Invalidation by extreme assay artefacts of immunoassayable urinary insulin as a parameter of pancreatic allograft function

Abstract

There are several reasons to believe that declines in pancreatic beta cell function should presage clinical rejection of pancreatic allografts; furthermore, assessments of urinary insulin might be more informative than plasma determinations to assess pancrcaticocystostomy function. However, the previous literature on urinary insulin is inconsistent. We examined endogenous insulin values, and recovery of exogenously‐added insulin in urine from control subjects and patients, before and after ultrafiltration to remove proteolytic enzymes derived from the pancreas. In patients with a pancreatic graft, major artefacts in the radioimmunoassay were found (attributable to degradation of tracer and/or antibody, presumably by proteases) which could not be obviated by ultrafiltration or other approaches. These included spuriously high endogenous “insulin” levels, and virtually total inability to recover authentic insulin added to PxTx (pancreas transplant) urine. These findings raise significant doubts about the validity of most previous studies on urinary insulin (or other markers for pancreatic rejection) based on the use of immunoassay measurements.