Alanine Aminotransferase: A Clinical and Regulatory Tool for Detecting Liver Injury–Past,Present, and Future
谷丙转氨酶：检测肝损伤的临床和调节工具—过去，现在和将来
JR Senior
Office of Pharmacovigilance and Epidemiology, Center for Drug Evaluation and Research, US Food and Drug Administration, Federal Research Center at White Oak,
Silver Spring, Maryland, USA. Correspondence: JR Senior ([email protected])
Received 8 May 2012; accepted 8 June 2012; advance online publication 1 August 2012. doi:10.1038/clpt.2012.108
Clinical pharmacology & Therapeutics
Assay of the serum activity of the enzyme alanine aminotransferase (ALT) has become the primary screening tool for detecting acute liver injury. But what does an elevated value mean? Not what it is too often mistakenly believed to indicate. It is not a test of liver function. It does not necessarily predict worse effects to come (in a given person). It is not
a valid measure of severity of liver injury or dysfunction. It is too unspecific to be reliable in screening for relatively rare effects on the liver. Although these are substantial limitations, ALT is a very useful biomarker if understood and used properly. It is important to consider how and why these erroneous concepts came to have such wide acceptance, and how
elevations of ALT activity for evaluating patients and subjects under study might be interpreted better.
[extracts]
A small set of serum chemical tests for liver injury or dysfunction is now considered “classic,” namely: the serum concentration of total bilirubin (TBL) and the serum activities of alkaline phosphatase (ALP), aspartate aminotransferase (AST),and ALT.
Of these tests, the elevated serum enzyme activities may indicate injury to hepatocytes or biliary cells, but only the bilirubin elevation indicates any loss of liver function. These tests have been “validated” and “qualified,” not by opinions of an expert committee, but by clinical use for more than six decades of successful application in millions of cases.
What Does All This Add Up To?
Since the 1978 Fogarty Conference and 1982 NCI toxicity criteria
proposal, vast amounts of data have been accumulated.
It is now obvious that elevations in serum enzyme levels alone
are not necessarily measures of clinical severity. Elevated levels
of ALT do not predict what will happen; they only indicate
what has happened up to the time of measurement. Serum
ALT enzyme activity levels are not measures of liver function.
It is not a function of the liver to regulate the levels of activity
of serum enzymes that have no function in the plasma but
represent leakage or release from damaged cells in which they
previously did have functions. In plasma, the levels of activity
are determined by the rates of release from cells and by their
rates of inactivation or degradation by nonspecific proteases.
To a greater extent the TBL or prothrombin time (or derived
international normalized ratio) are better measures of the
ability of the whole liver to perform its true functions, such
as clearing the plasma of bilirubin or synthesizing important
coagulation proteins.
I now propose that the gradings of severity (Table 1) published
since 1982 by the NCI as the Common Toxicity Criteria
for Adverse Events be considered instead as measures of urgency.
To do what?—(i) to repeat the tests immediately (in local laboratories
to avoid the delay involved in obtaining results when
specimens are sent to distant central laboratories); (ii) to confirm
the findings, and establish the direction and rapidity of
change; (iii) to initiate additional studies to clarify the probable
cause (Table 2) of the findings if they become worrisome; (iv)
to preserve serum samples for possible retrospective tests to be
done later as needed.
Elevations in the concentration levels of serum TBL, or of the
direct-reacting, conjugated bilirubin, are highly specific to liver
problems but are not very sensitive as markers; also, the elevation
often occurs only late in a disease process. The search for
new biomarkers should accept the combined biomarker of (ALT
and TBL) as the standard to surpass, rather than ALT alone.
Under the new regulations of 2010,45,46 it is the responsibility of
the sponsors of drug trials to supervise local investigators and
guide them on the procedure to follow when abnormal liver test
results are found; all data are required to be recorded in the case
report forms, both in local and central laboratories.
Despite the limitations on the predictive value of ALT elevations
in individual subjects, it has been found repeatedly that the
increased occurrence of such elevations in groups that are being
studied is highly predictive of what is later likely to be found
in other groups. This is an important point, the significance of
which is often misunderstood.47–49
Recommendations
Proposed here are some new interpretations relating to elevated
levels of ALT activity, and suggestions for how they might lend
themselves to better general use, to trigger discussions, objections,
or corrections, to stimulate debate, and to aim at consensus
and more consistent application:
1. The measurements of serum ALT (as well as AST and ALP)
are not tests of liver function, but of liver cell injury.
2. Whole-liver dysfunction is what really determines the severity
of the injury.
3. Serum bilirubin concentration does measure at least one
function of the liver, namely, to clear plasma of bilirubin,
and depends on whole-liver functional capacity.
4. The combined biomarker (ALT and TBL) has the high sensitivity
of ALT and the very great specificity of TBL, and is the
current standard to surpass for proposed new biomarkers.
5. Urgency (to repeat the test and initiate special study) should
replace severity in the NCI CTC guideline relating to serum
enzyme activity levels.
6. Elevated ALT activity does not necessarily predict what will
happen to the liver for an individual but is of predictive value
when found in groups.
7. There is need for standardization, both of the methods used
to measure ALT activity and of what should be adopted as
the truly “normal” range.50,51
8. Suitable revisions in guidances, teaching approaches, and
clinical practice will need to reflect these points.