Keywords
COVID-19 - pneumonia - thymus - computed tomography - volumetric measurement
Introduction
On March 11, 2020, the World Health Organization (WHO) declared the new novel coronavirus
disease 2019 (COVİD-19) as an epidemic which was first reported in Wuhan, China, as
a pandemic.[1] It has been reported that the risk of children being infected with the pandemic
agent COVİD-19 is equal to that of adults, but they mostly present as asymptomatic
or with milder symptoms, and infected individuals are generally diagnosed as a result
of screening a family contact. In publications to date, it is stated that pediatric
cases that have contracted COVİD-19 have a better prognosis and shorter duration of
hospitalization than adults, similar to the severe acute respiratory syndrome (SARS)
and Middle East respiratory syndrome (MERS) pandemics which are also members of the
coronavirus family.[2]
[3]
According to the published data, children constitute approximately 1.2 to 5% of COVİD-19
cases and less than 1% of related hospitalizations.[4]
[5]
[6] The COVİD-19 infection has been reported to have a severe course (dyspnea, central
cyanosis, and O2 saturation < 92) in 5.2% of children, and critical (acute respiratory failure syndrome,
encephalopathy, heart failure, coagulopathy, and acute renal failure) in 0.4%.[7] Several reasons have been suggested to explain why the COVİD-19 infection has a
milder course in children compared with adults. The most important explanations are
that angiotensin-converting enzyme 2, which is detected as the binding protein of
COVİD-19, does not show sufficient maturation in children; repeated viral agent exposure
in children can play a role in COVİD-19 immune defense; and the thymus gland, which
plays an important role in the immune response, is active in children.[8]
[9]
The thymus is a lymphoepithelial organ that plays an important role in the maturation
of the immune system and provides an environment for T lymphocyte differentiation
and selection. In addition, B-cells are essential elements that contribute to the
formation of protective immunity against pathogens, and the thymus contains an important
subset of resident CD20 + B-cells.[10] It is known that plasma cells in the perivascular space in the thymus gland provide
long-term protection against viruses by secreting antigen-specific antibodies. Studies
have reported that the thymus gland develops antimicrobial immunity and specific T
lymphocytes after local infection with some microbial agents.[11] In our study, we aimed to investigate the relationship between the thymus gland
volume and severity of pneumonia and blood laboratory findings in COVİD-19 infected
pediatric patients that underwent thoracic computed tomography (CT).
Materials and Methods
This retrospective study was approved by the Institutional Ethics Committee of Kütahya
Health Sciences University (document date and number: July 20, 2020-E.6262). The case
group consisted of patients who were found to be positive for COVİD-19 according to
the reverse-transcription polymerase chain reaction (RT-PCR) test conducted at the
Pediatric Department of Kütahya Health Sciences of University Evliya Çelebi Education
and Research Hospital after March 11, 2020, when the World Health Organization (WHO)
accepted COVİD-19 as a pandemic agent, as well as patients with a history of contact
with infected people and typical radiological findings of COVİD-19 described in the
literature. The typical radiological findings of COVİD-19 pneumonia were accepted
as peripheral-bilateral ground-glass opacities, multifocal-round ground-glass opacities,
inverted halo sign, and other findings of organized pneumonia.[12] In our study, thoracic CT examinations of patients infected with COVİD-19 were evaluated
at the time of the diagnosis. Patients with a negative RT-PCR test and CT findings
defined as atypical for COVİD-19 pneumonia in the thoracic CT examination and those
with CT scans with insufficient diagnostic quality or unclearly defined thymus margins
were excluded from the study. According to the anamnesis information obtained at the
time of admission, the mean duration of symptoms was 2.4 days (1–4 days). The blood
tests of the patients taken at the time of admission to the hospital were included
in the study. All patients included in the study were pandemic pediatric service patients
but no intensive care patients.
