RU2750525C1 - Method for diagnosing disseminated pulmonary tuberculosis in patients with hiv infection and negative results of bacterioscopic and immunological methods - Google Patents

Abstract

FIELD: medicine, phthisiology in particular.

SUBSTANCE: invention relates to medicine, namely to phthisiology and pulmonology, and can be used for the diagnosis of disseminated pulmonary tuberculosis in patients with HIV infection. The method involves examining the patient's white blood cell count in a general blood test. Additionally, the level of CD4-lymphocytes in the blood is determined, the level of hepcidin and serum iron is determined in the blood serum, the number of red blood cells is determined in the general blood test, the probability of tuberculosis (p) is calculated using the formula р=1/(1+2,7182^{(5,0016-0,0000081*H+0,1993*I+0,3874*W+1,80963*R+0,01844*С)}), where H is the level of hepcidin, pg/ml; I is the level of serum iron, mmol/l; W is the number of white blood cells, * 109/l; R is the number of red blood cells, *1012/l; C is the number of CD4 lymphocytes, cl/ml; 2.7182 is the base of the natural logarithm; 5.0016 constant and 0.0000081, 0.1993, 0.3874, 1.80963, 0.01844 coefficients obtained experimentally. According to the obtained value of p>0.5, the conclusion is made about the tuberculous etiology of disseminated lung damage, p ≤ 0.5 about the absence of tuberculosis.

EFFECT: use of the invention makes it possible to increase the efficiency of differential diagnosis of tuberculosis and pneumonia in persons with HIV infection and negative results of bacterioscopic and immunological methods by using objective laboratory quantitative indicators of blood analysis.

1 cl, 2 ex

Description

The invention relates to medicine, namely to phthisiology and pulmonology.

Diagnosis of pulmonary tuberculosis, according to the Federal Clinical Guidelines, is carried out taking into account the clinical picture of the disease, anamnesis data, the results of X-ray examination methods, immunological skin tests, molecular genetic and laboratory methods (bacterioscopic, microbiological, pathomorphological studies) [1]. Despite the improvement in the epidemiological situation with regard to tuberculosis, there is an increase in the incidence of tuberculosis among HIV-infected patients. The most common form of tuberculosis in patients with HIV infection in the advanced stages is disseminated tuberculosis. Verification of tuberculous etiology in dissemination syndrome presents great difficulties. Laboratory methods for detecting mycobacteria are more often negative. The low efficiency of diagnostic methods is noted with the similarity of clinical-anamnestic and social factors of disseminated lung diseases of infectious etiology. An important link in the pathogenesis of dissemination of tuberculous etiology is the stage of bacteremia. The search for differences in blood tests seems to be promising for their use in the diagnosis of tuberculosis.

There is a known method for diagnosing tuberculosis [2], which consists in the fact that mononuclear cells of peripheral blood are isolated, cultivated in a nutrient medium, a sample of the supernatant of a culture of mononuclear cells is obtained and ELISA is carried out using a solution of antigens of mycobacterium tuberculosis as a sorbent; about the presence of tuberculosis. The disadvantage of this method is its multistage and duration of execution.

A known method for the diagnosis of tuberculosis [3], based on the determination of a complex of serum cytokines: tumor necrosis factor - alpha, interleukin - 10 and soluble receptor interleukin - 6. With indicators of tumor necrosis factor - alpha from 19.72 to 23.87 pg / ml, interleukin - 10 from 15.48 to 21.86 pg / ml and soluble interleukin receptor - 6 from 770.76 to 1800.31 pg / ml with a viral load of more than 10,000 cop / ml and CD4 + less than 500 TB is diagnosed in HIV-infected patients ... The disadvantage of this method is the high cost of determining serum cytokines.

