Urgency of an issue COPD

Urgency of an issue. COPD is a disease that evolves over decades often with a substantial amount of emphysema and airflow obstruction at the time of diagnosis (1). Although airflow obstruction parameters in the pulmonary function test (PFT) by spirometry is essential in COPD diagnosis. These parameters fail to quantify the proportionate impact of emphysema and small airways disease individually (2) and these tests are unable to discriminate emphysema from airways disease (3) specially when emphysema does not elicit airway obstruction in every individual (4).

For this reason, in recent years tools as computed tomography (CT) have been used to further characterize the lung structure (3). Moreover, serial CT imaging improved the assessment of the lung architecture and is becoming increasingly important (3). Software programs have since been developed, which can objectively measure the severity of emphysema (5, 6).

Quantitative CT, and in particular CT densitometry has been demonstrated to be a surrogate for emphysema, and lung function measures airflow limitation as an indication of airways disease including COPD (1, 7).

An alternative method for quantitating pulmonary emphysema was introduced by FLENLEY's group in Edinburgh. This concept is based on percentile densities (PD) that are derived from the frequency histogram of the pixel attenuation values of the lung. Thus, the 15 th percentile density is defined as the cut-off density at which 15% of all pixels have lower densities (8-9). For example, a 15th percentile density (PD15) indicates the HU value below which the lower 15% of all voxels are distributed. The lower the PD15 values in grams per liter (ie, closer to 0), the more emphysema is present.This approach has been typically used in COPD studies (1,10).

Objectives: To determine the significance of MSCT densitometric parameters, depending on the severity of COPD, in order to assess the severity of structural lesions of the lung tissue.

Material and methods: The results of clinical and radiological studies of 62 patients with COPD under the age of 65 depending on the generally accepted criteria for COPD severity were analyzed, among which 14 (22.3%) patients had asthma-COPD overlap syndrome (ACOS).

The severity of bronchial obstruction was determined according to GOLD-17 (11) according to spirometry data. In the presence of the Tiffeneau index (FEV1/FVC) less than 0.70:

GOLD 1 - FEV1 more than 80%; GOLD 2 - FEV1 more than 50-80%; GOLD 3 - FEV1 30-50%; GOLD 4 - FEV1 less than 30%.

The severity of COPD (A, B, C, D) was determined according to GOLD taking into account the severity of bronchial obstruction, the severity of dyspnea (according to the modified scale of the British Medical Council (mMRC) and the frequency of exacerbations of the disease) (12).

MSCT is performed on a SOMATOM EMOTION device from Siemens or ASTEION from TOSHIBA. The technical parameters of the scan were as follows: 120 KVp, 250 mA, tube rotation speed 0.5 s, FOV 400 x 400 mm, matrix 512 x 512, slice thickness 1 mm. Scanning was done at maximum inspiration (inspiratory) and maximum exhalation (expiratory). The quantitative assessment was based on the study of tomograms obtained at the level of the upper and lower parts of the lungs during inhalation and exhalation. For this purpose, in this study we used the Canon application “Lung density analysis”. The principle of digital processing of tomograms using the Lung density analysis program is based on determining the optical density of tissues in Hounsfield units (HU), in addition the program allows to calculate the following indicators:

Low density volume (LDV) -in ml;

Medium density volume (MDV) -in ml;

High density volume (HDV) -in ml;

Lung volume (no high density volume) (LV) -in ml;

15 Percent. Density (PD15) - This is the density point at which 15% of the voxels of the lungs have a low lung density (can be converted to lung tissue density values (g/L) by adding 1,000 to the HU),

Additionally there are calculated:

Low Density Volume Index (LDVI) -in %;

Upper Lung Low Density Volume Index (ULLDVI) -in%;

Lower Lung Low Density Volume Index (LLLDVI) -in%;

Upper/Lower Lung Low Density Volume Index Ratio (U/L LLD IR),

PD15 (g/l), LDVI (%) and U/L LLD IR were selected for further analysis, which are included in the standard automatic density analysis of CT images (Lung density analysis).

Statistical processing was performed using the statistical software package PSRR (open program, does not require a license). For data processing, the methods of descriptive statistics and correlation analysis were applied, taking into account the data checking for normality by the Kolmogorov-Smirnov test. The qualitative indicators were analyzed using the x2 - Pearson criterion; for comparison of quantitative indicators with a normal distribution of t - Student's criterion; if abnormal distribution - the criterion of Man-Whitney. Correlation analysis was performed using the Spearman criterion. The difference was considered statistically significant when the probability of the null hypothesis was less than 5% (p <0.05).

Research results. It should be noted that all the indicators that were analyzed had a fairly large range of individual values (Table 1). In particular, the PD15 value ranged from 1.0 to 101.0 g/l and had an abnormal distribution according to the Kolmagorov- Smirnov criterion. In patients with COPD, the PD15 value ranged from 1.0 g / l to 71.0 g / l, the median was 26.5 [7.8; 37.0] g / l; in patients with ACOS from 3.0 g / l to 101.0 g / l, the median is 8.3 [4.9; 25.3] g / l (p = 051 according to the Man-Whitney test compared with patients with COPD). That is, the lowest values of PD15 were observed in patients with ACOS. A decrease in PD15 indicates large emphysematous changes in lung tissue. The distribution of the percentage of Low Density Volume Index (LDVI), or emphysema index was normal (p = 0.405), but also had a fairly large distribution of individual values - from 4.3% to 87%.

