Most people learn that our DNA lives on chromosomes. Textbooks show 46 of them, neatly paired, each containing thousands of genes in a fixed, orderly sequence. Cancer, as we understand it, arises when mutations disturb this order.

But that picture is incomplete.

Inside many tumor cells, loops of DNA float freely outside the chromosomes altogether. These are called extrachromosomal DNA, or ecDNA. They are not a minor curiosity. They are capable of driving some of the most aggressive cancers we know, and they operate by entirely different rules.

Until recently, ecDNA was largely overlooked. New research has changed that. We now know that ecDNA is present in roughly one in seven cancers. It can amplify dangerous genes to extreme levels. It helps tumors resist treatment and evolve rapidly. And, crucially, it can now be detected – opening the door for its use as a diagnostic and prognostic marker.

In pathology, this matters. ecDNA is not a distant concern for research laboratories. It is present in the specimens crossing our benches today.

What ecDNA does to cells

To understand why ecDNA is so dangerous, it helps to understand what makes it different from a standard chromosomal amplification.

When a gene on a chromosome is amplified, the copies are fixed in place. Every time the cell divides, each daughter cell receives a predictable share. The inheritance is orderly, even if the amplification is harmful.

ecDNA breaks that rule entirely.

Because ecDNA lacks a centromere, it cannot attach to the machinery that distributes chromosomes evenly during cell division. Instead, ecDNA molecules are distributed at random. Some daughter cells may receive many copies. Others may receive few, or none.

This seemingly simple quirk has enormous consequences.

Within a single tumor, individual cells can end up with wildly different amounts of ecDNA. That means they also carry differing levels of the oncogenes on those circles. The result is a form of genetic diversity within the tumor – known as intratumoral heterogeneity – that is far more dynamic than anything generated by chromosomal alterations.

Now consider what happens when treatment is applied. A chemotherapy drug or targeted therapy exerts selective pressure. Cells carrying more copies of a resistance-enabling gene survive better. Because ecDNA copy number can shift dramatically from one division to the next, a tumor can rapidly enrich for high-oncogene cells even within days.

This is not a slow evolutionary process. It can happen fast. When treatment stops, the pressure is lifted and copy numbers may rebalance. The tumor becomes, in effect, a shape-shifter adapting to therapy without permanently fixing a new mutation. That is what makes ecDNA so clinically formidable. Resistance is not merely acquired. It is dynamically modulated.

There is another layer to this. ecDNA does not just carry oncogenes. It also captures the genetic switches that control how actively a gene is expressed. By bringing oncogenes and their enhancers into close proximity within a circular loop, ecDNA creates unusually potent gene expression circuits. Studies have shown that genes on ecDNA are expressed at four to five times the level of the same genes sitting on a chromosome, even at equivalent copy numbers.

Inside the nucleus, ecDNA molecules cluster together in what are called "ecDNA hubs" – concentrated hotspots of transcriptional activity. These hubs generate extraordinarily high levels of oncogene output. They are, in a very real sense, cancer’s amplifiers.

Implications in oncology

ecDNA has been found across many cancer types, but a few examples illustrate its clinical weight particularly well.

• Glioblastoma – In this most aggressive primary brain tumor, ecDNA frequently carries amplified copies of EGFR. In some cases, the ecDNA carries a mutant form called EGFRvIII, which is constantly active even without a growth signal. The presence of ecDNA-mediated EGFR amplification has direct implications for clinical trial eligibility and treatment planning.

• Neuroblastoma – MYCN amplification – one of the oldest and most reliable markers of high-risk disease – is now understood to be largely ecDNA-mediated. Pathologists have been indirectly detecting ecDNA for decades through MYCN FISH testing, without always recognizing it as such.

• De-differentiated liposarcoma – ecDNA frequently co-amplifies CDK4 and MDM2. These two genes sit on the same circular element. They simultaneously dysregulate the cell cycle and suppress p53 activity. Their co-location is not coincidental but reflects the spatial co-amplification that ecDNA uniquely enables.