A known method for the diagnosis of pulmonary tuberculosis [4], including the incubation of whole blood with mycobacterial antigens, representing a mixture of proteins ESAT-6, CFP-10, TB 7.7, and without them, centrifugation of the sample with plasma separation, determination of gamma-interferon in supernatants, characterized by that in the supernatants of persons infected with mycobacteria, in addition to antigen-induced production of gamma-interferon, the content of antigen-induced interleukin-6 and spontaneous production of transforming growth factor-alpha are determined and at a level of gamma-interferon ≥6.4 IU / ml, or interleukin-6 ≥2039 pg / ml, or transforming growth factor-alpha ≥17.0 pg / ml, active tuberculosis infection is noted. The disadvantage of this method is the lack of information about the characteristics of the production of these substances in the late stages of HIV infection.

A known method for the diagnosis of tuberculosis [5], based on the determination of biomarkers IP-10, interleukins IL-6, IL-10, IL-4, FOXP3 and IL-12 in the blood. The disadvantage of this method is the high cost of determining interleukins.

There is a known method for the differential diagnosis of pneumonia and pulmonary tuberculosis [6] using a blood test, characterized in that a patient with an inflammatory infiltrate in the lung counts the number of leukocytes, determines the lipid-releasing ability of blood leukocytes (LLSL), and with an LLSL index equal to or less than 0.119 mmol / l, infiltrative pulmonary tuberculosis is diagnosed, and more than 0.119 mmol / l - pneumonia. The disadvantage of this method is its knowledge in a group of patients with only one clinical form of tuberculosis - infiltrative. In addition, the possibility of its use in persons with HIV infection has not been shown. We have chosen this method as a prototype.

The aim of the claimed invention is to create a method for the diagnosis of disseminated pulmonary tuberculosis in patients with HIV infection and negative results of bacterioscopic and immunological methods.

This goal is achieved by the fact that the number of erythrocytes is additionally determined in the general blood test; serum levels of hepcidin and serum iron are determined; determine the level of blood CD4 lymphocytes; calculate the probability of tuberculosis (p) by the formula:

p = 1 / (1 + 27182 ^{(5.0016-0.0000081 * G + 0.1993 * W + 0.3874 * L + 1.80963 * E + 0.01844 * C)} ),

where G is the level of hepcidin, pg / ml;

W - serum iron level, μmol / l;

L - the number of leukocytes, * 10 ⁹ / l;

E - the number of erythrocytes, * 10 ¹² / l;

C - the number of CD4-lymphocytes, cells / μl;

2.7182 is the base of the natural logarithm;

5.0016 - constant and 0.0000081, 0.1993, 0.3874, 1.80963, 0.01844 - coefficients obtained experimentally;

and the obtained value of p> 0.5 makes a conclusion about the tuberculous etiology of disseminated lung lesions, p≤0.5 - about the absence of tuberculosis.

These coefficients and the constant were obtained by us during mathematical and statistical processing using the method of logistic regression with subsequent logit transformation of clinical and laboratory data of examination of 96 patients with HIV infection and pulmonary dissemination syndrome, 42 of whom had tuberculosis and 44 - pneumonia. Based on the data obtained, the claimed formula is derived. To analyze the relationship between a qualitative binary trait (in our case, the presence or absence of tuberculosis) acting as a dependent variable and a subset of quantitative traits (laboratory parameters) acting as independent variables (predictors), we applied the statistical method of logistic regression with a stepwise algorithm inclusion and exclusion of predictors. The accepted level of significance was p <0.05. In patients with HIV infection and pulmonary dissemination syndrome, 5 predictors with the highest value were selected: hepcidin, serum iron, erythrocytes, leukocytes, CD4 lymphocytes. The assessment of the quality of the regression was carried out on the basis of calculating the coefficient of determination R2, which takes values from 0 to 1 and indicates a high quality of regression when approaching one (R2 according to Nagelkerk = 0.7736 - good quality of regression). To assess the diagnostic efficiency of the method, taking into account the consequences of false decisions, ROC curves (Receiver Operating Characteristic curve) were used. When constructing the ROC curve, the specificity of the method is taken as the abscissa axis, and the specificity is taken as the ordinate. The sensitivity of the method was 92.86% and the specificity was 90.91%. The percentage of correctly classified cases was 91.86%. The ROC-analysis method was used to determine the quantitative value of the reliability of the difference in the information content of the studied methods. For this, the area under the curves (AUC) was calculated. In our model, AUC = 0.958, which corresponds to high information content.

