A cross-sectional analytical design was used in an applied study covering the period from January to August of 2022. This study was carried out according to the Declaration of Helsinki principles [15]. It included 73 Cuban children attending the immunogenetics service of the National Medical Genetics Center (CNGM according to its Spanish acronym) in Havana because of their recurrent infections. The sample was divided into three groups according to the classification of thymus development, as a function of its total area [9, 16]. Hence, group 1 (G1), normal (thymic area of 1000−1500 mm2); group 2 (G2), slight hypoplasia (thymic area of 800−999 mm2); and group 3 (G3), moderate-severe hypoplasia (thymic area less than or equal to 799 mm2). The latter was formed by subgroup 3a (G3a), moderate hypoplasia (thymic area of 500−799 mm2) and subgroup 3b (G3b), and severe hypoplasia (thymic area equal to or lower than 499 mm2). In the study groups, the sample was distributed as follows G1 = 25, G2 = 19, and G3 = 29 (G3a = 24 and G3b = 5). The patients within each group were selected according to the order of their arrival to the service.

Inclusion criteria were as follows: (1) age ranging from 6 months to 6 years of age and (2) and patients that presented one of the following recurrent infection criteria: (a) two or more severe infections in 1 year (with persistent evidence of inflammation, lack of response to oral antibiotics, and/or the need of intravenous antibiotics or hospitalization); (b) severe infection with an unusual pathogen or infections produced by bacteria, which normally do not produce disorders in children of the same age; (c) six or more infections of the respiratory tract (1 of which could be pneumonia, including severe pneumonia) during 1 year in children of 1–3 years of age; (d) five or more infections of the respiratory tract (1 of which could be pneumonia, including severe pneumonia) during 1 year in children of 3–6 years of age; (e) two mild events of pneumonia confirmed by clinical criteria and/or X-rays in 1 year; (f) completing more than four cycles of antibiotics in 1 year; and (g) and prophylactic therapy with antibiotics to prevent infections.

Exclusion criteria for patients included the following: (1) nutritional value less than percentile 3 and higher than percentile 97; (2) under steroid treatment for up to 45 days before the start of the study; (3) treatment with immunomodulators at the time of the study; (4) personal history of genetic chromosomal diseases and/or congenital malformation, of thymic neoplasia, leukemia, lymphoma, and or chronic hepatic diseases; (5) history of partial or total surgical treatment of the thymus; and (6) and those received mediastinal radiotherapy.

We assessed the thymic area of the selected patients after receiving the informed consent of the parents and/or legal tutors. The area was measured using mediastinal echography, by the same researcher and echography method in all cases, at the “Dr. Ángel Arturo Aballí Arellano” Hospital. For the thymus echography, we used a real-time mobile SAL30A device with a flat surface linear pediatric transducer of 5 mHz. The parasternal line was the cutting plane used, and the area of the longitudinal echography section of both thymus lobules was determined between the upper edge of the second rib and the lower edge of the fourth rib [9, 13]

The thymus area was calculated using the following formula: \(\;Total\;area\;of\;the\;thymus(^2)\;=\left(lengthRL\times widthRL\right)+\left(lengthLL\times widthLL\right)\), where RL refers to the right lobule and LL is the left lobule. We determined severe hypoplasia when the thymus area was < 500 mm2, moderate hypoplasia when the range was of 500–799 mm2, and slight hypoplasia at the range of 800–999 mm2. The normal area was recorded as having values in the range of 1000–1500 mm2, since values above 1500 mm2 were classified as thymic hyperplasia.

