Performance comparison of EasyFix G26 and HYDRASYS 2 SCAN for the detection of serum monoclonal proteins

Abstract Background Serum protein electrophoresis (SPE) is a widely used laboratory technique to diagnose patients with multiple myeloma (MM) and other disorders related to serum protein. In patients with MM, abnormal monoclonal protein can be detected by SPE and further characterized using immunofixation electrophoresis (IFE). There are several semi‐automated agarose gel‐based systems available commercially for SPE and IFE. In this study, we sought to evaluate the analytical performance of fully automated EasyFix G26 (EFG26) and semi‐automated HYDRASYS 2 SCAN (H2SCAN) for both SPE and IFE. Methods Both instruments were operated according to manufacturer's instructions. Samples used include a commercially available normal control serum (NCS) and patients' specimens. The following were evaluated: precision and comparison studies for SPE, and reproducibility and comparison studies for IFE. Statistical analyses were performed using Microsoft Excel. Results For SPE repeatability study, our results showed that EFG26 has higher coefficient of variation (%CV) compared with H2SCAN for both samples except for monoclonal component with %CV of 0.97% and 1.18%, respectively. Similar results were obtained for SPE reproducibility study except for alpha‐1 (4.16%) and beta (3.13%) fractions for NCS, and beta fractions (5.36%) for monoclonal sample. Subsequently, reproducibility for IFE was 100% for both instruments. Values for correlation coefficients between both instruments ranged from 0.91 to 0.98 for the five classic bands. Conclusion Both instruments demonstrated good analytical performance characterized by high precision, reproducibility and correlation.


| INTRODUC TI ON
Serum protein electrophoresis (SPE) is an established technique used for screening abnormal protein known as monoclonal proteins (M proteins, paraproteins, monoclonal immunoglobulins) in the serum. 1 SPE aids in the diagnosis of patients with multiple myeloma (MM) and other serum protein disorders. In more than 80% of patients, abnormal proteins can be detected by using SPE technique. 2 Based on the principle of zone electrophoresis, SPE separates the serum protein into five major fractions (albumin, alpha-1 globulins, alpha-2 globulins, beta globulins, and gamma globulins) which primarily according to their charge at a given pH.
As a result, these fractions defined the pattern of SPE. In comparison with normal serum pattern, patient with MM usually produced an abnormal serum pattern which can be evaluated visually on SPE gel and also by electropherograms that were generated by the densitometer.
The confirmation of the presence and identification of monoclonal protein type can be accomplished by immunofixation electrophoresis (IFE); a "gold standard" method which consists of an electrophoresis phase and a fixation phase. Fixation phase allows a protein to be anchored after electrophoresis phase by forming an insoluble complex between antiserum containing anti-IgG, anti-IgA, anti-IgM, anti-light chain kappa, and anti-light chain lambda with specific antibody in the serum therefore allowing the protein to be characterized. [3][4][5] Since year 2008 our laboratory has been using a semi-automated HYDRASYS 2 SCAN (H2SCAN) to perform SPE and IFE test for diagnostic purposes. The availability of a fully automated instrument raised an interest for us to compare the performance between both systems. There was a study between both instruments published by Napodano et al in 2017, but only on the comparison of protein immunofixation electrophoresis. 6 To our knowledge, this is the first performance comparison study that has been performed by using EasyFix G26 (EFG26) which claimed to be the first instrument that could perform IFE fully automatically. Therefore, the aim of this study was to evaluate the analytical performance of a fully automated system, EFG26 with H2SCAN.

| Gel electrophoresis using easy fix G26
Fully automated agarose gel SPE and IFE were performed according to the manufacturer's instructions using the EFG26 (Interlab) with SRE636K and SRE628K kits (Interlab), respectively. EFG26 automatically performed all the procedures for SPE which include the application of samples on agarose gel plate; electrophoretic migration (pH 8.9 at 220 V and 20°C); gel denaturation; gel staining (with Acid Blue); gel destaining; gel drying; and densitometric reading using Elfolab™ densitometer (Interlab).
Procedures for IFE were also fully automated beginning from dispensation/dilution of samples, followed by application of samples on agarose gel plate; electrophoretic migration (380 V at 20°C); application of the fixative solution and antisera (IgG, IgA, IgM, kappa:κ, or lambda:λ); gel incubation; gel blotting; gel denaturation; gel washing; gel staining (with acid violet); gel destaining; and final gel drying (80°C). 7

| Gel electrophoresis using HYDRASYS 2 SCAN
Semi-automated SPE and IFE were performed using the H2SCAN For the detection of monoclonal bands, visual inspection of stained gels was also performed for both instruments apart from densitometric reading.

