Forecasting the future of cancer diagnostics


One of our experts answers questions about innovations in tissue and digital pathology.

February 20, 2024


Innovation and partnership are more important than ever in creating diagnostic tools to meet the needs of laboratories as they face staffing shortages and an increase in cancer cases. Changes in the diagnostic environment also are opening up opportunities to better leverage the latest technologies and partnerships. For example, recent changes in government regulation have paved the way for pathologists and other laboratory personnel to more easily use digital tools to review lab data, digital results and digital images remotely.

Katie Robertson, Ph.D., Roche Diagnostics’ disease area network lead for oncology, answered a few questions about how Roche is using innovation, like digital pathology and companion diagnostics, that rely on partnerships to adapt and grow to meet lab challenges. Robertson highlights the latest lab conversations on digital pathology and Roche’s pipeline.


There’s been a lot of buzz about digital pathology in the last few years, as hugely important to medical technology. Who has been adopting it?

Digital pathology is at a really interesting intersection right now. Many labs may say they have digital pathology, but most labs now have only scanners that support whole slide imaging. That’s pretty much table stakes. The minority of labs are fully digitized. Most labs fall into the category of Should we adopt it? Is now the right time?


What clinical questions are customers asking you about digital pathology?

Customers are asking about clinical launches – what will they be and when will they be ready? Specifically, they are asking if whole slide imaging solutions will be FDA approved. Right now there are only a few solutions approved. Roche has a pending FDA clearance.


Why should whole slide imaging be important to labs?

It supports the ability of clinical labs to image a large number of slides quickly. Digital images can be magnified so that they feel as if you are viewing them under a microscope. We know this transition happened in radiology. We went from film to digital images. Now pathology labs are feeling the pressure. Labs know they have to embrace innovation to modernize their pathology services. Fully embracing digital innovation means more than acquiring whole slide imaging, but it can be a good first step.


What’s the big hurdle for fuller adoption of digital pathology?

There are a few things. Labs have to be both technically and operationally ready, and of course, there are equipment and other related costs. Those are the three hurdles we hear. You can’t move into digital at half pace.

All the pathologists need to be on board, all the techs, all the IT staff. The whole lab needs to be ready to transition and have a budget for implementation and data storage. A lot of pre-work and pre-decisions need to come together to make sure the lab is ready, and it's no small feat. 


Are there ways that pathologists can help get their laboratories ready to go in that direction?

A champion can work to get everyone onboard and help colleagues understand that this will take time and commitment. You have to take time to learn about it. It's something new. If everyone is open and gives each other grace, that will go a long way to a smoother implementation. 

There may be some stumbling blocks in the beginning of figuring out the workflow. You have to think, We’re doing this, we’re going to make mistakes, but everyone has to have a positive attitude. Everyone needs to be willing to go on the journey together. 


What is the latest innovation in Roche digital pathology?

We heard from customers that they needed a HER2 quantitative image for breast cancer that aligns with current ASCO-CAP guidelines, which our test previously didn’t. So we developed a new algorithm that employs Deep Learning AI trained on a robust dataset of images that should be available in the near term. We’ve been focusing our investments in solutions that meet the needs of customers. Many of these are research-use-only (RUO) innovations. 

We have built a flexible RUO API (application programming interface) framework to enable third-party companies to deploy their image analysis algorithms tools within navify Digital Pathology (nDP). This has created an environment where we can collaborate with other companies making innovative algorithms, in addition to building our own. 

The RUO IBEX Galen First Read Prostate and Breast will be available in the near term. Galen Breast is a clinical-grade solution that assists pathologists in improving the detection and grading of breast cancer. Galen Prostate uses AI to analyze biopsies ahead of pathologists’ review, providing them with diagnostic insights to guide their diagnosis. These are algorithms embedded directly with navify Digital Pathology.


If you look out at all the disease states that tissue diagnostics supports, where do you see the most impact in the next few years from Roche? 

Tissue is always going to play a big role in breast and lung cancers. We are starting to play a role in gastric cancer. We have a potential gastric companion-diagnostic test pending approval. Our focus will be on the new companion diagnostic tests and quantitative algorithms in the disease states we currently support.


How is the future tissue diagnostics pipeline looking?

Lung cancer is always of interest. Despite advances in treatment, lung cancer remains the leading cause of cancer-related death in both men and women. Recent research suggests that c-Met overexpression is present in approximately 37% to 67% of NSCLC cases. The only known ligand for c-Met is HGF (hepatocyte growth factor), and overexpression of c-Met is correlated with more aggressive clinical behavior. There are no currently-approved cancer therapies specifically targeted for patients with c-Met overexpressing NSCLC, but there is a lot of exciting research in targeting c-Met overexpression, and there are a few clinical trials I am watching.

Another test in the pipeline is one to test for high expression of Claudin 18 (CLDN 18) for gastric cancer. Gastric cancer causes about 2% of cancers in the U.S. Recent studies have shown that approximately 36% of gastric and metastatic gastric cancers exhibit high CLDN18 expression.

Data also suggests that there is limited overlap in the expression of other common gastric biomarkers such as PD-L1 and HER2, meaning that many patients with high CLDN18 expression fell into a previously undefined patient population. This will be a very exciting opportunity for patients. 


What clinical trials are you watching? 

I’m watching several trials that are being conducted by partners of ours. 

I am following the Zolbetuximab Biologics License Application for the treatment of patients with locally advanced unresectable or metastatic human epidermal growth factor receptor 2 (HER2)-negative gastric or gastroesophageal junction (GEJ) adenocarcinoma whose tumors are claudin (CLDN) 18.2 positive. 

Another is the Phase 2 LUMINOSITY trial for patients with previously treated non-small cell lung cancer. 

Also, I’m watching the Phase 2b HERIZON-BTC-01 clinical trial on progression-free survival of the bispecific antibody zanidatamab in previously treated HER2-amplified biliary tract cancers. About 5% to 19% of patients with BTC have tumors that express HER2 and there is no targeted HER2 therapy currently available. We are working to develop a companion diagnostic for this therapy.


Katie Robertson, Ph.D., is a disease area network lead for oncology at Roche Diagnostics. She leads a team that provides disease area expertise and clinical input into medical strategies across pathology and sequencing for oncology.