Diagnosis for digital pathology

July 1, 2009

The pathology industry is undergoing dramatic changes
as new technologies transition the field into the digital age with
technology enabling pathologists not only to improve workflow but also
to use these new analyses, algorithms, and abilities to collaborate
globally. These new tools allow pathologists to offer patients more
detailed diagnostic insights, which are critical when taking a
personalized-medicine approach to treatment.

Pathologists depend on image quality to make an
accurate diagnosis; the image is the single most important factor in an
accurate, consistent diagnosis. Digitization of slides makes images
available for pathologists’ use in ways that are not possible with
manual microscopy. These advances are particularly important as
personalized medicine becomes a reality in areas (i.e., breast cancer)
and as physicians look to pathologists for critical support.

Image analysis can more accurately and reproducibly
quantify various features in different types of studies, including
immunohistochemistry, immunofluorescence, and fluorescent in situ
hybridization (FISH), which are being used by pathologists to arrive at
a specific diagnosis and to guide clinicians in therapeutic decisions
for a subset of their patients. Depending on study results, therapies
are personalized for patients, for more precise, targeted treatment and
disease-state monitoring.

Traditionally, community pathology practices are
geographically fragmented and small. Systems must be designed with space
restrictions in mind; a small footprint is critical. If a system relies on
other components (e.g., power supplies, cooling systems, or computers),
locate those components remotely and flexibly.

Imaging, computer, and storage technology have advanced
to the point where digital pathology is available to every lab to support
better decision making, workflow, and, ultimately, patient care. Web-based
software applications designed to work with slide scanners enable
pathologists, histo-technologists, lab administrators, and clinicians to
improve the efficiency and quality of the various steps in the anatomical
pathology workflow. Technologies now allow users to first scan glass slides
and — with whole slide images in hand — to view, manage, manipulate,
analyze, report, and collaborate within a digital environment.

Moving to a digital format has increased the need for
pathologists to stay connected with their professional networks to keep up
to date with the latest advances. Launched in March, PathXchange.org (PX), a
not-for-profit professional networking portal for the global pathology
community, brought pathology into the digital age with Web 2.0 features
designed to promote exchange of pathology cases, ideas, knowledge,
information, products, and services. Cases and information can originate
from anywhere, from any platform, and can be shared with sub-groups of a
user’s choice.

The pathologist will play a critical role in testing for
specific biomarkers to predict patient response to given therapeutic agents,
as will the primary physician analyzing and processing the tissue. Digital
pathology can provide these results. Within a few years, the number of
specific biomarker assays is expected to expand rapidly as a result of the
increased number of drugs being developed with companion diagnostics.
Processing cases in a digital, mineable form is important so clinical
decisions can be extracted and made available at the point of care, allowing
physicians and caregivers to make the most appropriate decisions in
personalizing medicine.

Herceptin highlights the crucial role digital pathology
can play in selecting patients for a given therapy. Based on the pattern and
quantity of the human epidermal growth factor receptor 2 (HER2/neu) protein
in the membrane compartment of breast cancer cells as assessed by
immunohistochemistry, patients are selected for treatment with the
monoclonal antibody Herceptin or, in equivocal cases, for a second
FISH-based assay that assesses amplification of the HER2/neu gene.
Interpretation of both the immunohistochemistry and FISH tests relies on
quantitative assessments facilitated by digital pathology and image
analysis. Misinterpretation of results by under- or over-estimating HER2/neu
signals can lead to suitable patients not being treated with a potentially
life-saving therapy and unsuitable patients being treated with an
unnecessary, high-cost drug with potentially devastating cardiac side
effects.

Pathology has only started to make the leap toward
adopting digital solutions in the past decade and faces challenges
associated with this transformation. As an example, many consider digital
pathology to be a “disruptive” technology that will have to displace
existing systems. Pathology labs face internal competition from other
departments for scarce capital resources. Systems are available, however,
with low or no capital outlay that can often be approved at the departmental
level, saving resources for other uses within the lab or hospital.

When deciding to purchase a digital pathology solution,
health systems need to consider how quickly and with what effort can a
system be deployed. Thanks to advancements in technology, systems are now
available that can be deployed rapidly — with little, if any, optimization —
and are intuitive to operate with minimal training requirements. Although
adoption is still in its adolescent stages, the move toward a digitized
realm in pathology is already proving to be “the missing link” as physicians
strive to make personalized medicine a practical, broadly based reality.

Ajit Singh, PhD,
is the CEO, and
Robert Monroe, MD, PhD
,
is the CMO of BioImagene, a digital pathology solutions company, based in
Sunnyvale, CA. BioImagene products are FDA-cleared for specific clinical
applications and are intended for research use for other applications. To
learn more, visit www.bioimagene.com.

ID 335171794 © Yuri Arcurs | Dreamstime.com
dreamstime_xxl_335171794
ID 141973761 © Vetre Antanaviciute-meskauskiene | Dreamstime.com
dreamstime_xxl_141973761
ID 325303940 © Karina Schultze | Dreamstime.com
dreamstime_xxl_325303940