| Analysis and evaluation
Q uantitative results of SPE were evaluated through precision studies which involve repeatability (within-run precision) and reproducibility (day-to-day imprecision). Repeatability for SPE (n = 30) was analyzed based on CLSI guideline EP15-A2 and Ministry of Health Malaysia guideline using a commercially available normal control serum (NCS; manufactured by Sebia) and patient's monoclonal sample. 9,10 Reproducibility for SPE (n = 15) was also evaluated using NCS and patient's monoclonal sample for five different days based on CLSI guideline EP10-A2 and Ministry of Health Malaysia guideline. Linear regression analysis for SPE was performed between the two instruments through comparison study using patients' samples (n = 50) submitted for routine SPE test and the statistical significance was set at P < .05. The agreement between both instruments was later assessed by Bland-Altman analysis.
Meanwhile for IFE, reproducibility study was done over a period of 3 days (n = 4) and the agreement between the two instruments was also observed (n = 4). Statistical analyses were performed using Microsoft Excel.

| Precision studies for serum protein electrophoresis
In the repeatability study, 30 aliquots of NCS and monoclonal sample were assayed by both EFG26 and H2SCAN. As revealed by NCS, our findings for the five classic bands expressed as coefficients of variation (%CV) ranged from 3.35% to 11.86% for EFG26 compared with 0.82%-5.37% for H2SCAN (   Table S1). The central 0.95 interval (mean difference ± 2 SD) indicated a good agreement

| Comparison study for immunofixation electrophoresis
Four serum samples were analyzed on both systems to measure the agreement between EFG26 and H2SCAN. A 100% agreement was observed between the two instruments which can be visualized in   cepted by the clinicians in reviewing the sample. 11 In this study, both EFG26 and H2SCAN electropherograms produced five major serum protein fractions from a healthy serum sample and from a patient with monoclonal protein ( Figure 5).

| D ISCUSS I ON
Assessment of precision is part of a process of verifying or validating a method which confirms the suitability for use. 12 Overall, EFG26 produced a higher %CV values compared with H2SCAN in both repeatability and reproducibility studies for the classic bands.
As for monoclonal component, EFG26 produced a lower %CV value (0.97%) in the repeatability study and a higher %CV value in the reproducibility study (0.67%) compared with H2SCAN which were 1.18% and 0.45%, respectively.
The %CV values obtained by both instruments were comparable to the %CV values reported by Kahn and Strony who performed their study using cellulose acetate electrophoresis. The study showed that the precision of densitometric quantification fraction %CV ranged from 1.3% to 5.1% for albumin, 4.4%-20.2% for alpha-1 globulins, 1.8%-9.2% for alpha-2 globulins, 1.6%-14.0% for beta globulins, and 2.5%-27.2% for gamma globulins. 13 In addition, a review by Bossuyt on protein fractionation using capillary method by Paragon 2000 ® and Capillary ® also showed similarity with both %CV values obtained for the five classic protein fractions were consistently <8% for intra-run (repeatability) and <11% for between-run precision (reproducibility) studies.
Their overall precision was 1.2% for albumin, 6.1% for alpha-1 globulin, 3.2% for alpha-2 globulin, 3.2% for beta globulin, and 4.3% for gamma globulin fractions. 14 Furthermore, EFG26%CV values for NCS also have a comparable performance. Therefore, the values obtained in our study were in accordance with those of the literature with both systems provided highly repeatable and reproducible results.
As with our comparison study, the results showed high correlation and good agreement between the two instruments. The highest correlation coefficient (r) was for the gamma globulins

ACK N OWLED G M ENTS
We thank the Director General of Health Malaysia for his approval to publish this article, the Director of Institute for Medical Research for her support, and staffs of Unit for Molecular Diagnostics & Protein for their kind assistance throughout the course of this research.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

AUTH O R CO NTR I B UTI O N S
FDAN designed the research study, carried out the statistical analysis, and drafted and wrote the manuscript. MKNMK co-designed the