Comparison of four matrixes for diluting insulin in routine clinical measurements

Abstract Objective In our laboratory, 2.36% (6626/280765) samples obtained for insulin evaluation have serum insulin concentrations higher than 300 mU/L, resulting in curves outside the linear range in the insulin release test (IRT). Accordingly, using appropriate dilution protocols to determine insulin concentration accurately is important. Here, we compared the effectiveness and economy of four different solutions for diluting high‐insulin serum in routine clinical measurements. Method Residual serum samples with high‐insulin concentrations ranging from 200 to 300 mU/L were collected in Peking Union Medical College Hospital from August to November 2017. Four different matrixes including a Siemens original diluent, pure water, 0.9% NaCl, and low‐insulin serum (labeled as A to D, respectively) were used to dilute the serum in the ratios of 1:2, 1:5, and 1:10. Results We found that the linear correlation coefficients of A to D were higher than 0.9. The recovery rates of A to D were 86.4%–104.0%, 73.2%–99.3%, 76.4%–101.3%, and 84.2%–99.7%, respectively. We conclude that the use of 0.9% NaCl, pure water, or low‐insulin serum to dilute high‐serum insulin (>300 mU/L) is feasible and cost‐effective. Conclusion We recommend a dilution factor of 1:5 on a Siemens ADVIA Centaur XP® instrument. The clinically reported range was 0.5‐1500 mU/L. For specific samples (>1500 mU/L), we recommended using low‐insulin serum samples for dilution.


| INTRODUC TI ON
Insulin is a peptide hormone synthesized and secreted by islet β cells. The main function of insulin is to regulate the concentration of glucose in the blood, and secondly, it plays an important role in the metabolism of lipids and proteins. From a clinical perspective, the measurement of serum insulin concentration provides useful information for the diagnosis of insulin deficiency and insulin resistance, 1,2 particularly in diabetes, neonatal hyperinsulinemia hypoglycemia, insulinoma, and polycystic ovary syndrome. [3][4][5] Moreover, in addition to its therapeutic application in diabetes, insulin therapy can improve lipid metabolism and decrease mortality for myocardial infarction patients. [7][8][9] Physiological insulin therapy with insulin analogs is now relatively simple to use and is associated with fewer episodes of hypoglycemia in diabetics. 6 To optimize its use, it is important to predict the degree of postprandial hyperglycemia and the likely response to prandial insulin. 2,7,9 Accurate measurement of insulin is also helpful for evaluating insulin therapy compliance and suspected overdose, 10 which is particularly important in diabetic mothers. 11 A useful test of endogenous insulin function is the IRT or oral glucose tolerance test (OGTT). It involves the administration of oral glucose to a fasting patient to increase blood glucose and stimulate the β cells to release insulin. Serum insulin concentrations are measured at fasting, 0.5, 1, 2, and 3 hours after taking sugar. Normal human insulin secretion often peaks at 60 minutes after taking sugar, and then returns to normal concentration within 2 hours. 6 However, when the patient has severe insulin resistance due to polycystic ovary syndrome, obesity, type 2 diabetes, or other illness, the serum insulin concentration at each of the above points may exceed the upper limit of the detection system. If a specific value cannot be detected at this time, the patient's insulin peak time cannot be determined, and multiple peaks are compared with the increase in fasting insulin levels. Otherwise, we also obtained the clinical laboratory real data to manifest the importance of an insulin dilution study. To obtain accurate serum insulin concentration at all time points of the OGTT, and thereby extend the applicability of this test to more patients, it is necessary to issue a dilution measurement report on recalcitrant specimens. The purpose of this study was to investigate the feasibility of using other diluents to replace the original diluent provided by the manufacturer to dilute samples with insulin concentrations above the upper limit of linearity.
Considering the impact of health economics, we aimed to find a low-cost and effective diluent for routine clinical working. Thus, we evaluated the effects of four different diluents including the original Siemens diluent, pure water, 0.9% NaCl, and low-insulin serum to dilute high-insulin samples. This study aimed to provide an enhanced protocol for routine clinical work to ensure accurate results with an efficient test.

| Precision
To evaluate precision of insulin measurements, we used quality

| Diluent preparation
In this study, four dilutions were used, including original Siemens diluent (main components: potassium thiocyanate buffer and sodium sulfide), pure water (Millipore), 0.9% NaCl, and low-insulin serum pools (serum samples without bilirubin, hemolysis, and lipemia) (labeled as A to D, respectively). The concentration of low-insulin serum samples ranging from 0.5 to 2 mU/L was compared.
Low-insulin serum samples (n = 16) were obtained from clinical residual serum between August 22 and November 29, 2017 and stored at −80°C until use. Before analysis, all of the low-insulin serum samples were brought up to room temperature and mixed together. The total amount of low-insulin serum pool obtained was 8 mL. The average concentration of the low-insulin serum pool was 1.71 mU/L.

