RU2810815C2 - Method of quantitative analysis of blood serum using ir spectroscopy - Google Patents

Abstract

FIELD: medicine; laboratory diagnostics.

SUBSTANCE: invention can be used for quantitative analysis of blood serum by IR spectroscopy. The IR spectrum of blood serum is taken in the range of 500–4,000 cm^-1. The intensities of absorption bands of 1,717 and 3,903 cm^-1 and the areas of absorption bands of 616, 3,750 and 3,903 cm^-1 are determined, from which the concentrations of 38 components of blood serum listed in Table 2 of the description are simultaneously determined, for the calculation of which the regression equation is

used

B_i means concentration of the i-th component, i=1,…,38; C_i means coefficients for calculation given in Table 1 of the description; H20 and H49 mean intensities of absorption bands of 1,717 and 3,919 cm^-1; S2, S42 and S49 mean areas of absorption bands of 617, 3,750 and 3,919 cm^-1.

EFFECT: method provides the ability to simultaneously determine the concentration of the main components of blood serum according to the characteristics of the IR spectra with an error of not more than 10% due to the quantitative analysis of blood serum by IR spectroscopy, which combines several approaches: using model solutions, which allows taking into account the complex composition of the sample matrix, carry out modeling in a wide range of concentrations, which corresponds to the range of concentrations of components of real blood serum, and also use chemometric methods to construct a regression equation and calculate the concentrations of components, thus improving accuracy.

1 cl, 3 dwg, 2 tbl

Description

The invention relates to the field of medicine, namely to methods for studying biological fluid using physical and chemical methods. Blood plasma and serum contain more than 300 types of proteins, as well as carbohydrates, lipids and amino acids, and up to 114,000 known metabolites in varying concentrations. Determination of clinical parameters of serum and blood plasma is widely used to diagnose various diseases, as well as as a way to monitor the effectiveness of treatment. In this regard, the demand for clinical tests is growing, leading to the use of automated analyzers, many of which are based on colorimetric reactions. These methods require the use of expensive and specific reagents. Given the massive nature of such studies, there is a clear need for new and low-cost alternative analytical procedures, especially as screening tools, as a possible indicator in situations where economic resources are limited and point-of-care diagnostic evidence is required.

Vibrational spectroscopy methods, in particular IR spectroscopy, can be used for these purposes, since they do not require labeling, are economical, easy to operate, and require minimal sample preparation. Sensitivity to subtle changes in biochemical composition makes them ideal diagnostic tools, and recent advances in technology and data analysis allow rapid and non-invasive analysis of body fluids. However, determination of clinical parameters in sera using vibrational spectroscopy can be difficult due to high matrix complexity and low concentration levels of some analytes, as chemometric algorithms are usually required to remove matrix effects. In general, the idea of determining clinical parameters of serum from IR spectra is not new and the potential for quantitative and semi-quantitative analysis of metabolites in plasma, serum and whole blood has been confirmed by a number of studies. Currently, using the method of IR spectroscopy of blood serum, it is possible to diagnose diseases such as urolithiasis (Patent RU 2 666 948 C1 dated July 6, 2017), diabetes mellitus (Patent RU 2 405 148 C2 dated February 25, 2009, Patent RU 2 198 402 C2 dated 03/11/2001), bronchial asthma (Patent RU 2 372 621 C1 dated 06/16/2008), for the differential diagnosis of schizophrenia and malignant brain tumors (Patent RU 2 664 444 C1 dated 07/03/2017 ), early differential diagnosis of acute pancreatitis and pancreatic cancer (Patent RU 2 602 689 C1 dated 07/01/2015). In these methods, the preparation of a blood serum sample is carried out by drying the sample and grinding the dry residue, followed by its suspension in vaseline oil. The blood serum sample is examined using IR spectroscopy in the spectral region of 1200-1000 cm ^-1 and the peak heights of the absorption bands are determined. Determination methods using IR spectroscopy have already been developed for a number of blood serum parameters. Thus, blood glucose, an important parameter for the control of diabetes and other common diseases, was determined in serum with acceptable accuracy. It has been shown that mid-infrared spectroscopy can serve as an alternative basis for the clinical measurement of urea and glucose in blood serum. The literature reports promising results for the determination of important serum parameters such as urea, total protein, albumin, triglycerides or total cholesterol. However, the possibility of simultaneous determination of several blood parameters has not been sufficiently studied. The closest in technical essence to the claimed method is the method of analyzing dry serum deposits using transmission spectroscopy, which revealed the possibility of quantitative determination of eight serum components (RA Shaw, S. Kotowich, M. Leroux, HH Mantsch. “Multianalyte serum analysis using mid-infrared spectroscopy” Ann. Clin. Biochem. An Int. J. Biochem. Lab. Med. 1998. 35: 624-632). The disadvantage of this method is the small number of components to be determined and low accuracy (for a number of parameters the error exceeds 100%).