To determine the dimensional changes caused by the COVİD-19 infection in the thymus
gland, patients who presented to the Emergency Service of Kütahya Health Sciences
University Evliya Çelebi Education and Research Hospital Pediatric Department for
noninfection reasons (e.g., traffic accident, assault, and fall from a height) and
underwent thoracic CT before the pandemic between March and May 2019 were included
in the control group. Similar to the case group, patients with thoracic CT scans with
insufficient diagnostic quality and unclear thymus margins were excluded from the
control group.
Thoracic CT examinations were performed with a 16-slice multidetector sequential CT
scanner (Aquilion, Toshiba Medical Systems, Otawara, Japan). Images were acquired
using window settings allowing the viewing of the lung parenchyma (window level, −500
to −700 Hounsfield's unit [HU]; window width, 1,200–1,500 HU) and the mediastinum
(window level, 20–40 HU; window width, 350 HU). The thoracic CT examinations anatomically
covered the area from the C1 vertebra to the bottom of the diaphragm. All patient
CT examinations were noncontrast examinations.
The CT examinations of the case and control groups were uploaded from the hospital
image archiving system to the workstation. The thymus (TV)-cardiac volumes (CV) of
the cases and controls ([Figs. 1],[2],[3],[4]) and the volumes of the total lung and pneumonic involvement areas (ground glass
and consolidation areas) in the case group ([Fig. 5]) were calculated using the Vitrea FX version 6.1 (Vital Images Inc., a Canon Group
Company, Minnetonka, Minnesota) software. Optimal volume measurements were undertaken
with manual corrections where necessary. Thymus/cardiac ratios (TCR) were obtained
by dividing TV by CV (TV/CV). In the case group, the severity of pneumonia was evaluated
according to the pneumonia severity index (PSI) obtained by proportioning the ground
glass opacity and consolidation areas to the total lung volume. The mean thymus density
(MTD) was measured in the axial plane where the thymus was the largest, with the region
of interest (ROI) covering the thymus and recorded in HU.
Fig. 1 Segmentation of the thymus gland in the axial (A), coronal (B), and sagittal (C) planes on noncontrast thoracic CT scans. CT, computed tomography.
Fig. 2 3D image of the thymus gland in reformatted images (A and B). 3D, three-dimensional.
Fig. 3 Heart segmentation in the axial (A), coronal (B), and sagittal (C) planes on noncontrast thoracic CT scans. CT, computed tomography
Fig. 4 3D image of the heart in reformatted images. 3D, three-dimensional.
Fig. 5 (A-C) Software-supported automatic measurements of the thymus gland and cardiac volumes
in reformatted images.
The Statistical Package for the Social Sciences (IBM SPSS, version 21.0.0 for Windows)
was used for statistical analyses. The conformance of the numerical dataset to normal
distribution was determined by the Kolmogorov–Smirnov test. All data showed normal
distribution. In the case group, the relationship between low O2 saturation, fever, cough, loss of taste and smell, headache symptoms, and TV, TCR,
and MTD was evaluated by independent-samples t-test. The relationship between TV, TCR, MTD and lymphocyte count and the case and
control groups was evaluated using an independent-samples t-test. A p-value of <0.05 was considered to indicate a statistically significant difference.
In the case group, the relationships between PSI, the TV, TCR, and MTD volumetric
values, blood parameters, namely, lymphocyte, white blood cell (WBC), C-reactive protein
(CRP) were evaluated using Pearson's correlation test. The correlation coefficients
were evaluated as follows: r ≥ 0.91, excellent; 0.90 ≥ r ≥ 0.71, good; 0.70 ≥ r ≥ 0.51, moderate; 0.50 ≥ r ≥ 0.31, weak; and r ≤ 0.3, no correlation. In all analyses, p < 0.05 was considered as statistically significant.
