Community prevalence of advanced liver fibrosis in Type 2 diabetes—How big is the challenge?

Abstract

Editorial

Current population estimates suggest over a third of adults in the world have hepatic steatosis, and the majority would meet criteria for metabolic dysfunction-associated steatotic liver disease (MASLD) [1]. This was previously termed non-alcoholic fatty liver disease until an international multi-society consensus to change the name in 2023 [2] to facilitate a proactive diagnosis rather than being a diagnosis of exclusion.

However, only a small proportion of patients with MASLD will have a major adverse liver outcome in their lifetime, such as decompensated cirrhosis or developing a hepatocellular carcinoma (HCC). The major risk for adults with hepatic steatosis is a major adverse cardiovascular event (MACE), particularly myocardial infarction, with an increased lifetime odds of coronary artery disease rising by 33% compared to adults without steatosis [3].

The counterfactual is MASLD is a leading indication for liver transplantation across Europe and North America, in part due to the success of the Hepatitis C eradication programmes internationally. MASLD is also a rapidly rising cause of liver-related deaths in the WHO global burden of disease study, behind alcohol-related liver disease and viral hepatitis. Modelling of trajectories of liver disease–related deaths from seven international liver disease cohorts suggests MASLD related deaths will have doubled between 2015 and 2030 [4]. It is also driving the rise in HCC in non-cirrhotic liver disease across the western world, with one multicentre French cohort reporting 72% of non-cirrhotic HCC diagnoses were made incidentally over an 11-year period [5].

Crucially, fibrosis progression rates (FPRs) in MASLD are slow, and addressing cardiometabolic risk factors can halt—and even reverse—progression whilst improving all-cause morbidity and mortality from MACE. Importantly, fibrosis is the only histological feature which is known to determine the risk of liver-related events, with progressive stages of fibrosis conferring the highest risk of adverse outcomes [6, 7]. Therefore, identifying patients with MASLD with advanced fibrosis is pivotal in terms of risk stratification. Emerging data suggested that Type 2 diabetes (T2D) is one of the strongest potentiators of fibrogenesis, and FPRs in those with T2D are enhanced [8]. However, a substantial proportion of this evidence comes from clinical trials or those attending secondary care diabetes services; thus, real-world evidence, particularly from primary care settings, is to be welcomed.

In this context, Balkhed et al. provide valuable normative data when screening for MASLD amongst T2D patients in the primary care setting [9]. Steatosis prevalence in T2D has historically been reported at 50%–70% [10], and multiple societies such as the European Association for the Study of the Liver and the American Diabetes Association support screening for MASLD and fibrosis in patients with T2D. Thus, proactive liver fibrosis screening in this higher risk group is supported.

In screening 308 patients with T2D in the community, the prevalence of steatosis using the non-invasive gold standard of MRI proton density fat fraction was 54.6%. Overall, 7% of patients had evidence of advanced fibrosis based on transient elastography (≥10 kPa) [9]. As expected, obesity amongst patients with T2D was strongly associated with odds of MASLD advanced fibrosis in multivariable logistic regression modelling. It is important for readers to consider different obesity thresholds that would need to be applied based on the ethnicity mix of the population you serve when considering risk [11]. Additionally, in this Swedish cohort of T2D patients, the proportion of patients with lean MASLD (with a normal BMI <25 kg/m²) would have been lower compared to if this study had been performed in a comparable South Asian or Southeast Asian cohort.

Balkhed et al. provide a valuable insight on Phosphatidylethanol (PEth) testing in this MASLD cohort [9]. There are persisting concerns regarding alcohol disclosure that patients admit to or can accurately recall, leading to significant discordance when ascribing a diagnosis of MASLD compared to MetALD (MASLD with greater alcohol consumption) [12]. PEth testing is now arguably the best available non-invasive test (NIT) for corroborating alcohol history, able to reflect alcohol history of the previous 4 weeks, although it is most accurate in the preceding 4–10 days [13]. Less than 1 in 20 patients tested had an elevated PEth test, whereas no patients with advanced fibrosis had an elevated measurement. The authors acknowledge their findings must be interpreted in the context of healthy volunteer bias, which challenges the generalizability to wider patient populations.

A total of 211 patients within the cohort had cardiac MRI, demonstrating reduced left ventricular stroke volume index in patients with MASLD compared to those without. These findings replicate findings in other cohorts, with such cardiac changes noted in non-diabetic patients with MASLD as early as their mid-twenties [14]. Left ventricular mass index was noted to be lower in patients with MASLD fibrosis compared to T2D patients without MASLD, further illustrating the multisystem impact of MASLD within the wider metabolic syndrome.

Balkhed et al. were able to offer multiple NITs for liver fibrosis to their T2D cohort and illustrated the challenges with screening patients for liver fibrosis in the low-prevalence setting. Almost all NITs would have been created and validated in a high disease prevalence setting in secondary care. Applying those tests in the low-prevalence setting leads to a spectrum effect with a fall in sensitivity. The authors challenged the negative predictive value of FIB-4 (a composite score of age, platelet count, alanine transaminase and aspartate aminotransferase) for ruling out advanced fibrosis, with FIB-4 being increasingly shown to have diminished utility in those with T2D [15]. Similarly, it is recognized that severe and morbid obesity may also falsely elevate liver stiffness measurements (LSMs) as an estimate of liver fibrosis [16].

The findings for NIT performance should be contextualized. First, FIB-4 has been repeatedly demonstrated to have good negative predictive value in multiple international cohorts for MASLD. Concordant NITs, for example, a positive FIB-4 and transient elastography, increase the diagnostic accuracy for identification of advanced fibrosis. It is telling that of 21 patients with an elevated LSM in the community who agreed to a liver biopsy, 4 patients had no fibrosis at all. This illustrates the challenges with using transient elastography in the low disease prevalence setting and the value of utilizing concordant NITs to risk stratify patients.

The authors are to be commended for undertaking a comprehensive study which adds to the growing literature providing robust real-world evidence of the prevalence of advanced fibrosis in persons with T2D in primary care.

So, the questions remain: Have we over-estimated the true prevalence of advanced fibrosis in T2D patients based on enriched populations from other studies? Do community-based prevalence studies overly depend on self-selecting patients who are willing to enrol and are not representative of true prevalence in a population? The reality is probably somewhere in between both of these extremes. Multi-ethnic cohort studies are needed to phenotype heterogeneity in population characteristics to determine true background risk of advanced fibrosis amongst patients with diabetes.

KWMA has received consultancy fees from Novo Nordisk and lecture honoraria from Advanz outside of the submitted work. PNB discloses consultancy fees for Novo Nordisk and Resolution Therapeutics and educational honoraria from Takeda and Novo Nordisk outside of the submitted work.