was not worth doing,” says Christianson. Responding to image-related queries from regulatory agencies was also tedious because data was stored in auxiliary storage devices, often scattered across dozens of internal and external locations. Images could potentially get lost during transfer. With the repository, “we’re able to do more internally. It’s easier if we have to do searches to find particular data [or a particular] study or image.”

Acquiring and implementing the technology was a year-long project. The change will be invisible to partnering contract research organizations, excepting requests for a copy of the original data, says Christianson

Many big pharmaceutical companies have clinical imaging repositories, although “they may not all use Oracle” as the foundation, says Christianson. For AstraZeneca, “the timing was perfect. Oracle was releasing a new version with the right functionality while we were looking for a solution. It was a marriage made in heaven.”

Early phase clinical studies can generate up to 100 images and later phase studies many times more, Christianson says. The image repository will “easily accommodate all we do for the next few years.” By this time next year, AstraZeneca expects the repository will exceed 100 terabytes of data.

Want to read more expert articles like this? Send an email to:
subscribe-ecliniqua@newsletters.bio-itworld.com.