The survival data are unambiguous. Patients with ecDNA-positive tumors have significantly worse overall survival compared to those whose tumors carry chromosomally integrated amplifications of the same genes. This is independent of copy number alone. It reflects the unique biology of ecDNA: its transcriptional efficiency, its heterogeneity, and its capacity for rapid adaptation.

From a therapeutic standpoint, ecDNA creates new opportunities alongside its challenges. Its dependence on specific transcriptional co-activators – particularly a protein called BRD4 – suggests a potential vulnerability to a class of drugs known as BET inhibitors. Emerging strategies targeting the DNA replication machinery that sustains ecDNA are also under active investigation. The goal is to destabilize ecDNA specifically, reducing oncogenic load without equivalent damage to the rest of the genome.

Detection of ecDNA

For pathologists, the central question is practical: how can ecDNA be detected? And can that detection fit into existing diagnostic workflows?

Several approaches are available, each with different trade-offs.

Conventional karyotyping can reveal double minutes – the cytogenetic footprint of ecDNA – in dividing cells. It is accessible but low-resolution, unable to characterize the content of the ecDNA or reliably distinguish it from other extrachromosomal structures.

Interphase FISH is more clinically practical. When oncogene-specific probes reveal multiple scattered signals in the nucleus outside the expected chromosomal regions, this pattern is consistent with ecDNA. Many diagnostic laboratories already have this capability. No additional infrastructure is required; only a trained eye and an appropriate index of suspicion.

Whole genome sequencing (WGS) is currently the most comprehensive approach. Paired with bioinformatics tools such as AmpliconSuite, WGS can reconstruct the architecture of ecDNA molecules, determining which genes they carry, how many copies are present, and how the circular structure is organized. This is not yet routine, but as sequencing costs fall, it is becoming increasingly accessible.

Long-read sequencing platforms offer even greater resolution, resolving complex junctions and repetitive regions that standard sequencing misses. Optical genome mapping provides an alternative approach for identifying circular structural signatures without sequencing – potentially offering a practical middle ground between FISH and WGS.

The path to integrating ecDNA detection into standard diagnostic practice is clearer than it may initially appear. For tumor types with well-established ecDNA associations – glioblastoma, neuroblastoma, liposarcoma, and others – reflex testing for ecDNA signatures could be embedded within existing molecular panels. In many cases, the information is already latent in data we are collecting. We simply need to look for it.

Future perspectives

ecDNA is a young field. Key questions remain unanswered. The answers will determine how quickly the discoveries of the past decade translate into clinical benefit.

We still do not fully understand why ecDNA forms in the first place. The underlying mechanisms are increasingly well characterized. But the factors that predispose a tumor cell to generate ecDNA rather than relying on other forms of amplification are poorly understood. Identifying those drivers could open routes to early interception.

We also need better tools to track ecDNA over time. Single-cell sequencing is beginning to reveal how ecDNA populations shift during treatment and disease progression. But this remains largely a research tool. The development of clinical-grade assays for ecDNA monitoring, including from circulating tumor DNA in liquid biopsy, would transform our ability to track resistance in real time.

On the therapeutic side, BET inhibitors, ATR–CHK1 pathway targeting, and other ecDNA-directed strategies are still largely in preclinical or early-phase investigation. Rigorous clinical trials with ecDNA status as a defined stratification criterion are essential to establish which patients benefit.

Most importantly, ecDNA needs to become part of the conversation in oncology diagnostics. To fully benefit from what we now know, we must prioritize its integration into comprehensive genomic profiling panels. That will facilitate the development of ecDNA-specific clinical trial stratification criteria. Meanwhile, research should continue to expose the determinants of ecDNA formation.

The pathologist has always been the person who looks most closely at the tumor. ecDNA is one of the most powerful things we have not yet been looking at. That needs to change.

Newsletters

Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.

I have read and understand the Privacy Notice*

Subscribe