DNA methylation can reveal patterns of disease that conventional pathology and genomic testing may not capture. Now, advances in comprehensive genomic profiling are creating opportunities to integrate methylation profiling into a broader multi-omics approach to precision oncology.
At a symposium organized by Illumina at the European Congress of Pathology, Helge Lubenow explored how methylation technologies could reshape diagnostics and advance precision pathology. The Pathologist caught up with Lubenow after the presentation to discuss how DNA methylation profiling could enable more precise tumor classification, inform patient management, and complement established pathology workflows.
Why is DNA methylation profiling such a valuable source of diagnostic information, and what can it reveal about disease that conventional genomic testing may not?
Methylation-based tumor classification has been extremely valuable in identifying tumor classes that cannot always be distinguished using established pathology workflows such as H&E staining.
The current WHO classification recognizes more than 150 tumor classes, and for some of these, methylation-based classification is required for accurate identification. In those cases, molecular characterization is not simply an additional recommendation; it can be an essential diagnostic criterion for assigning a patient’s tumor to the correct class.
Importantly, methylation profiling has also helped shape the classification framework itself. In central nervous system (CNS) tumors, for example, methylation-based classification contributed to establishing the nomenclature now used in the WHO system. That is why you see methylation classification represented on the cover of the current WHO CNS “Blue Book”: it reflects how fundamental this approach has become to defining and diagnosing certain tumor entities.
Which clinical problems do you think methylation-based technologies are particularly well suited to solve?
One of the main applications is matching patients to the correct disease classification. That is different from many molecular biomarkers, where a particular finding may correlate directly with a specific treatment. Methylation-based classification can provide information across the entire disease landscape, so its implications extend beyond selecting a particular drug.
It can help determine how a patient should be managed. That might mean adopting a watch-and-wait approach or downgrading a diagnosis and avoiding unnecessary treatment and its associated side effects. It might inform the timing of treatment, or, in some cases, the selection of a particular therapy. We are therefore talking about both escalation and de-escalation of patient management.
My company's focus is on pediatric brain tumors, and that has strongly influenced our perspective. Avoiding unnecessary treatment and its long-term side effects is particularly important in very young patients, who may have to live with those consequences for the rest of their lives. That makes the clinical proposition quite different from many adult cancers.
How might methylation profiling fit into a diagnostic workflow?
We see methylation profiling as something that integrates with the established pathology workflow rather than replacing traditional pathology. More recently, we have begun cross-training image-based analysis of conventional H&E slides with methylation classifiers. The aim is to help pathologists identify which patients are most likely to benefit from more detailed molecular classification.
We would not propose molecular characterization for every patient. Instead, by combining image-based analysis with methylation profiling, we can help triage cases and identify those where molecular testing could add meaningful diagnostic information. That allows methylation-based classification to be incorporated into the workflow in a more targeted and cost-effective way.
Is the future increasingly about bringing these different diagnostic layers together?
Absolutely. The important thing is to integrate these technologies in a way that can be deployed broadly. H&E staining and conventional pathology are performed globally and will remain the foundation of diagnosis, irrespective of a laboratory’s level of sophistication.
We envisage a two-tiered approach. Routine pathology can identify patients who may benefit from more specialized molecular characterization, with those cases then referred to expert centers or networks for further testing and, where appropriate, a second opinion. We are already speaking with groups in different regions that want to establish these kinds of expert networks.
That is what we mean by integrating molecular technologies with standard pathology. We are not suggesting that every pathology laboratory will eventually deploy molecular methods. Instead, the goal is for laboratories to recognize which patients should be referred for more advanced testing and which can safely continue to be managed through the existing pathway. In that way, we can make advanced diagnostics accessible much more broadly.
Do you expect molecular testing to remain centralized, rather than being offered by every facility?
That is still an open question. Many molecular technologies are currently most cost-effective when they are run at scale. One possibility, therefore, is that testing remains centralized and efficient sample logistics become increasingly important. Alternatively, smaller, more affordable platforms may emerge that allow more centers to adopt these technologies locally. At this stage, we do not know which model will prevail, and the answer may differ between healthcare systems.
Through our MNP Outreach initiative, for example, we work with organizations in regions where access to advanced diagnostics can be challenging, including countries across the Global South. We discuss the feasibility of centralized versus decentralized testing at a country level, and there is considerable variation. We are working with groups from Brazil to Pakistan and Thailand, and there will not necessarily be a one-size-fits-all solution.
Access has always been particularly important to us because of our roots in pediatric brain tumors. We even have an arrangement with the National Health Service in England that provides free access for pediatric cases. Although pediatric tumors represent a relatively small proportion of brain tumors overall, brain tumors are a major cause of cancer mortality in children. That clinical need helped create an early impetus to explore methylation-based diagnostics.
What other factors might affect the accessibility of DNA methylation profiling, going forward?
A major factor is reimbursement. Gaining a CPT code for CNS tumor methylation profiling in the US helped demonstrate that healthcare systems recognized the clinical value of the approach and made adoption more viable for laboratories. We are now working toward expanding reimbursement into areas such as sarcoma.
Ultimately, cost will remain important, and this is where integration with sequencing could help. Array-based methylation profiling was instrumental in establishing the field, but it primarily provides methylation information. If methylation can be incorporated into sequencing workflows alongside other molecular information, laboratories could obtain more information from a single workflow. That has the potential to make advanced molecular characterization more cost-effective and, ultimately, more accessible.
Looking five to 10 years ahead, how significant a role could methylation profiling play in precision pathology, and where are the biggest opportunities?
We believe there is potential utility across all tumor types. Not every patient will require methylation profiling – many diagnoses will continue to be resolved under the microscope – but we estimate that around 20 to 25 percent of patients presenting with tumors could benefit from it. Within 10 years, we expect it to become an established part of diagnostic practice across many well-resourced healthcare systems.
The bigger challenge is developing models that can make these technologies accessible globally. That question is becoming increasingly important in discussions around disease classification. We are seeing greater representation from resource-constrained settings in WHO expert panels, bringing more focus not only on what is diagnostically desirable, but also to what can realistically be deployed in different healthcare environments.
That is one area where methylation has an important practical advantage. RNA-based gene expression can provide valuable information, but RNA is comparatively unstable, which places greater demands on sample storage and transportation. Those considerations become particularly important when you are deciding whether testing should be centralized or decentralized.
Methylation, by comparison, is extremely stable. We can go back to samples that are 25 years old and reassess them. So, two technologies may potentially provide similar diagnostic insights, but turning that science into a deployable diagnostic solution is a different question. Sample stability, transport, infrastructure and accessibility all matter. Those practical advantages could make methylation profiling particularly valuable as precision pathology becomes more widely adopted.
How do you see methylation-based testing complementing conventional pathology?
The most important point for us is that we are not talking about replacing conventional pathology. Pathology is already extremely efficient, both in terms of cost and turnaround time.
Turnaround time is particularly important to me because, behind every sample, there is a patient and a family waiting for an answer. Even today, a molecular workflow will typically take around seven days. We are becoming increasingly efficient on the cost side, but my ambition is to make molecular testing much faster as well.
I think we will get there. In the meantime, conventional pathology does a wonderful job, and our aim is to strengthen what pathologists can already achieve by adding another layer of information where it is needed – not to threaten or replace it.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
