As personalized medicine in oncology continues to progress, companion diagnostics (CDx) are becoming ever more critical in the future of treatment and in its success. CDx help determine the patients most likely to respond to a therapy, but while detecting predictive biomarkers is a strong focus for the industry that develops these invaluable tools, identifying and developing robust CDx is still a very challenging process. It starts with tissue analysis – and digital pathology, tissue image analysis and other data such as genomics – are essential. It is at this stage where the challenges begin. First, the biomarkers being identified and targeted for CDx can be inducible; thus, if a patient has already undergone treatment such as chemotherapy, the expression of those markers may be influenced and results will be impacted.
An additional primary obstacle currently is pathologists’ inability to visualize the spatial distribution of immune cells and their correlation to the tumor. Understanding the spatial relationship between immune cells within the tumor setting is increasingly important in helping identify biomarkers in immuno-oncology, and it requires powerful tissue image analysis tools. As such, standardizing the use of digital pathology – which includes digitizing tissue slides and using automated tissue image analysis to extract relevant data about cancer cells, their behavior and their spatial relationships to other cells – is a key next step in advancing the diagnosis and treatment of cancers. While many are of the understanding that tissue image analysis simply automates routine tasks such as counting cells, there is so much more to these tools; they are able to dig deep into basic biology, providing the kinds of information traditionally found in quantitative liquid biopsies, but with the added morphological context that only tissue provides. Today’s digital pathology and image analysis tools enable pathologists to extract and quantify relevant information from digitized tissue images, helping them to not only identify markers, but also understand the correlation of different markers and their spatial relationships to one other. After gathering this tissue data, they can also view it in context with genomic and clinical data. This kind of “X-ray vision” assists pathologists in reading information beyond plain eyesight and coming to conclusions based on biologically relevant and clinically meaningful features in the tissue along with other important patient data. For pathologists in R&D, this means they have the ability to discover meaningful biomarkers; and for pathologists in the clinic, it means they can retrieve data that demonstrates how a patient will benefit from a particular treatment. Tissue image analysis makes pathology an equally quantitative versus merely qualitative discipline. It’s the only way to effectively develop tissue-based CDx, which will allow pathologists to successfully support clinical decision-making. And significant technological advancements are being made.
The integration of multidimensional omics, or “multi-omics”, which looks at a combination of genomic, proteomic, epigenomic, tissue phenomic and other relevant information as part of a big data approach to personalized medicine, is an important advancement. Multi-omics enables researchers and pathologists to rely on curated data correlations rather than isolated data sets to gain additional insights, answer research questions and make diagnostic and prognostic determinations. Additionally, as tissue image analysis tools become more powerful and precise, they are also being evaluated for use in cancer immunoprofiling. Immunoprofiling helps provide a more personal picture of an individual patient’s disease, how it is behaving and how the immune system is responding, and subsequently can support more effective clinical decision-making and patient treatment. Ultimately, pathologists must be able to provide good guidance to oncologists on treatment options. In order to accomplish this, it is essential that industry formalizes the use of digital pathology, providing the tools for pathologists to make better decisions about disease diagnosis and prognosis than they are able to reviewing tissue samples by eye alone.
While in principle all intelligent components for the standardized use of digital pathology across the industry exist, in actuality there are still a number of challenges to overcome. Digital pathology by nature requires the scanning of available slides and the digitization of the tissue images, and while there are of course tools to do this, a lack of different resources across labs often creates limitations here. Additionally, since digital pathology platforms can digest millions of data points from tissue and other samples – which of course is where the opportunity lies for advanced CDx and clinical decision support – it can be a challenge to understand and analyze all of this data and then correlate it with patient outcomes. Finally, some skepticism across the industry – from companies to patients to payers – about this approach limits the growth of digital pathology. There is still work to be done to convince all parties involved that digital pathology is crucial to the advancement of cancer drugs and diagnoses, but, as industry invests in more powerful tools and globally standardized lab and tissue testing procedures, pathologists will be in an even greater position to support oncologists and patients in making truly personalized treatment decisions.
Newsletters
Receive the latest pathology news, personalities, education, and career development – weekly to your inbox.
