Defining 'Normal': Reference Intervals and Biological Variation in Dogs

How much cfDNA is too much? We review the data on healthy baselines, the impact of age and breed, and why '0.0' is not the goal.
Defining "Normal": Reference Intervals and Biological Variation in Dogs
A common misconception among veterinarians new to liquid biopsy is that a healthy patient should have zero cell-free DNA (cfDNA).
In reality, cfDNA is a normal physiological byproduct. Every day, billions of cells in a dog’s body reach the end of their lifespan and undergo apoptosis (programmed cell death). The liver and kidneys work constantly to clear this debris. The cfDNA level we measure in the blood is the net balance between this production (cell death) and clearance (organ function).
To interpret a test result, we must first define what "normal" looks like.
The Healthy Baseline
While reference intervals vary slightly by methodology (e.g., fluorometry vs. qPCR), multiple studies have converged on a consistent range for healthy, adult dogs.
* Typical Range: 0.5 – 1.2 ng/mL (or 500 – 1,200 µg/L).
* The Cutoff: Most studies consider values consistently above 1.5 – 2.0 ng/mL to be abnormal.
For example, in a study by Kim et al. (2021), healthy dogs had a mean cfDNA concentration of approximately 0.88 ng/mL. In contrast, dogs with lymphoma averaged over 2.0 ng/mL, with some advanced cases reaching levels 10x or 100x higher.
Factors Influencing Baseline
Before flagging a result as "pathological," consider these biological variables:
1. Age
Older dogs tend to have slightly higher baseline cfDNA than puppies or young adults. This is attributed to "inflammaging"—the low-grade, chronic inflammation associated with aging—and a higher rate of cellular senescence. A 12-year-old Golden Retriever might run at 1.0 ng/mL, whereas a 2-year-old might run at 0.4 ng/mL. Both can be considered healthy, but the "buffer" for the older dog is smaller.2. Exercise
If a client brings in a working dog (e.g., a Border Collie or hunting dog) immediately after a field trial, hold off on the draw. The Data: A study by Hunt et al. (2017)* showed that strenuous exercise could transiently elevate cfDNA levels by ~20%.
* Why? Exercise causes minor tissue micro-trauma and oxidative stress, leading to a burst of cell turnover.
* Recommendation: Rest the dog for 24 hours before collecting a sample for baseline monitoring.
3. Breed and Size
Currently, there is no strong evidence that breed significantly alters cfDNA concentration per milliliter of plasma. While a Great Dane has more total blood volume (and thus more total circulating DNA) than a Chihuahua, the concentration remains relatively constant. However, high-risk breeds (Boxers, Bernese Mountain Dogs) are the primary candidates for screening, so establishing individual baselines for these patients during wellness visits is highly valuable.The "Gray Zone"
What do you do with a result of 1.6 ng/mL?
This is the diagnostic "gray zone." It is too high to be perfectly normal, but too low to be a definitive cancer signal (like the >5.0 ng/mL spikes seen in septic peritonitis or metastatic hemangiosarcoma).
Differential Diagnoses for the Gray Zone:
1. Mild Inflammation: Gingivitis, dermatitis, or early osteoarthritis.
2. Pre-analytical Error: Was the sample delayed in the mail? Did minor hemolysis occur?
3. Early Disease: A small, localized tumor or early-stage organ dysfunction.
The Strategy: Do not panic. Recheck the sample in 2–4 weeks.
* If it returns to <1.0 ng/mL, it was likely transient (e.g., a resolved minor injury).
* If it rises to 2.5 ng/mL, you have a trajectory. A rising trend is far more clinically significant than a single borderline value. This warrants a "confirmatory hunt"—detailed imaging and exams to find the source.
Summary
* Normal: < 1.0 ng/mL.
* Gray Zone: 1.0 – 2.0 ng/mL (Monitor trends).
* Abnormal: > 2.0 ng/mL (Investigate for neoplasia, trauma, or sepsis).
Understanding these baselines allows us to use cfDNA not just as a binary "yes/no" test, but as a nuanced physiological monitor.
(References: Kim et al. 2021; Hunt et al. 2017; Letendre & Goggs 2018)