The objective of the technical solution being developed is to simultaneously determine the concentration of the main components of blood serum based on the characteristics of the IR spectra with an error of no more than 10%.

This technical result is achieved by the fact that in the method of quantitative analysis of blood serum using IR spectroscopy, several approaches are combined, namely: (1) model solutions are used, which makes it possible to take into account the complex composition of the sample matrix; (2) carry out simulations over a wide range of concentrations, which corresponds to the range of concentrations of components of real blood serum; (3) they use chemometric methods to construct a regression equation and calculate the concentrations of components, which improves the accuracy. The described combination of methods makes it possible to simultaneously determine the concentrations of 38 components using the intensities of absorption bands of 1717 and 3903 cm ^-1 and the areas of absorption bands of 616, 3750 and 3903 cm ^-1 with an error of less than 10%. The results obtained confirm the potential clinical use of IR spectroscopy as a reagent-free rapid method for analyzing blood serum.

The method is illustrated by pictures:

figure 1 - Example of the IR spectrum of blood serum (A); Comparison of the IR spectrum of blood serum and model solution (B).

figure 2 - IR spectra of model solutions when the original solution is diluted 5 times (curve 1), 3 times (curve 2) and 2 times (curve 3), respectively.

figure 3 - Changes in the intensities and areas of characteristic absorption bands upon dilution of model solutions.

The method is carried out as follows.

To prepare model solutions, a set of lyophilized control sera based on bovine blood serum, certified by approximately 38 indicators (SPINREACT, SA Ctra. Santa Coloma, Spain), was used. The solution of the initial model serum was diluted 2, 3, 5, 7, 10 times (in three parallels). Aliquots of 50 μl are removed and dried for 30 minutes on a zinc selenide substrate in an incubator at 37°C. IR absorption spectra are recorded on an FT-801 IR Fourier spectrometer (SIMEX) in the range of 500-4000 cm ^-1 . The spectra are recorded with a number of scans of 32 with a resolution of 4 cm ^-1 .

Select absorption bands that are present in the IR spectra of the serum, regardless of dilution. The intensity (H) and area (S) for the corresponding absorption bands were determined.

S and H are selected in such absorption bands that have a high correlation with the concentrations of the components ( B _i ). They turned out to be H20, H49, S2, S42, S49, where H20 and H49 are the intensities of the absorption bands 1717 and 3919 cm ^-1 ; S2, S42 and S49 - absorption band areas 617, 3750 and 3919 cm ^-1 . Then a Delphi program is created and a combination of K of these parameters is selected that has a high correlation with Bi when the concentration changes (equation 1, 2). Since the operations of multiplication and division of variables can enhance correlations, and multiplication by a constant, addition and subtraction of variables does not affect correlations, as a result, from all combinations K those for which there is a strong correlation with each B _i (Pearson correlation coefficient r>0.999) are selected and constitute their weighted average sum, which has a minimum sum of squared deviations of the values from B _i . The vector of weighting coefficients in this sum is sought by solving the nonlinear problem using the generalized reduced gradient method. Next, the weighting coefficients Ci are identified, at which the square of the deviation of Bi from the value of K multiplied by Ci is minimal.

