The occasional finding of elevated serum aminotransferase values in individuals without symptoms is a more and more broadening phenomenon, particularly in areas where preventive medicine results in frequent health screenings that include liver enzyme measurement. In most cases, hypertransaminasaemia is related to liver cell injury, with alanine aminotransferase (ALT) values generally more increased than those of aspartate aminotransferase (AST).1 In both adults and children, the AST to ALT ratio can be sporadically elevated in non-alcoholic steatohepatitis, liver fibrosis/cirrhosis and Wilson disease, but their ratio typically does not exceed two. In the adult population, an AST to ALT ratio ≥2:1 may indicate alcohol abuse and alcoholic liver disease especially when it is associated with an elevated gamma-glutamyl transferase (GGT) activity. When AST values are higher than ALT, sources other than liver should however also be considered, including cardiac, muscular, haematological, renal, cerebral and endocrine cells/organs. In these cases, serum AST levels increase jointly with some other associated enzymes depending on the pathology, for example, lactate dehydrogenase in heart disease or haemolysis, creatine phosphokinase and/or aldolase in muscular diseases and/or intense exercise. AST released from cellular lysis comprises the cytosolic (cAST) and mitochondrial (mAST) isoenzymes, encoded by GOT1 (10q24) and GOT2 (16q12) genes respectively. The largest part of serum AST activity is cytosolic, while the increase of serum mAST indicates the extent of ischaemic necrosis better than cAST.2

When AST values are increased separately without an evident corresponding organ damage, finding the origin of this hyperAST may be extremely difficult to understand or solve. In such a puzzling scenario in which studies on large paediatric series are scarce, the article by Magen-Rimon and colleagues appearing in this issue of Acta Paediatrica3 is particularly welcome. In fact, these authors retrospectively reviewed the medical charts of 32 children diagnosed with prolonged isolated AST elevation (ranged 1.2–12 folds upper limit of normal). Similar to other literature studies and case series/reports, their patients needed to go an extensive traditional stepwise workup1 which included abdominal ultrasonographic studies, laboratory screening for major and minor hepatotropic viruses, autoimmunity, celiac disease, a quite large number of inborn errors of metabolism, and renal, muscular and haemolytic diseases as well. Using both polyethylene glycol (PEG) precipitation and ultrafiltration (UF) testing, at last hyperAST resulted most likely due to the presence of macroAST in approximately one of 4.5 patients (22%). The authors did not find differences in AST levels between patients with/without macro-AST.3

Although most of this information is not entirely new, the study contributes to shed further light on several aspects of macro-AST, a condition which is still frequently neglected and as such responsible for unnecessary costly, iterative and invasive evaluation, for example, the liver biopsy considered essential in one child of their series.4, 5 As shown in Table 1, macro-AST belongs to the group of the so-called macroenzymes which also include a number of other hepatic and non-hepatic enzymes as well. Macroenzymes are normal enzymes forming high molecular-mass complexes unable to undergo renal clearance because of their size, and, therefore, they persist in the serum causing interference with routine measurement and diagnostic confusion.6 As other macroenzymes, also macro-AST may be categorised in two types. In Type I, the pathogenesis appears to be likely autoimmunity related, with the Fab portion of immunoglobulins IgG, or IgM or Ig A targeting serum enzymes as antigens in the setting of a molecular mimicry phenomenon. In Type II, the macroenzymes are transiently formed when the enzyme either self-polymerises or associates with a foreign chemical, for example, a drug or the apoprotein moiety of lipoprotein carriers.5, 7 The prevalence of most macroenzymes is quite low, spanning 0.5%–2.5%; although they are generally described as an incidentaloma in healthy people, they have been associated also with various pathological conditions, mostly autoimmune or infectious in origin (Table 2). Macro-AST, specifically, has been reported to range between 13% and 60% of adults evaluated in tertiary centres for isolated elevation of AST levels.5 Except for rare reports (e.g., association with inflammatory bowel disease and Kawasaki disease), data suggest that in children and adolescents macro-AST is a benign condition and is not accompanying other disorders.

TABLE 1. Origins of macroenzymes Prevalent hepatic origin Non-exclusive hepatic origin Non-hepatic origin Aspartate aminotransferase (AST) Alanine aminotransferase (ALT) Gamma-glutamyltransferase (GGT) Lactate dehydrogenase (LDH) Alkaline phosphatase (ALP) Acid phosphatase (ACP) Prolactin (PRL) Alanine aminopeptidase (AAP) Amylase, lipase (AMS, LPS) Creatine kinase (CK) Glucose-6-phosphate dehydrogenase (G6PDH) TABLE 2. Conditions which have been reported in children with macro-AST Immune related disorders Liver disorders Rheumatoid arthritis Systemic lupus Immunotherapy Monoclonal gammopathy Inflammatory bowel disease/celiac disease Drugs Chronic liver diseases Acute and chronic hepatitis C Hepatic malignancies