Comparison of the proposed method with others known in modern medicine in medicine has shown its compliance with the criteria and invention.

The method is implemented as follows: in a patient with HIV infection and suspected disseminated pulmonary tuberculosis, who has negative results of bacterioscopic and immunological methods, hepcidin and serum iron are determined in the blood serum. The level of CD4-lymphocytes is determined. In a general blood test, the level of erythrocytes and leukocytes is determined. Each indicator is given an experimentally established weight value in the form of its product by a constant coefficient, the obtained values are summed up and a constant is added. According to the formula obtained, the probability (p) of tuberculosis is calculated:

p = 1 / (1 + 27182 ^{(5.0016-0.0000081 * G + 0.1993 * W + 0.3874 * L + 1.80963 * E + 0.01844 * C)} ),

where G is the level of hepcidin, pg / ml;

W - serum iron level, μmol / l;

L - the number of leukocytes, * 10 ⁹ / l;

E - the number of erythrocytes, * 10 ¹² / l;

C - the number of CD4-lymphocytes, cells / μl;

2.7182 is the base of the natural logarithm;

5.0016 - constant and 0.0000081, 0.1993, 0.3874, 1.80963, 0.01844 - coefficients obtained experimentally;

and by the value of p> 0.5, a conclusion is made about the tuberculous etiology of disseminated lung lesions, p≤0.5 - about the absence of tuberculosis.

The claimed invention is illustrated by the following clinical examples.

Example 1.

In patient N., 43 years old with HIV infection and pulmonary dissemination syndrome, no mycobacteria were found on the X-ray microscopy of sputum and bronchial lavages. Mycobacteria were not detected by polymerase chain reaction. The sputum was sent to the bacteriological laboratory for inoculation. Skin test with Diaskintest is negative. On the first day of hospitalization, venous blood was taken. The following values of diagnostic parameters were determined: hepcidin 10885 pg / ml, serum iron 4.9 μmol / l, leukocytes 5.8 * 10 ⁹ / l, erythrocytes 3.5 * 10 ¹² / l, CD4 200 cells / μl. The value of the probability of tuberculosis, taking into account these indicators, is equal to:

p = 1 / (1 + 27182 ^{(5.0016-0.0000081 * 10885 + 0.1993 * 4.9 + 0.3874 * 5.8 + 1.80963 * 3.5 + 0.01844 * 200)} ) = 0.867.

Since 0.867 is more than 0.5, it is concluded that this patient has tuberculosis. The patient started a specific etiotropic anti-tuberculosis therapy. After 18 days, a positive culture result was obtained, the growth of the culture of Mycobacterium tuberculosis was detected, the diagnosis of tuberculosis was confirmed.

Example 2.

In a 37-year-old patient R. with HIV infection and pulmonary dissemination syndrome, no mycobacteria were found on the X-ray microscopy of sputum and bronchial lavages. Mycobacteria were not detected by polymerase chain reaction. The sputum was sent to the bacteriological laboratory for inoculation. Skin test with Diaskintest is negative. On the second day of hospitalization, venous blood was taken, the following values of diagnostic indicators were determined: hepcidin 6665 pg / ml, serum iron 9.9 μmol / l, leukocytes 3.2 * 10 ⁹ / l, erythrocytes 2.51 * 10 ¹² / l, CD4 11 cells / μL. The value of the probability of tuberculosis, taking into account these indicators, is equal to:

p = 1 / (1 + 27182 ^{(5.0016-0.0000081 * 6665 + 0.1993 * 9.9 + 0.3874 * 3.2 + 1.80963 * 2.51 + 0.01844 * 11)} ) = 0.072.