Next, the model is tested and the constant error is adjusted by subtracting the error, since the calculated values are always approximately 5% larger (equation 3). Model development is carried out on the basis of model solutions of normal blood serum (SPINREACT normal), while model solutions of pathological serum (SPINREACT pathologic) are used for testing. The concentrations of all components in the test model solution correspond to the concentration range of the initial model solutions.

The IR spectra of blood serum and the proposed model solutions are compared (Fig. 1). It has been shown that the contribution of proteins to the IR spectra is in the frequency ranges 3400-3030 cm ^-1 , 1720-1480 cm ^-1 and 1301-1229 cm ^-1 - region I; lipids in the ranges of 3020-2819 cm ^-1 , 1750-1725 cm ^-1 and 1480-1430 cm ^-1 - region II; carbohydrates and nucleic acids (DNA/RNA) in the frequency range 1200-900 cm ^-1 - region III (Fig. 1a). The maximum differences between the IR spectra of serum and model solutions are observed in regions III and IV and are due to the absence of a number of carbohydrates and nucleic acids in model mixtures (Fig. 1b).

IR spectra of model solutions with different concentrations of constituent substances are obtained (Fig. 2a). To construct the regression equation, four absorption bands are selected: H20 and H49 - absorption band intensities of 1717 and 3903 cm ^-1 ; S2, S42 and S49 are the areas of absorption bands 616, 3750 and 3903 cm ^-1 (Fig. 2). In fig. 2c - 2e show the change in absorption bands with changes in the concentrations of model solutions. When diluting the corresponding solutions, the intensities and areas of the selected absorption bands decrease (Fig. 3).

Based on the proposed algorithm, an equation is obtained for calculating the concentrations of the components of the model solution (B _i ), where i=1,...38 (equation 3). The coefficients ( _Ci ) for calculation are given in Table 1. Table 2 shows a complete list of the components of the model solution to be determined, the concentration range that was used to build the model, as well as the true concentrations of the test solution found during testing of the model. According to the data presented, in all cases the error in determining the concentration does not exceed 10% (Table 2).

Table 1. Coefficients (C _i ) for calculating the concentrations of components B _i (i=1, ..., 38) i 1 2 3 4 5 6 7 8 9 10 C _i 1019 136.7 61.42 458.7 1507 21.46 5.474 32.66 7.744 2.815 i eleven 12 13 14 15 16 17 18 19 20 C _i 2.768 4.447 3.685 5.474 167.9 51.32 65.93 106.8 37.16 3.81 i 21 22 23 24 25 26 27 28 29 thirty C _i 2.084 227 3.623 531.8 78.53 53.8 317.2 768.2 8554 88.48 i 31 32 33 34 35 36 37 38 C _i 329.7 309.4 247.2 220.8 131.4 1157 65.47 489.8

