Protein Interference with Common Laboratory Tests (download)
Objective: Protein as an interferent in underappreciated. Therefore, we evaluated interference by proteins on clinical laboratory tests. Methods: Base Pool was prepared using a serum pool. A High Pool was prepared by spiking proteins into the Base Pool (Concentrated human albumin and gamma-globulins 1:1; ASSURANCE™ Interference Test Kit, Sun Diagnostics, New Gloucester, ME). The Base Pool and High Pool were intermixed to create five levels of total protein (3.5, 6.9, 9.4, 12.4, and 15.3 g/dL). Multiple analytes were measured on the Beckman AU5800. Results: Minimal effects were seen with alkaline phosphatase (ALP), amylase (AMY), direct bilirubin (DBIL), total bilirubin (TBIL), creatine kinase (CK), gamma-glutamyltransferase (GGT), glucose (GLU), lactate dehydrogenase (LDH), lipase (LIP), phosphorus (PHOS), urea nitrogen (BUN), uric acid (UA), lipoprotein(a) (LPA), small dense low-density lipoprotein cholesterol (sdLDL-C), high sensitivity C-reactive protein (hsCRP), ferritin (FER), low-density lipoprotein cholesterol (LDL-C), high density lipoprotein cholesterol (HDL-C), potassium (K), chloride (CL), triglycerides (TRIG), and total cholesterol (CHOL). Positive bias with increasing protein concentration was seen with alanine aminotransferase (ALT), aspartate aminotransferase (AST), calcium (CA), creatinine (CRE), iron (FE), magnesium (MG), homocysteine (HCY), and non-esterified fatty acids (NEFA). Because proteins bind analytes such as calcium, magnesium, iron, HCY, and NEFA, and the albumin interferent is minimally processed human albumin, we believe these increases were artifactual. Negative bias was seen with carbon dioxide (CO2), apolipoprotein AI (apo AI), and apolipoprotein B (apo B). Conclusion: Human proteins (albumin and gamma-globulin) interfere with the measurement of many routine clinical chemistry assays, and manufacturers and laboratorians need to pay more attention to proteins as a potential interferents. The mechanism may be photometric or a volume depletion effect, whereby a very high protein concentration will reduce the available water so that the analyte concentration is artifactually low.
Proteins as potential interferents are overlooked by laboratorians and IVD manufacturers. The assumption is that protein interference is caused by paraproteins (monoclonal immunoglobulins) [1,2], but this is an oversimplification. Elevated proteins, including albumin and gamma-globulins may affect many assays, including bilirubin, phosphate, HDL cholesterol, GGT, CRP, and glucose . The manufacturer may note on the package insert that monoclonal gammopathies may result in errors, but rarely do they perform a comprehensive study of protein interference. The Clinical Laboratory Standards Institute (CLSI) recommends testing with both albumin and gamma-globulins at a total protein concentration of 12 g/dL . The interference may be due to physical or chemical alteration in the signal (e.g. increased light scattering) or it may be due to a volume displacement effect . The practical effect of an increased concentration of proteins is a decrease in serum water or volume depletion: the aspirated sample is diluted with solid (protein and/or lipid) so that the analyte concentration is artifactually low. Viscosity is also a common problem with specimens containing monoclonal immunoglobulins, especially IgM, and this may affect the accuracy of sample delivery.
A Base Pool was prepared using leftover, de-identified patient sera. A Test Pool was prepared by adding 1 part concentrated human proteins (ASSURANCE™ Test Kit, Sun Diagnostics, LLC, New Gloucester, ME) to 1-part Base Pool. The concentrated proteins are ~50% human albumin and 50% human gamma-globulins. A Control Pool was prepared by mixing 1-part Base Pool with 1-part saline to maintain equal analyte concentrations. The Test Pool and Control Pool were intermixed to create five levels of protein interferents. Multiple analytes in each of the five levels were measured in duplicate on the Beckman AU5800 at HDL, Inc. (Richmond, VA). Bias was calculated for levels 2 to 5 versus level 1, or level 4 versus level 1. Bias < ½ the total allowable error was considered acceptable.
- Minimal interference with total protein concentrations up to 15.5 g/dL were seen for ALP, AMY, DBIL, TBIL, CL, CHOL, CK, FER, GGT, GLU, HDL-C, HDL3-C, hsCRP, LDH, LIP, LPA, LDL-C, PHOS, K, sdLDL-C, TRIG, BUN, or UA.
- Positive bias was seen with increasing protein concentrations for ALT, AST, CA, CRE, HCY, FE, NEFA, and MG.
- Negative bias was seen with CO2, APOAI, and APOB.
- Because human proteins bind calcium, homocysteine, iron, nonesterified fatty acids, and magnesium we measured these analytes in the concentrated protein interference material and subtracted this from the measured concentrations. Positive bias remained for all analytes except NEFA, which converted to a negative bias.
Table 1: Measured Protein Concentrations of Five Levels of Test Pools, g/dL
Table 2: List of Analytes, Test Concentration, Allowable Error, and Bias Due to Protein
(Significant Bias due to Protein Interferent is highlighted in RED)
Bias Plots: Interference Due to Protein
- Human proteins (albumin and gamma-globulins) interfere with a number of common laboratory tests.
- Protein interference is not described by most manufacturers, except to note possible interference by monoclonal paraproteins.
- Laboratorians and manufacturers need to pay greater attention to proteins as a source of interference with chemistry assays.
- Dimeski G. Interference testing. Clin Biochem Rev 2008; 29 Suppl (i): S43-S48.
- Kroll MH, Elin RJ. Interference with clinical laboratory analyses. Clin Chem 1994; 40:1996-2005.
- Kazmierczak SC, Catrou PG. Analytical interference: more than just a laboratory problem. Am J Clin Pathol 2000; 113:9-11.
- Artiss JD, Zak B. Problems with measurements caused by high concentrations of serum solids. Crit Rev Clin Lab Sci 1987; 25: 19-41.
Adapted from: John H. Contois, Jason P. Nagy, and Rae-Anne Nguyen, Sun Diagnostics, New Gloucester, ME. Protein interference with common laboratory tests. Presented at the Annual Meeting of the AACC, Atlanta, GA; July 2015.