Since 0.072 is less than 0.5, it is concluded that there is no tuberculosis in this patient. No specific etiotropic anti-tuberculosis therapy was performed. After 14 days, the patient was discharged with improvement and pronounced positive dynamics on the roentgenogram. The culture method (seeding) also did not confirm the diagnosis of tuberculosis.

Thus, the method makes it possible to quickly, informatively, reliably diagnose tuberculosis in patients with HIV infection and pulmonary dissemination syndrome and can be used in cases when the results of bacterioscopic and immunological methods are negative if tuberculosis is suspected.

Increasing the efficiency of differential diagnosis of tuberculosis and pneumonia in persons with HIV infection through the use of objective laboratory quantitative blood test indicators is an advantage and advantage of the proposed method. The industrial applicability of the method is confirmed by the possibility of its reproduction in the conditions of any medical institution with a department of clinical laboratory diagnostics.

Information sources:

1. Guidelines for improving the diagnosis and treatment of respiratory tuberculosis, approved. Order of the Ministry of Health of the Russian Federation of December 29, 2014 No. 951.

2. A method for the diagnosis of tuberculosis and differential diagnosis of tuberculosis and latent tuberculosis infection. Kazanova A.S., Lyadova I.V., Kondratyuk N.A., Panteleev A.V., Vasilyeva I.A., Tarakanova Yu.N. Invention patent RU 2576833 C1, 03/10/2016. Application No. 2014114430/15 dated 04/14/2014.

3. Method for early diagnosis of tuberculosis in HIV infection. Sklyar L.F., Markelova E.V., Sotnichenko S.A. Invention patent RU 2416801 C1, 20.04.2011. Application No. 2009137802/15 dated 12.10.2009.

4. Method for the diagnosis of pulmonary tuberculosis. Vasilyeva E.V., Verbov V.N., Totolyan A.A., Lyadova I.V., Nikitina I.Yu. Invention patent RU 2503005 C1, 27.12.2013. Application No. 2012128311/15 dated 02.07.2012.

5. Host biomarkers of clinical relevance in tuberculosis: review of gene and protein expression studies. John S.H., Kenneth J., Gandhe A.S. Biomarkers. 2012 Feb; 17 (1): l-8.

6. Method for differential diagnosis of pneumonia and infiltrative pulmonary tuberculosis. Varlamov P.N., Mishlanov V.Yu., Varlamov O.P., Morozova N.S. Patent for invention RU 2444015 C1, 27.02.2012.

Application No. 2011102649/15 dated 01.24.2011.

Claims (10)

A method for diagnosing disseminated pulmonary tuberculosis in patients with HIV infection and negative results of bacterioscopic and immunological methods, including examining the number of leukocytes in a general blood test in a patient, characterized in that the level of blood CD4-lymphocytes is additionally determined; serum levels of hepcidin and serum iron are determined; in a general blood test, the number of erythrocytes is determined; calculate the probability of tuberculosis (p) by the formula p = 1 / (1 + 2.7182 ^{(5.0016-0.0000081 * G + 0.1993 * W + 0.3874 * L + 1.80963 * E + 0.01844 * C)} ), where G is the level of hepcidin, pg / ml; W - serum iron level, μmol / l; L - the number of leukocytes, * 10 ⁹ / l; E - the number of erythrocytes, * 10 ¹² / l; C - the number of CD4-lymphocytes, cells / μl; 2.7182 is the base of the natural logarithm; 5.0016 - constant and 0.0000081, 0.1993, 0.3874, 1.80963, 0.01844 - coefficients obtained experimentally; and the obtained value of p> 0.5 makes a conclusion about the tuberculous etiology of disseminated lung lesions, p≤0.5 - about the absence of tuberculosis.