Evaluation of the Interference of HbE on HbA1c Results on the Detection System Produced by Shenzhen Lifotronic


 Objective: To analyze the interference of HbE on the HbA1c results on detection systems manufactured by Shenzhen Lifotronic. Methods: Volunteers were recruited from Zhongshan city, Guangdong province during June 2020 to November 2020. A total of 55 specimens were obtained, including 40 cases without HbE(23 cases in normal group and 17 cases in diabetes group) and 15 cases with HbE variant. HbA1c of these two groups (normal and HbE group) were analyzed by Shenzhen Lifotronic H8, H9 and GH-900Plus, Bio-Rad Variant II Turbo 2.0 and D-10 from USA, Arkray HA-8180 from Japan, statistical analysis was conducted on the difference between detection results of normal group and HbE variant group. Bio-Rad Variant II Turbo 2.0 was used as the comparison system to calculate the bias and to evaluate the interference of HbE on each system. Results: In normal group, the results of each detection system were comparable to those of Variant II Turbo 2.0, with a bias of -6%~6%. In HbE variant group,system bias of H8, H9, GH-900Plus and D-10 was within -6% and 6%, and the results showed no significant difference (P>0.05), while the negative bias of HA-8180 showed significant difference (P<0.05). Conclusion Compared with Bio-Rad Variant II Turbo 2.0, HbA1c can be correctly detected by H8, H9 and GH-900Plus system without the interference of HbE.


Introduction
Hemoglobin A1c (Glycated hemoglobin, HbA1c) is widely used in clinical practice as an important indicator to assess the average level of blood glucose control. China has paid increasingly attention to the standardization of HbA1c test results, and the research and development of HbA1c test system has been constantly innovating. It is particularly important to develop a detection system with good precision and strong anti-interference ability to accurately meet the needs of clinical detection 1,2 .HbE is one of the most common abnormal hemoglobin proteins in the world, with high incidence rate in China, India, Southeast Asia and North America 3 . Hemoglobin variation can interfere with HbA1c test results to varying degrees 4 , the purpose of this study is to discuss the HbA1c testing results in the presence of HbE variant by Shenzhen Lifotronic's H8, H9 and GH-900Plus detection systems. The comparison between Shenzhen Lifotronic's three system and Bio-Rad Variant II Turbo 2.0 with regard to the interference degree of HbE was evaluated. The results are reported as follows.

Materials And Methods
Source of specimens Volunteers from Zhongshan Hospital of Sun Yat-Sen University were recruited from June 2020 to November 2020. A total of 55 EDTA-K 2 anticoagulant venous blood samples were obtained in this study, including 40 cases in normal group,which were used as normal specimen for comparison experiment without hemoglobin variant, hyperlipidemia, uremia and other interfering factors. In addition, 15 HbE variant samples were preliminarily determined HbE presence by capillary electrophoresis, and then genotype was detected and con rmed by gene sequencing. All the volunteers signed informed consent forms, and the study was reviewed by the Medical Ethics Committee of Zhongshan Hospital.

Specimen collection
Fasting blood of 2mL EDTA-K 2 anticoagulant venous blood was collected. These samples were used for capillary electrophoresis, gene sequencing and HbA1c test. All collected specimens were pipetted and stored at -70℃.

Detection system and method
HbA1c was detected by Ion Exchange High Performance Liquid Chromatography (IE-HPLC) using Shenzhen Lifotronic's automatic H8, H9 and GH-900Plus HbA1c analyzers and reagents, calibrators, control material and consumables.
HbA1c was detected by Ion Exchange High Performance Liquid Chromatography using Variant II Turbo 2.0, D-10 automatic HbA1c analyzers and reagents, calibrators, control material and consumables (Bio-Rad Company).
HbA1c was detected by Ion Exchange High Performance Liquid Chromatography using Arkray HA-8180 automatic HbA1c analyzer and reagents, calibrators, control material and consumable.
The NGSP quality evaluation material was retested at Variant II Turbo 2.0 before experimental operation to ensure that the system met the requirements of NGSP laboratory certi cation, so as to ensure the traceability of the comparison system. The relative bias between detection results of different systems and Variant II Turbo 2.0 was calculated to be within -6% ~ +6%. Meeting this condition means that detection results of other systems are comparable to those of Variant II Turbo 2.0.

