📄 medicine

Circulating Cell-Free DNA fragmentomics distinguishes pancreatic ductal adenocarcinoma from non- malignant conditions

This prospective study demonstrates that analyzing circulating cell-free DNA fragmentomics, specifically shorter fragment lengths and higher concentrations, can effectively distinguish pancreatic ductal adenocarcinoma from chronic pancreatitis and healthy controls, offering a promising minimally invasive diagnostic approach.

Original authors: Romualdas Riauka, Darius Stukas, Arenida Bartkeviciene, Inga Zievyte, Sandra Ivanauskiene, Rita Kupcinskaite, Ingrida Grabauskyte, Zilvinas Dambrauskas, Giedrius Barauskas, Povilas Ignatavicius

Published 2026-06-28

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Romualdas Riauka, Darius Stukas, Arenida Bartkeviciene, Inga Zievyte, Sandra Ivanauskiene, Rita Kupcinskaite, Ingrida Grabauskyte, Zilvinas Dambrauskas, Giedrius Barauskas, Povilas Ignatavicius

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Problem: Finding a Needle in a Haystack

Imagine the pancreas is a small, hidden factory inside the body. Sometimes, this factory gets damaged by a very dangerous "saboteur" called Pancreatic Ductal Adenocarcinoma (PDAC). This is a type of cancer that is hard to find early because it doesn't show obvious symptoms until it's too late.

Currently, doctors have a hard time telling the difference between this dangerous cancer and a less dangerous condition called Chronic Pancreatitis (which is like a chronic, long-term inflammation or "irritation" of the factory). It's like trying to tell the difference between a factory that is just having a noisy, messy renovation (pancreatitis) and one that is secretly being destroyed by a criminal gang (cancer). Standard blood tests often fail to make this distinction clearly.

The New Idea: Listening to the "Shredded Papers"

The researchers from the Lithuanian University of Health Sciences decided to look at something called Cell-Free DNA (cfDNA).

Think of your body's cells as libraries filled with books (DNA). When cells die or get damaged, they break apart, and tiny shreds of these books float out into the bloodstream. This is the cfDNA.

The Theory: If a factory is being destroyed by a criminal gang (cancer), the shreds of paper floating in the air might look different than if the factory is just having a messy renovation (inflammation).

The team wanted to see if they could use these floating paper shreds to tell the difference between cancer patients, pancreatitis patients, and healthy people.

How They Did the Experiment

They gathered 83 people for a study:

56 people who had confirmed pancreatic cancer.
12 people who had chronic pancreatitis.
15 healthy people with no pancreatic issues.

Before any surgery or treatment, they took a blood sample from each person. They then looked at the "shredded papers" (cfDNA) in the blood using special high-tech microscopes and computers. They measured three main things:

How long were the shreds? (Fragment length)
How many shreds were there? (Concentration)
Were there any typos in the text? (Mutations/SNPs)

What They Found

The results showed two very clear differences between the cancer group and the others:

1. The "Shred Size" Difference

The Analogy: Imagine a shredder. A healthy shredder makes medium-sized confetti. A cancer shredder is broken and makes tiny, dust-like particles.
The Result: The people with cancer had shorter DNA fragments (average 292.7 units long) compared to the pancreatitis group (313.9 units) and healthy people (315.7 units). The cancer DNA was significantly more "crumbled."

2. The "Shred Quantity" Difference

The Analogy: If a factory is collapsing, you might expect to see a lot more debris floating in the air than if it's just being renovated.
The Result: The cancer patients had more DNA floating in their blood (45.6 ng/mL) compared to the pancreatitis group (21.8 ng/mL) and healthy people (20.2 ng/mL).

3. What Didn't Work

The researchers also looked for "typos" (mutations) in the DNA text. They found that counting these typos or looking for specific genetic errors did not help distinguish cancer from pancreatitis or healthy people. The "typos" were too similar across all groups to be useful for this specific test.

How Good Was the Test?

The researchers used a statistical tool (like a grading curve) to see how well these measurements could sort the people into the right groups.

Fragment Length: It was pretty good at spotting cancer. If the DNA shreds were shorter than a certain size, it was likely cancer.
Concentration: It was also helpful. If there was a lot of DNA floating around, it was more likely to be cancer.

The Catch (Limitations)

The authors were honest about the limits of their study:

Small Group: They only had 83 people, and the pancreatitis group was quite small (only 12 people). This is like trying to judge a whole city's weather based on a few days of data.
Not Perfect: While the test was good, it wasn't perfect. There was some overlap, meaning a few healthy people might look like they have cancer, or vice versa, if you only look at one person.
Other Causes: Things like infections, trauma, or other illnesses can also change the amount of DNA in the blood, so this test isn't a magic bullet that works 100% of the time on its own.

The Bottom Line

This study suggests that looking at how short the DNA fragments are and how much DNA is in the blood could be a new, less invasive way to help doctors tell the difference between pancreatic cancer and chronic pancreatitis.