Results
The mean age was 12.4 ± 4.5 (minimum: 0.3, maximum: 17) years for the case group and
11.7 ± 4.2 (minimum: 0.1, maximum: 17) years for the control group, with no statistically
significant difference. There were 19 (58%) boys and 14 (42%) girls in both groups.
TV, MTD, and TCR were not related to gender. While TV and MTD were not associated
with age, TCR decreased with age and showed a moderately significant inverse correlation
with age ([Table 1]).
Table 1
Pearson's correlation analysis of the relationship of TV, TCR, and MTD with age and
gender
|
Age
|
Gender
|
|
TV
|
r: 0.083, p = 0.5
|
r: −0.242 p = 0.05
|
|
TCR
|
r: −0.521, p < 0.001
|
r: 0.041 p = 0.7
|
|
MTD
|
r: −0.198, p = 0.1
|
r: 0.079 p = 0.5
|
Abbreviations: MTD, mean thymus density; TCR, thymus/cardiac ratio; TV, thymus volume.
Only two of the patients had O2 saturation below 95% and they did not have tachypnea. In patients with low O2 saturation, O2 levels returned to normal with nasal O2 and no additional intervention was required. Eight of the patients described decreased
sense of taste and smell. There was cough in 22 patients, fever in 21 patients, and
headache in 7 patients. There were no patients with gastrointestinal symptoms or skin
rash. No significant correlation was found between the presence of symptoms and TV,
TCR, MTD, and lymphocyte counts ([Table 2]). The mean hospital stay of the patients was 3.1 (2–5) days.
Table 2
Independent samples t-test results of the mean TV, TCR, MTD values, lymphocyte counts, and clinical symptoms
in the case group
|
Fever
|
Cough
|
Loss of sense of taste and smell
|
|
+(n:22)
|
−(n:11)
|
p
|
+(n:21)
|
−(n:12)
|
p
|
+(n:8)
|
−(n:25)
|
p
|
|
TV
|
24.71 mL
|
22.21 mL
|
0.592
|
23.97 mL
|
23.71 mL
|
0.956
|
17.97 mL
|
25.76 mL
|
0.123
|
|
TCR
|
0.109
|
0.1
|
0.732
|
0.101
|
0.115
|
0.6
|
0.093
|
0.110
|
0.573
|
|
MTD
|
−20.88
|
−170.29
|
0.058
|
−27.72
|
−145.85
|
0.131
|
−41.80
|
−79.92
|
0.669
|
|
LC
|
2.16
|
2.28
|
0.641
|
2.17
|
2.25
|
0.743
|
2.39
|
2.14
|
0.375
|
|
Low O2 saturation
|
Headache
|
|
+(n:2)
|
−(n:31)
|
p
|
+(n:7)
|
−(n:26)
|
p
|
|
TV
|
19.50 mL
|
24.16 mL
|
0.614
|
17.98 mL
|
25.46 mL
|
0.159
|
|
TCR
|
0.170
|
0.102
|
0.190
|
0.125
|
0.101
|
0.419
|
|
MTD
|
45.60
|
−78.18
|
0.438
|
−2.87
|
−88.94
|
0.355
|
|
LC
|
2.64
|
2.17
|
0.367
|
2.44
|
2.13
|
0.302
|
Abbreviations: LC, lymphocyte count; MTD, mean thymus density; TCR, thymus/cardiac
ratio; TV, thymus volume.
In the independent-samples t-test performed between the case and control groups, the MTD and TCR values were found
to significantly differ. In the COVİD-19 infected group, MTD was lower and TCR was
higher. Although the TV value was higher in the case group, it did not statistically
significantly differ from the control group ([Table 3]).
Table 3
Results of the independent-samples t-test of the mean TV, TCR, and MTD values and the lymphocyte counts of the case and
control groups
|
Case group
|
Control group
|
p-value
|
|
TV
|
23.87 mL
|
19.25 mL
|
0.09
|
|
TCR
|
0.1
|
0.07
|
0.04
|
|
MTD
|
−70.68 HU
|
23.35 HU
|
0.01
|
|
LC
|
2.42
|
2.07
|
0.36
|
Abbreviations: HU, Hounsfield's unit; LC, lymphocyte count; MTD, mean thymus density;
TCR, thymus/cardiac ratio; TV, thymus volume.
