Developing a Digital Pathology Strategy

Benefits of digital pathology are being realized across many laboratory and scientific spectrums, as it enables researchers and pathologists to look at digital images from any investigative site worldwide. In addition to eliminating issues associated with transporting and storing human tissue, digital pathology for biomarker research, in particular, affords the ability to view sample images almost instantly, reduces the drug development timeline and saves significant costs to a development program.

What follows is a discussion on best practices and lessons learned for labs interested in developing a global digital pathology program.

Benefits of Digital Pathology

Consider oncology, where biomarker work commonly involves stains on patient tissue that highlight the presence of a certain protein or genetic target. Digitally capturing this information allows a pathologist to leverage the richness of data contained within the patient's tumor morphology. Today's technology also allows us to digitally capture an entire slide, share it almost in real time and mine the image for quantitative and ­qualitative data.

The use of digital pathology also can be extended to clinical trials. The ability to identify a specific patient population reduces the overall time needed to conduct a trial. By knowing exactly what to look for on a slide sample, researchers can immediately narrow the potential pool of patients to a specific genotype or protein level. Biopharma as a whole is certainly beginning to embrace this emerging technology, and many organizations have established digital pathology networks with in-house scientists tuning digital algorithms to analyze samples. These activities may lead to further improvements in personalized medicine where patients are selected more accurately to receive specific targeted therapies.

Finally, a centralized digital pathology network in which all image analysis and reading is conducted in one location-using a ­standardized process-eliminates variability, increases accuracy, provides data storage security and enables near real-time dissemination of vital image data. Working with an organization that has invested in a robust, scalable and secure digital repository allows drug developers, for example, to extract these benefits. Substantial thought has to be invested in such a repository, but once it has been setup and deployed on a global basis, the drug developing organization can begin harvesting the benefits of expanded collaboration and reduced shipping and logistics cost.

Setting Up a Digital Pathology Network

Before beginning the process of establishing a digital pathology network, having a very clear picture of the ultimate objective is essential.

The process of converting systems to a digital platform can be lengthy and costly, and a very concise business plan needs to be in place to ensure the investment will be worthwhile. To prevent redundancies in workflow, all elements need to be examined and aligned with the end goal of an organization.

As with any significant information technology undertaking, an architecture assessment is a critical early step when considering a digital pathology network. This consists of reviewing available technology for compatibility with existing systems and ensuring that the overall platform matches the organization's needs. An architecture specification is a more detailed document in which the specific technology is named and a process for implementation is explained. This is eventually handed off to the implementation team, essentially as a recipe.

A key component in establishing a digital pathology network is data migration. The organization needs to consider how old, slide-based data, the gold standard of several decades now, will be imported into a new system.

Keys to Establishing a Network

Additional items of consideration in the early stages of establishing a digital pathology network are robustness, resiliency, reliability, scalability and security. The necessary technical components must be in place, so ensure your information and digital imaging systems easily and quickly scale all aspects of system requirements. One of the biggest issues to address, for example, is security and potential system vulnerabilities. From a security perspective, one of the things we struggled with was setting up a security model. The trick was to keep it as simple as possible without compromising the security of the server. Naturally, ­security-and proper documentation thereof-is of utmost importance if the images are to be exposed to external customers. We require our clients to clearly articulate their expectations of our digital network and hold detailed discussions around role specifications of the users and defining access protocols.

Lessons Learned

Global implementations are complex and require detailed planning. Education and working with your internal information technology department are enablers for success. The IT group needs to be involved from the very beginning to ensure that the whole implementation process is as smooth as possible.

Digital pathology can be a very disruptive technology for an organization. The much revered microscope, sacred to many pathologists, is transformed from an isolated unit to, well, the world. As such, clear communication between the organizational leadership, the lab and the information technology team is critical.

At Quintiles, our network will allow us to leverage digital pathology to help develop more targeted therapies for patients. Using tools such as TMA lab and image analysis algorithms allows us to create additional value for our assay development work. As we set the standard for utilizing this technology to optimize the drug development process, we take one step closer to realizing our ultimate vision of helping people live healthier lives.

Christopher Ung is vice president, Strategic Business & Operations, Oncology, Quintiles, and Pete Tearle is enterprise architect, Quintiles.

7 Step Strategy for Implementing Digital Pathology

Define the business case for digital pathology (DP). The process is initiated by identifying the bottlenecks in the current workflow and the market drivers underlining the need for adopting DP. This information will help labs identify their DP use cases that can include one or more of the following: whole slide imaging, image analysis, remote sign-outs, second opinions, peer reviews, quality control, remote consultations, frozen sections, digital archives and/or reporting. A Return on Investment (ROI) analysis is also conducted to determine financial viability of the business case.

Analyze implementation requirements and objectives. The laboratory defines the case volume, targeted growth and implementation options and sets measurable objectives for implementation. These might include target turnaround time, site configuration such as VPN/Remote/Local/Http/Citrix, access model such as Single/Multi/Satellite location/Centralized, data retention policies, revenue targets and usage targets.

Develop solution architecture. Using the information from steps 1 and 2, the solution architecture is formulated and the components of the solution are designed. An implementation plan is created that includes the deployment timeframe and addresses deployment logistics such as space, connectivity, power and security.

Implement solution. Implementation of the DP solution at the customer site is the critical phase of deployment. Success of this step depends on the participation of dedicated resources from all stakeholders. The DP solution is tested and verified within the production environment.

Integrate with existing systems. The deployed system is integrated with existing laboratory systems such as bar codes, stainers and LIS systems. Integration testing and verification of the end-to-end solution is conducted.

Train and roll-out. A roll out plan is created. This covers training the users of the DP solution such as staff pathologists and histotechnologists, as well as sales and marketing people. The solution is "rolled out" to the lab's customers. The DP solution is live at this stage.

Measure success of solution. The DP solution implementation is measured against the objectives set in step 2. Monthly reports for usage, performance, revenue and cost are generated to ensure the solution is performing as expected. The outcome of this step could launch subsequent phases that would involve a new round of implementing the seven step strategy to implement digital pathology.

Information provided by BioImagene