In recent years the most rapidly emerging disease in general hepatology is the Metabolic
dysfunction-Associated Fatty Liver Disease (MAFLD) [1]
[2]. This a new definition, which was suggested to replace and extend what was previously
known as Non Alcoholic Fatty Liver Disease (NAFLD) and Non Alcoholic Steatohepatitis
(NASH) [1]. The new proposal of MAFLD was set forward because the criterion for diagnosing
NAFLD/NASH requires absence of any other chronic liver condition (e. g. viral hepatitis,
immune-related cholestatic liver disease, etc) and absence of any over threshold intake
of alcohol [1]
. The terms NAFLD/NASH are correct to establish a diagnosis of liver disease purely
caused by metabolic dysfunction, useful to better understand the natural history of
the disease, especially in a scientific setting, but does not reflect the real-life
situation, where many patients may actually suffer from metabolic liver disease on
top of other liver conditions and the two conditions may or may not both contribute
to the progression and complications of the liver disease with different and not well
definable relevance [3]. For instance, the persistence of minimally elevated liver enzymes in slightly overweight
patients after cure of HCV or after full dosage of ursodesoxycholic acid in Primary
Biliary Cholangitis or Primary Sclerosing Cholangitis or after drastic reduction or
the suspension of alcohol intake often raise the question whether an underlying MAFLD
is present. This adds obviously to the question of whether some elevation of liver
enzymes in the absence of any known etiological factor may be due to NAFLD when ultrasonography
does not detect an obvious bright liver. In all these and in other situations the
precise assessment of liver fat content may be crucial, when this is not obvious on
conventional ultrasonography.
Historically and still at present the reference standard for the assessment of an
excess of liver fat content, for establishing presence and grade of inflammation (i. e.,
hepatitis) and degree of consequent fibrotic deposition (staging) is histology [4]
[5]. In the absence of an excess of liver fat content (histologically set at > 5 % of
hepatocytes) [5] a contribution of metabolic fat conditions to liver abnormalities can be ruled out.
However, biopsy is a (mini-)invasive procedure and thus hardly applicable to very
large number of symptomless patients. Therefore, non-invasive techniques have been
developed trying to answer any possible question otherwise requiring biopsy. In this
issue of the journal the article by Bauer and colleagues [6] reports on the use of shear wave ultrasound elastography methods to screen patients
with NAFLD for the presence of significant fibrosis or cirrhosis, a topic which has
been the focus of some relevant publications to which Bauer adds another significant
piece of information. However, we would like to bring to attention here the value
of research in the field of the step before fibrosis quantification in NAFLD patients,
which is whether there is an excess of fat content and to which extent. Do we need
to have this information?
In simple words, do we need liver fat quantification?
-
The extent of steatosis is commonly evaluated and reported semi-quantitatively on
bioptic specimens. The most reproducible method follows the acinar architecture dividing
the liver parenchyma and assessing percentage involvement by steatotic hepatocytes
according to three steatotic classes: 5 %–33 %-mild, 33 %–66 %-moderate or > 66 %
– severe steatosis, being 0 %–5 % the condition of no steatosis. Various ultrasound-based
methods to assess liver steatosis exist, which are extensively described in a recently
published position paper [7]. Most of the methods have been compared to histologically established liver steatosis,
showing in general a significant correlation between ultrasound parameters and semi-quantitative
grades of steatosis, with however a significant overlap between ultrasound estimated
classes [8]. This also implies that such methods cannot be utilized to definitively establish
a steatotic grade at present, but most importantly also not to establish whether the
patient has no steatosis or rather mild steatosis in dubious cases (generally corresponding
to a fatty infiltration of only 5 % to 15 %, sufficient to cause liver enzymes abnormalities
in predisposed patients, but hardly recognizable at ultrasonography). At present therefore
the need to accurately distinguish no steatosis from mild steatosis in dubious cases,
which would be a relevant need, remains insufficiently satisfied by currently available
ultrasound-based liver fat quantification methods, due to the overlap in measures
between no or very mild steatosis.