| Sample collection
A total of 19 residual serum samples with insulin concentrations between 200 and 300 mU/L, without bilirubin, hemolysis, and lipemia were collected from Peking Union Medical College Hospital for evaluating the effect of the different matrixes.

| Laboratory measurements
Serum insulin concentration was detected using a Siemens ADVIA Centaur XP ® automatic chemiluminescence immunoassay analyzer, with its corresponding reagents and calibrators provided by the manufacturer. Calibration was performed according to the manufacturer's instructions. The analytical sensitivity of this assay was 0.5 mU/L. Measurements were performed according to the standard operating procedure (SOP). 12,13 The instrument was calibrated and prophylactically maintained every year. Our laboratory also participated in external quality assessments by the National Center for Clinical Laboratories and the College of American Pathologists to guarantee the accuracy and reliability of results.

| Dilution protocol
The matrixes were divided into four groups for experiments: A (Siemens original diluent); B (pure water); C (0.9% NaCl); D (low-insulin serum). All of the 19 high-insulin samples were diluted by four different matrixes. The dilution factors were 1:2, 1:5, and 1:10. Among these, 1:2 and 1:5 dilutions were made using the instrument's automatic dilution procedure, and 1:10 dilution was made by using a twofold manual dilution method followed by a five-fold auto-dilution using the analyzer.

| Recovery
Insulin recovery results of the different matrixes are shown in Figure 2 and Table 2. The recovery rates were 86.4%-104.0% (original diluent), 73.2%-99.3% (pure water), 76.4%-101.3% (0.9% NaCl), and 84.2%-99.7% (low-insulin serum). Among these, the recovery results of the original diluent and low-insulin serum were better than those of the others and satisfied clinical requirements.  increasing the incidence of ovarian and endometrial cancer. 17,18 In addition, insulin level also increases in neonatal hypoglycemia.

| D ISCUSS I ON
Finally, the clinician also relies on a precise measurement of insulin to determine the peak time in the insulin release test and the rate of insulin increase compared with fasting. Thus, for many reasons, having an accurate measure of serum insulin concentration is necessary.
In this study, we compared the effectiveness of the four matrixes pure water, 0.9% NaCl, low-insulin serum, and the original manufacturer's diluent, to dilute high-insulin serum (>300 mU/L).
We found that using the original diluent was much better than the Based on these results, we conclude that dilution factors 1:2 and F I G U R E 1 Passing-Bablok regression of insulin level between the original and dilution results. A to D represent four different dilution matrixes including the original diluent, pure water, 0.9% NaCl, and low-insulin serum, respectively. The blue line represents 1:2, the red line represents 1:5, and the green line represents 1:10 F I G U R E 2 Recovery rates for the original and dilution results. A to D represent four different dilution matrixes including original diluent, water, 0.9% NaCl, and low-insulin serum. The numbers 1-3 represent the different dilution factors (2, 5, 10) 1:5 were much better than 1:10 for both pure water and 0.9% NaCl. Additionally, pure water and 0.9% NaCl are, of course, much easier to obtain.
To confirm the clinical application of this study, we collected the residual serum samples (>300 mU/L) and used different diluents to determine the actual distribution of samples (>300 mU/L).
Sixteen samples with insulin concentrations higher than 300 mU/L were collected from October 2019 to January 2020. Thus, we used the four diluent matrixes to detect the insulin concentration in samples (>300 mU/L). The distribution of insulin concentrations higher than 300 mU/L is shown in Figure 3. Among the 16 samples with insulin concentrations higher than 300 mU/L, only one sample showed a concentration higher than 1500 mU/L after 1:5 dilution.  Table 3.

TA B L E 3 Basic performance of sample diluents from various manufacturers
The results indicate that, where insulin reagents are used to measure serum insulin concentrations greater than 300 mU/L for clinical application, evaluation of the effects of different diluents should be considered to optimize cost and effectiveness.

| CON CLUS ION
From the perspective of health economics, this study confirms that high-insulin serum (higher than upper limit of linearity) can be diluted by pure water, 0.9% NaCl, or low-insulin serum on the Siemens ADVIA Centaur XP ® instrument, but the dilution factor should be 1:5 or lower. In order to standardize the operation, we recommend using 0.9% NaCl to dilute high-insulin serum (>300 mU/L) with a dilution factor of 1:5 on the Siemens ADVIA Centaur XP ® instrument. Under these conditions, the clinically reported range is 0.5-1500 mU/L, which meets clinical requirements. However, for specific samples (>1500 mU/L), it is better to use low-insulin serum samples for dilution with a dilution factor of 1:10.

This study was supported by grants from Education Reforming
Funding of PUMC [2018zlgc0119].

CO N FLI C T O F I NTE R E S T
There was no conflict of interest.