Our laboratory is occasionally asked to evaluate isolated enzyme elevations for the presence of macroenzymes. Our evaluation is essentially a first-pass assessment for immune complexes, based on the susceptibility of macroenzymes to polyethylene glycol (PEG) precipitation (1). A recent case was found to be consistent with macro-AST. The physician asked whether, despite this finding, one could still rule out a circumstance of overproduction of AST. To rephrase the question: is the patient’s abnormal AST normal for macro-AST?

There is a reasonable question. Our approach to answering this question -- albeit an indefinite approach -- was described in a 2015 AACC abstract (2). The accepted explanation of elevation of a macroenzyme is that it is due solely to the extended lifetime of enzyme in circulation, due to its association with an immunoglobulin, rather than to any abnormality in its rate of appearance in circulation (its plasma production rate). A simple mass balance can thus relate macroenzyme (E’) concentration to normal enzyme (E) concentration: E’ = (k/k’) E, where k (enzyme) and k’ (macroenzyme)) are first-order elimination rate constants consistent with the half-life (t1/2) of each entity. The overall premise is simply that there is a projected “reference interval” for E’, if k and k’ are known.

For our clinician’s patient, the back-projection calculation (E = (k’/k) E’) gave a value for E at the 14th percentile of the reference interval for AST( 7-42 U/L), assuming t1/2 for E (normal AST) of 0.7 days, and assuming t1/2 of E’ to be equal to that of IgG (20 days). This calculation was a reasonable indication that AST production by itself was not elevated in this patient.

It was of interest, then, to examine this premise with respect to values for macro-AST reported in the literature. We found a bimodal distribution for reported concentrations of macro-AST that, in short, was very consistent with (was bounded by) E’ distributions projected from the reference interval for normal AST for an overlapping mixture of IgA complexes (approximately 70% of reported values for macro-AST, with assumed t1/2 of 6 days) and IgG complexes (approximately 40% of reported values, with assumed t1/2 of 20 days). In fact, the bimodal distribution was even better characterized when t1/2 for E’ was reduced by 35% from the textbook values of t1/2 for both IgA and IgG – and it is reasonable to suppose that t1/2 for immune complexes might indeed be less than t1/2 for uncomplexed immunoglobulins.

In short, literature data for AST (as an example) support a means of calculation of a “back-projection” for E from E’, or of the “forward projection” of a “reference interval” for E’ from E. This information may be helpful (and reassuring) in discussion with clinicians regarding the interpretation an isolated enzyme elevation shown to be consistent with a macroenzyme. It is likely that, in most cases, the macroenzyme concentration will be found to be a “patently normal” abnormal value.


  1. Davidson DF, Watson DJ. Macroenzyme detection by polyethylene glycol precipitation. Ann Clin Biochem 2003;40:514-20.
  2. Rubin AS, McCloskey LJ, Goldsmith BM, Stickle DF. Serum concentrations of macro-AST relative to normal AST: limits consistent with IgG or IgA complexes and with normal rates of serum AST production. AACC, Atlanta, July 2015, #A-129