-
Another critical step is that the reproducibility of the demonstration of ultrasound
methods to precisely measure changes in liver fat content over time has not been provided
so far, since very difficult to achieve. In fact, when patients lose weight (or gain
weight) ultrasound methods may display changes in quantification variables. However,
whether this corresponds strictly to histological changes of fat content has been
simply assumed for true, but never definitively demonstrated, since usually patients
are not submitted to repeat biopsy to demonstrate changes in fat content, thus preventing
the possibility of achieving solid scientific evidence. This may only happen in pharmacological
experimental trials, in which, however, another variable (drug treatment) is in place.
Such possibility, namely to declare a decrease or an increase in liver fat content,
would be a good tool to motivate patients to change unhealthy life-styles or reinforce
the benefits of the adopted changes. We have all experienced patients coming to standard
ultrasonography exams after having lost weight, who put the question to the operator:
“is my liver less steatotic now”? or when they gain weight they whisper “has my liver
become more steatotic”? This question is currently inaccurately answered by the visual
impression of the operator.
-
In keeping with the latter issue of changes over time and in connection with the need
of liver fat quantification we would like to remind that extremely few drugs have
provided some evidence of clinical benefit in NASH so far (and mainly on histology,
without long term clinical endpoints). The difficulties in achieving positive trials
in this field is so big, that new forms of studies are envisioned [9]. For these few effective drugs (and so far none of them widely accepted for reimbursement
by National Agencies in Europe) there is no evidence that the decrease in liver fat
content can be utilized as a surrogate marker of drug efficacy, limiting the need
for liver fat quantification. However, if a new drug will ever arrive, providing clinical
benefit to an extent to make it recommended in clinical practice for NASH, and changes
in liver fat content will be demonstrated (which is not unlikely) to be a good biomarker
associated with response, then the need for non-invasive ultrasound quantification
of liver fat content will become very strong. Therefore, trying to develop reliable
ultrasound methods for liver fat quantification now is desirable, in order that we
will be found prepared then. In fact, there would be insufficient availability of
MRI scanners to satisfy such need with alternative methods, such as MRI-PDFF (Magnetic
Resonance Imaging-estimated Proton Density Fat Fraction) [8].
-
In addition to establishing the presence of MAFLD, clarifying to presence or absence
of significant or advanced fibrosis is crucial for the best management of such patients.
To this end liver ultrasound elastography methods are reliable and convenient methods
[10], as also shown by Bauer in this issue of the journal [6] and thus recommended. Furthermore, liver stiffness values contribute to build accurate
prognostication score for liver disease evolution [11]. However, Petta and colleagues have demonstrated that higher degrees of liver fat
infiltration increase the risk of overestimation of liver fibrosis assessed by ultrasound
elastography [12] and that the combination of liver fat quantification with elastography makes the
non-invasive liver staging in NAFLD patients more accurate [13]. This might therefore be held as a clinical need for better ultrasound fat quantification,
especially whenever this will become confirmed also with methods embedded in conventional
ultrasound equipment.
At last, whenever the need for ultrasound liver fat quantification will take a definitive
place, also potentially because of adequate quantification modalities, as illustrated
above, attention will have to be paid to individual methods and units expressing the
information. In fact, different modalities exist. Some methods, such as the UDFF (Ultrasound-Derived
Fat Fraction), quantify liver fat content in terms of percentages [8] similarly to the MRI based PDFF methods. Worth to outline, that such percentages
do not correspond to histological percentages. For instance, an UDFF value of 25–30 %
suggests the presence of already severe histological fibrosis (but corresponding to
> 66 % fat hepatocytes at histology). Other methods quantify the acoustic attenuation
and express the values in terms of dB/m [12]
[13] with values falling in the range of approximately 150–300 dB/m, while others use
dB/cm-MHz [8], with values ranging 0.5–1.5 dB/cm-MHz. Other methods use ultrasound speed estimation
measured in mm/microsec, with values usually in the range of 1500–1600 mm/microsec
[14], making the field of liver fat quantification methods animated. Most likely the
methods will not be interchangeable to satisfy the clinical needs and evidence should
preferably be provided for each technique.
To conclude, the needs for ultrasound liver fat quantification are limited at present,
also in view of the potential of the current methods, but they may become much requested
in the future and the ultrasound community and industry must let themselves be found
prepared.