Cellular immunophenotyping was made by flow cytometry (eight-color Gallios flow cytometry, Beckman Coulter, France) with peripheral blood obtained through venous puncture, using the anti-coagulant K2-EDTA and the lysis solution VersaLyse (Beckman Coulter, France). According to the manufacturer’s recommendations (Beckman Coulter, France), we used a protocol of red blood cell lysis without washing [17]. To rapidly and accurately identify and count the subpopulations of T cells, we designed a polychromatic tube with six lymphocyte antigens. The following monoclonal antibodies conjugated with fluorochromosomes of MACS Miltenyi Biotec (Germany) were added: 0.5 µL of anti-CD45 PE-Vio770 (Clone 5B1), 2.5 µL of anti-CD3 PE (Clone BW264/56), 0.5 µL of anti-CD4 APC-Vio770 (Clone M-T466), 0.5 µL of anti-CD8 PerCP-Vio700 (Clone BW135/80), 1.0 µL of anti-CD45RA APC (Clone T6D11), and 1.0 µL of anti-CD27 FITC (Clone M-T271). To each volume of the conjugate, we added 100 μL of blood; they were mixed for 3 s and incubated in a dark chamber for 15 min at room temperature. Later, we included 1 mL of the lysis buffer VersaLyse™ (Beckman Coulter, Francia) and incubated this for 10 min under the same conditions as the previous step. Finally, the next step was to immediately acquire the sample through the cytometer. Quality control was carried out accordingly (see Supplementary material). For data acquisition, we used the Kaluza Acquisition v1.0 software through which we obtained a minimum of 50,000 total events. For the analysis and results output, we used Kaluza Analysis v1.5a. The absolute cell counts of the lymphocyte populations were made through a dual platform. A manual selection and sequential window strategy were designed with biparametric graphs (see Supplementary material).

The immunophenotypes of T cells were characterized by taking into account the antigenic marking on the lymphocyte selection window in the biparametric graph of CD45 vs SS (side scatter). We quantified the subpopulations of T lymphocytes CD3+ (CD45+, CD3+), T CD4+ (CD45+, CD3+, CD4+) and T CD8+ (CD45+, CD3+, CD8+). Additionally, we quantified the extended immune phenotypes where we included the following: (a) naïve T cells, T CD4+naïve (CD45+, CD3+, CD4+, CD45RA+, CD27+) and T CD8+naïve (CD45+, CD3+, CD8+, CD45RA+, CD27+), (b) central memory T cells (TCM), TCM CD4+ (CD45+, CD3+, CD4+, CD45RA−, CD27+) and y TCM CD8+ (CD45+, CD3+, CD8+, CD45RA−, CD27+), and (c) effector memory T cells (TEM), TEM CD4+ (CD45+, CD3+, CD4+, CD45RA−, CD27−) and TEM CD8+ (CD45+, CD3+, CD8+, CD45RA−, CD27−), as well as terminally differentiated effector memory T cells known as TEMRA cells (T-effector memory re-expresses CD45RA), TEMRA CD4+ (CD45+, CD3+, CD4+, CD45RA+, CD27−), and TEMRA CD8+ (CD45+, CD3+, CD8+, CD45RA+, CD27−).

The summary measures used here were percentage and proportion for qualitative variables. The comparison of the distribution of the qualitative variables was made through the Pearson χ2 test. The odds ratio (OR) was used as a measure of association and their 95% confidence intervals (CI). The quantitative variables were analyzed through central trend measures of position and dispersion. We evaluated the normal distribution of the continuous quantitative variables using the Shapiro Wilks test from group samples with n < 50 and through the Kolmogorov-Smirnov test for samples with n > 50. The variables with a normal distribution were expressed through their mean and standard deviation, while for those with a different type of distribution, we used the median and interquartile range (IQR). Univariate analyses were made using parametric methods (t-test and ANOVA test) and nonparametric methods (Mann-Whitney U- and Kruskal-Wallis test). The correlation coefficient between the values of the T-cell subpopulations identified by flow cytometry and other variables of interest was analyzed using parametric (Pearson) or nonparametric (Spearman) tests, accordingly. Multivariate analyses of logistic regressions were made to evaluate the influence of the type of delivery, weight at birth, duration of breastfeeding, size, and weight (adjusted by age and sex) on the development of the thymus according to their total area (normal, mild hypoplasia, and moderate-severe hypoplasia). Statistical significance was considered when p < 0.05. For data analysis, we used IBM SPSS Statistics (version 26.0 for Windows, NY, USA). The graphs were obtained through GraphPad Prism (version 9.0 for Windows, CA, USA).