The release is an overview of a recent journal article that reports findings on a new blood test that detects liver fibrosis — or scarring of the liver — based on differences in “DNA methylation.” The release says the new technique can also detect the severity of the fibrosis in patients. The release does not, however, discuss cost, potential harms (such as the possibility of misdiagnosis), how this technique may be better than other early diagnostic techniques (such as using magnetic resonance enterography [MRE]), or adequately discuss the limitations of the study. It also engages in some fear-mongering and unjustifiable language.
Non-alcoholic fatty liver disease (NAFLD) is not uncommon. An aptly named 2011 paper in the BMJ, “How big a problem is non-alcoholic fatty liver disease?,” points to existing research that states “94% of obese patients…67% of overweight patients…[and] 25% of normal weight patients” in a European population had NAFLD. The paper also notes that “The overall prevalence of NAFLD in people with type 2 diabetes ranges from 40% to 70%.” A 2008 paper estimated NAFLD prevalence at between 20 and 30 percent in developed countries. NAFLD, in itself, is “often considered a relatively benign condition,” according to the 2008 paper. But it can also be a precursor to more serious conditions, such as cirrhosis. In depth analysis of the condition has historically hinged largely on liver biopsies, which require invasive procedures. However, there are now a number of techniques aimed at collecting liver fibrosis data at various stages of the condition’s development. While this release compares the new technique to the use of biopsies, it’s not clear how it compares to any of the other techniques that are in various stages of development.
The technique is likely not at the stage of commercial availability so costs would be uncertain. Still, cost isn’t mentioned at all, even in an abstract way. Could this be comparable to the cost of existing diagnostic tools? Cheaper? More expensive? There’s still much work to do to demonstrate the value of this test.
The benefits here are not clear, much less quantified. At one point, the release quotes a researcher as saying: “We know that the presence of even mild fibrosis of the liver predicts a worse long-term outcome for patients with NAFLD and so it’s important to be able to detect liver scarring at an early stage.” But why is that important? Are there therapeutic treatments or interventions available to change the course of the disease or otherwise affect the patient’s health and well-being? Lower down, a second researcher is quoted as saying “If we are able to accurately tell the extent of a person’s liver damage with a blood test, and even track the scarring as it gets better or worse, it could provide reassurance for patients, save NHS [National Health Service] resources and avoid patients having to undergo a liver biopsy.” So, the benefit is that patients could avoid having a biopsy? Is that the benefit?
Without knowing the prognostic value of the test results, it’s hard to see how it could provide reassurance for patients. As noted above, the value of knowing this diagnosis in terms of directing targeted therapy is uncertain and the authors actually note that their data do not suggest superiority to existing fibrosis biomarkers.
There’s no discussion of potential harms. How accurate is the new blood test? Is there a risk of false positives or misdiagnosis — and, if so, how high is that risk? The investigators assembled only a small number of carefully selected patients. Many were excluded if they had other liver diseases as well as complications from alcohol. And the as-yet undetermined potential for both false positive and false negative results could certainly undermine its value in reassuring patients.
The release notes that the study looked at 26 patients with NAFLD. There was no mention of the control group included in the paper. There was no discussion of the study design. There was no mention of the test’s accuracy. There was no discussion of the study’s limitations, which are stated clearly in the paper itself. For example: “Clearly, the sample size is relatively small, with no independent validation within a separate cohort.”
According to the published research, the investigators actually studied whether the new test could more accurately detect advanced fibrosis than the currently used NAFLD fibrosis score. They also found that their test could accurately detect fibrosis in patients with alcoholic liver disease–here with only 13 subjects. Controls for these analyses had no evidence of liver disease, but we can’t tell whether they had liver biopsies or just provided blood–suspect the latter.
As noted above, NAFLD is both fairly common and — for most patients — is not cause for undue alarm. In a small number of patients, it can progress to more serious conditions. That is not clear from the release. Readers who know little about NAFLD could be forgiven for thinking that NAFLD is likely to progress to cirrhosis and liver failure. A little bit of extra context would have been good here; for example, noting that approximately 5 percent of patients with NAFLD are likely to deal with long-term liver failure. Is NAFLD something that should be taken seriously? Absolutely. But it needs to be discussed responsibly.
Funding sources are clearly identified, as are researchers who are quoted and their affiliations related to the work. There do not appear to be any relevant conflicts of interest.
The release discusses the new technique and liver biopsies. It doesn’t discuss other techniques, such as the use of MREs or other noninvasive diagnostics (like those discussed here or here or here). How does the proof-of-concept approach discussed in this news release compare to any of these other techniques? It’s not clear.
Imaging tests including ultrasound, CT and MRI can detect fatty changes in the liver, although not necessarily fibrosis. The NAFLD fibrosis score used by the investigators as a comparison test is probably the most useful for detecting fibrosis.
The release includes a quote that says “We are now working on confirming these findings in a larger group of patients.” Based on that, readers can infer that the technique is not yet available — so it gets a pass. However, it would have been better to state that more explicitly. As the paper itself says: “With validation, this blood-based biomarker could become an important contributory clinical tool alongside other epigenetic, genetic and biochemical biomarkers, mitigating the future need for biopsy to evaluate fibrosis.”
The release makes clear that this is the first diagnostic tool to assess fibrosis based on DNA methylation, which it describes as “chemical changes on tiny amounts of ‘cell-free’ DNA that are released into the blood when liver cells are injured.”
“…this could be a replacement for a liver biopsy.” That line is just under the headline, and it’s an overreach. It could just as easily have said that this may not be a replacement for a liver biopsy — particularly given that the study looked at only 26 patients. Here’s the thing: this research is really interesting. It could stand on its own, with all of the relevant qualifiers, and still be fascinating. It highlights some interesting things about how the human body works, how it responds to disease, and how researchers can monitor those changes to understand what’s happening in our bodies. The release didn’t need to go over top, but it did.
A couple other examples of unjustifiable language:
“This scientific breakthrough has great promise because the majority of patients show no symptoms,” according to a quoted researcher. In the preceding paragraphs the test under development is described as “proof of principle research,” and “this first stage of research,” and it’s noted that the test is now being confirmed in a larger group of patients. We cry foul on the use of “breakthrough” for this preliminary work.
“…[G]reat promise because the majority of patients show no symptoms.” We don’t know that asymptomatic patients will benefit from early detection.