Statistical analysis
HbA1c tests were respectively performed on the experimental systems and the comparison system for each group. Statistical analysis was carried out for the bias of detection results. All the data were statistically analyzed by Excel and SPSS 22.0 software. Kolmogorov-Smirnov test, a single sample, was used to test the normality of the measurement data, which were in line with the normal distribution, and paired sample t-test was used as well. P<0.05 was considered statistically signi cant.

Results
Retesting results of NGSP in Variant II Turbo 2.0 system All the NGSP quality assessment and target value bias detected on the same day were less than 6%, detailed results are shown in Table 1. Indicating that the accuracy of the analysis and testing system reaches acceptable standards to ensure that the system meets the requirements of NGSP laboratory certi cation, so as to ensure the traceability of the comparison system. Comparison and bias analysis of HbA1c test results of the ve detection systems The correlation coe cient r of regression analysis between the ve detection systems including H9, H8, GH-900Plus, D-10 and HA-8180 and the comparison system of Variant II Turbo 2.0 were all greater than 0.975, and the percentage bias of the experimental system and the comparison system were all less than ±6%, the speci c results were shown in Table 2. Chromatogram analysis of HbE variant in different HbA1c detection systems The elution time of H9 was 360 seconds, and the peak names were HbA1a, HbA1b, HbF, LA1c, HbA1c, P3, P4, HbA0 ,HbA2 and HbE/D/S/C (if any) respectively. The average retention time of HbE was 220 seconds, which is after HbA2. Moreover, it was identi ed and reported separately in the chromatogram, and the system could directly indicate suspected HbE variant.
The elution time of H8 and GH-900Plus was 130 seconds, and the peak names were HbA1a, HbA1b, HbF, LA1c, HbA1c, HbA0 and variant window respectively. The average retention time of HbE was 89 seconds which is after HbA0 in the chromatogram, and the system gave V-window alarm.
The elution time of Variant II Turbo 2.0 was 90 seconds, and the peak names were HbA1a, HbA1b, HbF, LA1c, HbA1c, P3, P4 and HbA0 respectively. The average retention time of the variant was 66 seconds, HbE's retention time is after HbA0, and it was next to HbA0, but the obvious presence of this variant could be identi ed from the chromatogram, and the system gave V-window alarm too.
The elution time of D-10 was 180 seconds, and the peak names were approximately HbA1a, HbA1b, HbF, LA1c, HbA1c, P3, and HbA0, respectively. The average retention time of HbE variant was 95 seconds, which is after HbA0, and it was next to HbA0, but the obvious presence of this variant could be identi ed from the chromatogram, and the system gave V-window alarm too.
The elution time of HA-8180 was 48 seconds, and the peak names were approximately HbF, LA1c, HbA1c, and HbA0, respectively. The retention time of HbE was between HbA1c and HbA0, and it was after HbA1c in the chromatogram. The system showed no variant alarm or there was no peak name or retention time.
Statistical analysis of HbA1c test results differences between normal group and HbE variant group.
There was no signi cant difference between HbA1c and HbE group detected by H8, H9, GH-900plus, Turbo 2.0 and D-10 (P>0.05), while there was signi cant difference between HbA1c and HbE group detected by HA-8180 (P<0.05), as shown in Table 3.  Table 4 and gure 2 for details. However, its elution time is longer, and the number of tests can be supported by chromatographic column and reagent is less, so D-10 is more commonly used in small laboratories.
The Arkray HA-8180 from Japan has the feature of fast speed.. Its elution time is 48 seconds, which is the fastest in the detection systems of this study. Its fast mode is often used for mass screening.
However, the elution time is too short to separate the HbE peak which leads to relatively large negative bias in the detection results. Therefore, HA-8180 was not suitable for the detection of HbA1c results for population with HbE variants.
In conclusion, H8, H9 and GH-900Plus HbA1c detection systems and Bio-Rad Variant II Turbo 2.0 have comparable results when testing HbA1c in population with HbE variants. And the anti-interference ability of these three detection systems from Shenzhen Lifotronic was better than that of Arkray's HA-8180. With continuous improvement of reagent formula and chromatographic column quality, HbA1c detection systems from Shenzhen Lifotronic have strengthened their ability to separate and identify the variants and have a good performance on anti-interference of variants. In regions where hemoglobin variants appear frequently, it is necessary to fully understand the anti-interference ability of each detection system against the variants, so as to avoid results bias of HbA1c, which would lead to misjudgment of average blood glucose when monitoring level of patients by clinicians. Figure 1 Chromatograms of normal and HbE group in 6 detection systems ← -Represents the HbE position in Chromatogram