Table 2. List of components to be determined and range of concentrations of model solutions No. Component Units Model concentration range Test East Find test Δ, % IN 1 Uric acid µmol/l 58.40 - 292.00 131.00 134.51 2.68 AT 2 Bilirubin (total) µmol/l 5.78 - 28.90 17.58 16.49 6.20 AT 3 Bilirubin (connection) µmol/l 3.06 - 15.30 7.90 8.51 7.74 AT 4 Creatinine µmol/l 18.02 - 90.10 59.00 63.65 7.88 AT 5 Fructosamine µmol/l 147.40 - 737.00 193.80 182.22 5.98 AT 6 Glucose mmol/l 1.04 - 5.22 2.76 2.95 6.87 AT 7 Lactate mmol/l 0.30 - 1.52 0.70 0.75 6.61 AT 8 Urea mmol/l 1.39 - 6.97 4.20 4.44 5.81 AT 9 Cholesterol mmol/l 0.48 - 2.39 1.00 1.07 7.21 AT 10 HDL (high density lipoprotein) mmol/l 0.14 - 0.72 0.36 0.34 5.74 AT 11 LDL (low density lipoprotein) mmol/l 0.27 - 1.34 0.36 0.38 6.98 AT 12 Phospholipids mmol/l 0.32 - 1.60 0.57 0.62 8.01 B13 Triglycerides mmol/l 0.25 - 1.24 0.47 0.43 8.29 B14 Calcium mmol/l 0.44 - 2.22 0.70 0.75 6.61 B15 Chlorides mmol/l 16.20 - 81.00 21.60 23.70 9.72 B16 Copper µmol/l 3.64 - 18.20 6.60 6.18 6.36 B17 Iron µmol/l 3.68 - 18.40 8.48 9.13 7.64 B18 Iron binding capacity µmol/l 13.74 - 68.70 13.74 15.11 10.00 B19 Potassium mmol/l 0.70 - 3.51 4.78 5.19 8.55 IN 20 Lithium mmol/l 0.20 - 1.00 0.49 0.52 6.68 AT 21 Magnesium mmol/l 0.17 - 0.87 0.27 0.25 5.90 B22 Sodium mmol/l 25.20 - 126.00 29.20 31.73 8.66 B23 Phosphorus mmol/l 0.27 - 1.36 0.47 0.50 7.75 B24 Zinc µg/ml 77.40 - 387.00 68.40 63.25 7.52 B25 Total protein g/l 13.24 - 66.20 10.10 10.81 7.02 B26 Albumen g/l 9.78 - 48.90 6.92 7.47 7.89 B27 Amylase E/l 16.40 - 82.00 40.80 44.05 7.97 B28 Creatine kinase E/l 29.00 - 145.00 98.80 91.94 6.95 B29 Cholinesterase E/l 1107.80 - 5539.00 1100.20 1181.62 7.40 B30 Acid phosphatase E/l 6.72 - 33.60 11.38 12.21 7.28 B31 Alkaline phosphatase E/l 15.20 - 76.00 42.40 45.00 6.12 B32 Gammaglutamyltransferase E/l 8.40 - 42.00 39.80 36.30 8.79 B33 AST E/l 10.60 - 53.00 31.80 34.71 9.16 B34 ALT E/l 10.40 - 52.00 28.40 30.61 7.79 B35 Lipase E/l 7.86 - 39.30 16.90 15.69 7.15 B36 Lactate dehydrogenase E/l 78.00 - 390.00 148.80 138.91 6.65 B37 Glutamate dehydrogenase E/l 4.98 - 24.90 8.42 8.98 6.65 B38 Alpha hydroxybutyrate
dehydrogenase E/l 31.00 - 155.00 63.00 67.94 7.85

Thus, in our study, we demonstrated for the first time the possibility of simultaneous determination of 38 parameters in blood serum with minimal sample preparation and an error not exceeding 10%, and numerous studies confirming the principle have revealed the potential clinical use of this method.

Claims (6)

A method for quantitative analysis of blood serum using IR spectroscopy, which consists in taking the IR spectrum of blood serum in the range of 500-4000 cm ^-1 , characterized in that the intensity of absorption bands of 1717 and 3903 cm ^-1 and the area of absorption bands of 616 are determined, 3750 and 3903 cm ^-1 , from which the concentrations of 38 blood serum components listed in Table 2 of the description are simultaneously determined, for the calculation of which the regression equation is used B _i – concentration of the i-th component, i=1,…,38; C _i – coefficients for calculation given in Table 1 of the description; H20 and H49 – intensities of absorption bands 1717 and 3919 cm ^-1 ; S2, S42 and S49 – absorption band areas 617, 3750 and 3919 cm ^-1 .