Table 1

INSPIRATORY DENSITOMETRIC INDICES OF MSCT IN PATIENTS INCLUDED IN THE STUDY

In patients with COPD, LDVI ranged from 19.0% to 87.0%, the average value was (60.3 ± 16.6)% (95% CI 55.5% - 65.1%); in patients with ACOS - from 5.8% to 84.1%, on average (69.4 ± 19.7)% (95% CI - 37.8% - 64.4%) (p = 0.086). Thus, the highest LDVI values were observed in patients with ACOS, which were not significantly more than in patients with COPD.

The distribution of the density ratio of the upper and lower lungs (U / L LLD IR) was not normal (p = 007 by the Kolmagorov-Smirnov criterion), so the median, 25th and 25th quartile, was used to describe it. In patients with COPD, this index ranged from 0.79 (the percentage of low density voxels in the lower sections was greater than the upper ones) to 214 (the percentage of low density voxels in the lower sections was less than the upper ones), a median 1.02 [0.96; 1.08]; in patients with ACOS, U / L LLD IR ranged from 0.67 to 1.17, the median was 0.91 [0,8,81; 1.02] (p = 0.015 compared with patients with COPD according to the Man-Whitney test). Normally, this indicator should be close to 1.0 or slightly more due to the large airiness of the lungs in the upper sections. The obtained data prove that in patients with COPD, emphysematous changes were more characteristic for the upper lung, while in patients with ACOS - emphysematous observed more in the lower lungs. But this index is difficult to interpret in the aggregate, since with diffuse emphysema with a uniform lesion of the lungs, this index will also be close to 1.0. But in the context of the research, the more interesting is index of total lesion of lung tissue, characterizing the spread of emphysema - PD15 and LDVI, rather than the prevailing process localization (U / L LLD IR). According to the results of the PD15 analysis, a clear dependence of its magnitude on COPD stage was established (Fig. 1).

Fig. 1. Box diagram of the distribution of inspiratory PD15 in patients with COPD, depending on the stage of the disease. Note: box borders - 25th and 75th percentiles; the line in the middle of the box is the median; whiskers - arithmetic mean ± standard deviation; points are data outliers with an observation number.

According to the data obtained, at stage A (n = 5), the median PD15 was 59.0 g / l [39.0; 66.5] (from 38.0 g / l to 71.0 g / l); at stage B (n = 15) - 29.5 g / l [12.6; 35.0] (from 4.5 g / l to 71.0 g / l); at stage C (n = 18) - 18 g / l [6.0; 29.0] (from 1.0 g / l to 39.0 g / l); at stage D (n = 10) - 5.6 g / l [4.8; 8.9] (from 3.0 g / l to 28.0 g / l). The 15th percentile of pulmonary tissue density decreases significantly with an increase in the stage of the disease (p <0.05 at each stage compared with the previous Mana-Whitney criterion). A slightly different distribution was obtained by analyzing the distribution of PD15 value depending on the degree of bronchial obstruction (Fig. 2).

Fig. 2. Box diagram of the distribution of inspiratory PD15 in patients with COPD, depending on the degree of bronchial obstruction (GOLD 2017).

The PD15 value in patients with COPD GOLD1 was 38.0 [32.0; 62.0] g / l (from 27.0 g / l to 71.0 g / l); GOLD2 - 30.0 [17.5; 45.0] g / l (from 4.2 g / l to 71.0 g / l); GOLD3 - 10.4 [4.9; 37.0] g / l (from 4.5 g / l to 39.0 g, l); GOLD4 - 8.4 [5.0; 28; 0] g / l (from 1.0 g / l to 36g / l). As in the previous comparison, the 15th percentile of lung tissue density decreases with increasing severity of bronchial obstruction, but the difference between GOLD1 and GOLD2, between GOLD2 and GOLD3, and also between GOLD3 and GOLD4 is not statistically significant (p <0.05 in each comparison with the Man-Whitney test).

A similar analysis was performed on the LDVI indicator (Fig. 3, 4). The mean value of LDVI in stage A COPD patients was (42.6 ± 6.7)% (95% CI 34.3 - 50.9%); at stage B - (51.1 ± 15.6)% (95% CI 42.5 -

59.8%); at stage C - (63.7 ± 14.2) (95% CI 56.7 -

70.8%); at stage D - (76.5 ± 5.5)% (95% CI 72.5 -

80.5%). That is, the emphysematous index increases

with increasing stage of the disease (p = 0.107 between A and B, p = 0.023 between B and C, and p = 0.002 between C and D).

Fig. 3. The value of inspiratory LDVI in patients with COPD depending on the stage of the disease (M ± SD; 95% CI)