The paper by Magen-Rimon and colleagues adds on to the few available data of paediatric macroAST series by reporting a macro-AST prevalence which is lower than, but quite comparable to, that of the large series of Caropreso et al. (i.e., 1 of 3), and may occur early in life.8 As in adults, macro-AST frequency should be considered cautiously because of a possible referral bias and of the numerous methods of macroenzymes testing4, 9-12 (Table 3). Magen-Rimon et al. studied their patients with PEG precipitation together with UF, and considered positive or negative only those sera with concordant results. Similar studies using the same cut-off of their PEG test showed that it was a valuable tool with good sensitivity and specificity having electrophoresis as the gold standard.8 The 5-year long-term follow-up of Magen-Rimon et al. patients, which in some way remains comparable to that of Caropreso et al.,8 is important because it allows to confirm a general benign course of this condition over the years. Even though macro-AST levels in some series show a natural fluctuation or a possible normalisation over time, in Magen-Rimon's study the macro-AST remained stably elevated. If this variance depends on the different type of macroenzyme (type 1 or 2) is still unclear because this issue has not been studied in most published series.

TABLE 3. Most common methods used to detect macroenzymes/macro-AST Test Meaning / Pros–Cons Enzyme testing before/after precipitation with PEG 6000 Cut-off values proposed for defining presence/absence of macroenzyme still uncertain. Enzyme testing before/after ultrafiltration Centrifugation using a 100 kDa pore size filter which allows for removal of large molecular mass components Precipitation with human antisera against IgG, IgA and IgM. Sera with macro-AST may be precipitated with human antisera against IgG, IgA and IgM. The test is specific to verify the binding of serum AST to an immunoglobulin Sequential lipoprotein fraction precipitation This test examines the possible aggregation of serum AST with lipids. Chromatography This test relies on size-exclusion, HPLC and Protein G-agarose beads to precipitate immune complexes of AST/immunoglobulin followed by centrifugation and AST activity measurements in the supernatant. Serum protein electrophoresis on cellulose acetate Macro-AST has a migration intermediate between cAST and mAST. Cumbersome. Measurements of heat stability

Sera storage at 4°C for 1 week shows a decrease in AST levels >90% in case of macro-AST.

This screening method is potentially useful to laboratories lacking the above more complex methodologies.

Abbreviations: AST, Aspartate aminotransferase; cAST, cytosolic AST; HPLC, high-performance liquid chromatography; Ig, Immunoglobulin; mAST, mitochondrial AST; PEG, polyethylene glycol.

Despite the description of familiar forms of macro-AST, its formation has not been confirmed to have a clearly defined inherited genetic basis. Recently, Kuleka et al. by screening patients with suspected familial macro-AST found a GOT1 missense variant (p. Gln208Glu and rs374966349) as a putative cause in more than half probands, while its prevalence in healthy controls was only 0.18%. Computational strategies showed that the amino acid at this position is a negatively charged glutamate which could plausibly be responsible for a strong anchorage of serum immunoglobulins on the GOT1 surface.2 Due to the positive results observed only in part of the families described by Kuleka, and others’ findings who were unable to find the presence of this mutation in a family with macro-AST,13 it is possible that familial macro-AST may have polygenic inheritance requiring further genetic research to solve this still unsolved puzzle.2

LIMITATIONS AND PERSPECTIVES

Magen Rimon and colleagues by studying a quite large sample of patients add to what we know about paediatric macro-AST.3 However, like other investigations on this topic, their well-designed study is not exempt from inevitable and evitable limitations, as the authors themselves acknowledge. The major inevitable limitations are the retrospective nature of the study and the poor generalisability (only Israelian participants). Evitable limitations are the lack of more detailed information about the possible aetiology of the remaining patients with macro-AST negative hyperAST. One could speculate that the variability of the upper reference limits of AST level during different periods of childhood might represent the extremes of distinct Gaussian distributions in the normal paediatric population at different ages.5 Precise data on the AST and ALT normal values for age and folds of increase of AST over ALT values in their series might have added weight to the authors’ conclusions and made possible more comparisons with other studies. Despite this, the article by Magen-Rimon et al. is of interest for at least two major indisputable reasons: (a) it quite clearly confirms the benign course of a prolonged elevation of paediatric isolated AST, in general, and of paediatric macro-AST, in particular, (b) it further confirms that isolated AST levels cannot be used to discriminate between negative, borderline or positive macro- AST, as even AST values slightly exceeding the upper reference range may be due to the macroenzyme presence.8

Based on all the above findings, the recommendation of avoiding extensive/invasive laboratory and imaging investigations in healthy children with prolonged isolated AST elevation appears reasonable also when macro-AST testing is not easily available.

CONFLICT OF INTEREST

There is no conflict of interest regarding this manuscript.