Validation of limulus tests for endotoxin evaluation in dialysate

Summary: To measure endotoxin (ET) levels in dialysate accurately, three commercial limulus tests (Endospecy, Seikagaku, Tokyo, Japan; Kinetic QCL, BioWhittacker, MD, USA; ET single test-ES, Wako, Osaka, Japan) were validated in order to determine whether or not they are enhanced or inhibited by bicarbonate dialysate. Two reference standard ET, one from US Pharmacopoeia (Escherichia coli reference standard endotoxin; RSE) and the other from European Pharmacopoeia (Salmonella RSE), were adjusted with dialysate at concentrations of 0.05, 0.10 and 0.15 EU/mL for measurement by each kit. A simple linear regression line forced to the origin was generated for each kit. the validation criteria, based upon accuracy and precision analysis, includes the fact that the range of slope of the regression line of 95% confidence must fall between 0.75 and 1.25, and the correlation coefficient must be no less than 0.99 with significance ( P < 0.05). No enhancement/inhibition was observed with Endospecy. Kinetic QCL showed enhancement with Salmonella RSE and no enhancement/inhibition with E. coli RSE. the ET single test-ES kit was inhibited with both RSE. Clinical dialysate samples were also measured for ET activity by each limulus kit which revealed that Endospecy and Kinetic QCL gave almost identical values; however, the ET single test-ES kit gave significantly lower values (41%) than the others ( P < 0.05). These results suggest that either the validation test using RSE or comparative analysis with an already validated kit (Endospecy) is necessary for measuring ET levels in dialysate.