Page 752 - Read Online

Targher et al. Hepatoma Res 2020;6:64 I http://dx.doi.org/10.20517/2394-5079.2020.71 Page 5 of 9

suggestive of steatohepatitis, who meet at least one of the following criteria: past or present evidence of
metabolic risk abnormalities that meet the criteria to diagnose MAFLD (as described in Figure 2) with at
least one of the following diagnostic criteria, namely documentation of MAFLD on a previous liver biopsy,
[5]
or historical documentation of hepatic steatosis by imaging methods . Notably, a history of past alcohol
[5]
intake should be considered as patients may have a dual disease aetiology with alcohol use disorder .
Finally, the panel of international experts has also proposed the new definition of MAFLD in order to
bring more clarity to the distinction between the diagnostic criteria used in clinical practice to define this
common liver disease and the inclusion criteria used for future research studies or randomized controlled
[5]
trials . However, these considerations in no way detract the conduct from ongoing trials, nor does it affect
the diagnostic criteria proposed for MAFLD. As also suggested by the panel of international experts, it is
reasonable to hypothesize that the use of the newly proposed “positive” criteria for the diagnosis of MAFLD
and the exclusion of individuals with hepatic steatosis who do not have coexisting metabolic dysfunction
will render study cohorts less heterogeneous, thereby increasing the probability of detecting a positive effect
[5]
of clinical approaches targeting MAFLD .

MAFLD and NAFLD: are two terms interchangeable?
It is possible that a consequence of implementing the proposed change in terminology will also highlight
some new categories of fatty liver disease. This might also foster new discoveries of the causes, mechanisms,
[5]
classification and treatment of fatty liver disease .
In line with this hypothesis, Lin et al. have recently compared the characteristics of MAFLD and NAFLD
[32]
in a nationally representative cohort of nearly 13,000 United States adults with liver ultrasonography and
complete laboratory data. Interestingly, using the National Health and Nutrition Examination Survey
(NHANES-III) 1988-1994 database, the authors reported that NAFLD was present in 33.2% of participants,
while MAFLD was present in 31.2%. In addition, compared to those with NAFLD, patients with MAFLD
were more likely to have multiple metabolic comorbidities (e.g., obesity, hypertension and type 2 diabetes)
[32]
and more cases with advanced liver fibrosis as detected by non-invasive fibrosis scores . These findings
suggest that MAFLD definition is more accurate than NAFLD definition in identifying those patients
with hepatic steatosis who are at higher risk of disease progression or have a greater risk of cardiovascular
disease.

As depicted in Figure 3, these “real-world” data clearly indicate that there is an excellent concordance
between MAFLD and NAFLD (with a calculated Cohen’s kappa coefficient of 0.92) . However, NAFLD
[33]
and MAFLD definitions do not identify exactly the same individuals and, therefore, the two terms are not
fully interchangeable. Looking at these “real-world” data from adult individuals in the United States, it is
evident that not all individuals with NAFLD have MAFLD and vice versa. Indeed, as shown in Figure 3,
there are some individuals with MAFLD and coexisting liver diseases who do not have NAFLD. There
is also a non-negligible proportion of individuals (620/4,347; nearly 15%) previously deemed as having
NAFLD who do not have MAFLD, and that cannot be attributed to significant alcohol consumption (based
on self-reported history that might also be under-reported in clinical practice) or other known causes of
[32]
[32]
liver disease . Notably, as reported in the study by Lin et al. , these latter individuals are predominantly
of female sex (~60% women), young (mean ± SD: 35 ± 13.4 years) and have “lean” NAFLD (mean body
mass index: 21.7 ± 2.1 kg/m ) without coexisting metabolic dysregulation, as defined by the diagnostic
2
criteria of MAFLD. Most importantly, these individuals also have a (relatively) low proportion of advanced
liver fibrosis . Unfortunately, no information is available about patatin-like phospholipase domain-
[32]
containing protein-3 (PNPLA3) I148M variant, trans-membrane-6 superfamily member 2 (TM6SF2) E167K
variant or other NAFLD-related genetic variants in these “lean” NAFLD individuals without metabolic
dysregulation.