However, the authors emphasize that this is still an early step. They need to test this on many more people before it can be used as a standard tool in hospitals. They also found that counting genetic "typos" wasn't helpful for this specific job.

Technical Summary: Circulating Cell-Free DNA Fragmentomics in Pancreatic Ductal Adenocarcinoma

Problem Statement
Pancreatic ductal adenocarcinoma (PDAC) is characterized by extremely poor survival rates, largely due to late-stage diagnosis and a lack of effective early detection methods. Current serum biomarkers, such as CA 19-9 and CEA, suffer from suboptimal sensitivity and specificity, particularly when differentiating PDAC from benign inflammatory conditions like chronic pancreatitis (CP). While liquid biopsy offers a minimally invasive alternative, tumor-derived circulating cell-free DNA (cfDNA) often exists at low concentrations in early disease, limiting its diagnostic utility. This study addresses the need for reliable, non-invasive biomarkers capable of distinguishing PDAC from CP and healthy individuals by investigating whether cfDNA fragmentomic features—specifically fragment length and concentration—provide superior discriminatory power compared to traditional mutation-based metrics.

Methodology
The research was conducted as a prospective, single-center study at the Lithuanian University of Health Sciences Hospital Kaunas Clinics. The study cohort consisted of 83 participants divided into three groups:

PDAC Group: 56 patients with histologically confirmed PDAC, recruited prior to any surgical or chemotherapeutic intervention.
Chronic Pancreatitis (CP) Group: 12 patients with confirmed CP and no concurrent oncological disease.
Healthy Control Group: 15 volunteers with no history of oncological disease.

Peripheral venous blood (10 mL) was collected in EDTA tubes and processed via double centrifugation to isolate plasma. cfDNA was extracted using the QIAamp Circulating Nucleic Acid Kit. The study analyzed four primary parameters:

cfDNA Concentration: Measured using Qubit™ dsDNA High Sensitivity and Broad Range Assay Kits.
Fragment Length: Determined via Oxford Nanopore Technologies sequencing (PromethION 2 device) using the Native Barcoding Kit. Data processing involved basecalling with Dorado (HAC model) and analysis via the Nextflow Nanoseq pipeline (including Minimap2 for mapping and Deepvariant for mutation calling).
Total SNP Counts: Aggregated single nucleotide polymorphism counts.
Clinically Relevant Mutations: Annotated using Open Cravat and the ClinVar database, filtering out synonymous and intronic variants.

Statistical analysis utilized non-parametric tests (Mann-Whitney U, Kruskal-Wallis H) due to small sample sizes. Diagnostic performance was evaluated using Receiver Operating Characteristic (ROC) curve analysis.

Key Results
The study identified distinct differences in cfDNA physical characteristics between the groups, while mutation-based metrics showed no significant separation:

Fragment Length: PDAC patients exhibited significantly shorter cfDNA fragments compared to both CP patients and healthy controls.
- PDAC median: 292.7 bp.
- CP median: 313.9 bp ( $p = 0.047$ ).
- Controls median: 315.7 bp ( $p = 0.003$ ).
Concentration: cfDNA concentration was significantly elevated in the PDAC group.
- PDAC median: 45.6 ng/mL.
- CP median: 21.8 ng/mL ( $p = 0.026$ ).
- Controls median: 20.2 ng/mL ( $p = 0.002$ ).
Mutation Metrics: No statistically significant differences were found in total SNP counts or the number of clinically relevant mutations among the three groups. Furthermore, within the PDAC cohort, fragment length, concentration, and mutation parameters did not correlate with tumor location, resectability, metastasis, histological grade, or survival outcomes.
Diagnostic Performance (ROC Analysis):
- Fragment Length: Showed moderate accuracy in distinguishing PDAC from CP (AUC 0.770; cutoff 311.65 bp) and from healthy controls (AUC 0.820; cutoff 294.5 bp).
- Concentration: Also demonstrated moderate accuracy, with an AUC of 0.737 for PDAC vs. CP (cutoff 23.65 ng/mL) and 0.796 for PDAC vs. controls (cutoff 38.7 ng/mL).

Significance and Claims
The authors claim that cfDNA fragmentomics, specifically the analysis of fragment length and concentration, offers a promising, minimally invasive approach to differentiating PDAC from chronic pancreatitis and healthy individuals. The study suggests that these physical features of cfDNA may provide better discriminatory power than total SNP counts or the number of clinically relevant mutations, which failed to show significant differences in this cohort.

The paper posits that shorter fragment lengths and higher concentrations in PDAC patients may reflect increased tumor cell apoptosis and necrosis, as well as higher tumor burden. The authors highlight the potential utility of these biomarkers in scenarios where radiological examinations are inconclusive. However, the authors maintain a modest stance regarding clinical implementation. They explicitly acknowledge limitations, including the small sample size, the exploratory nature of the molecular analysis due to varying plasma volumes, and the lack of disease specificity (as cfDNA levels can be influenced by non-tumor conditions like trauma or sepsis). Consequently, the paper concludes that while these findings support the potential of liquid biopsies, further development and validation are required before routine clinical use can be recommended.