In the case group, there was no significant correlation between the presence of pneumonic
involvement on thoracic CT and the TV, TCR, and MTD values. Among the COVİD-19 cases
with pneumonic involvement, the relationships between PSI, TV, TCR, and MTD, and lymphocyte,
WBC, and CRP values were evaluated with Pearson's correlation test and a moderately
significant correlation was observed between the PSI and WBC and CRP values, but no
correlation was determined between the remaining parameters ([Table 4]).
Table 4
Pearson's correlation analysis of the relationship between PSI and volumetric and
laboratory measurements
|
TV
|
r: -0.025, p = 0.8
|
LC
|
r: 0.157, p = 0.3
|
|
TCR
|
r: 0.233, p = 0.1
|
WBC
|
r: 0.419, p = 0.01
|
|
MTD
|
r: 0.183, p = 0.3
|
CRP
|
r: 0.451, p = 0.01
|
Abbreviations: LC, lymphocyte count; MTD, mean thymus density; PSI, pneumonia severity
index; TCR, thymus/cardiac ratio; TV, thymus volume.
Discussion
To the best of our knowledge, our study is the first to evaluate thymus gland sizes
in patients with COVID-19, and it is important both in terms of its contribution to
the literature and quantitative measurement of thymus size in thoracic CT examinations.
In our study, the TCR and TV values in children infected with COVİD-19 were higher
compared with the controls, and the increase in the TCR value was statistically significant.
This finding can be considered as an indirect indicator of the active role of the
thymus in COVID-19 infection. Most studies on the thymus gland, lymphocyte count,
and immune system have been undertaken in HIV-infected cases. In studies conducted
with HIV-infected cases, it was reported that the thymus gland volume was positively
correlated with the CD4 lymphocyte count and negatively correlated with HIV viral
load, and the thymus volume was suggested to guide lymphocyte activity. A dimensional
increase in the thymus gland has been described in the CT measurements in HIV-infected
patients with an increase in the number of CD4+ cells by >100 or more.[13] In the study of Varga et al, it was reported that lymphocytes count in the peripheral
blood and thymus size showed a statistically significant positive correlation in newborns.[14] However, there are different mechanisms that affect the lymphocyte count in the
peripheral blood independent of the thymus size in COVID-19 infection. CD8+ T-cells
release perforin and granzyme A and B which induce apoptosis in virus-infected cells,
chronic inflammatory state continuously stimulates T-cells and causes depletion of
T-cells, and upregulation of NKG2A expression in inflammation leads to functional
depletion of CD8+ and natural killer (NK) cells. These has been blamed for lymphopenia
in COVID-19 infected patients.[15]
[16] It has been reported that increased secretion of proinflammatory cytokines, such
as interleukin (IL)-1, interferon (IFN)-γ, and IL-6, to compensate for the maladministration
of depleted lymphocytes plays a critical role in the induction of lymphopenia.[17] In addition, COVID-19 can directly target lymphocytes and destroy lymphoid organs.
Since blood lactic acid levels are high in patients with severe COVID-19 infection,
it has been reported that lymphopenia may be due to such metabolic causes.[18]
Lymphocytopenia has been reported at a rate of 3 to 3.5% in children infected with
COVİD-19.[19]
[20] Lymphocytopenia is seen especially in elderly patients and severe cases of COVİD-19.[21]
[22]
[23]
[24] In a study by Liu et al, it was shown that the outflow of lymphocytes from the thymus
was induced, which positively contributed to mortality in cases of lymphopenia that
were given thymosin α 1 treatment.[25] In our study, lymphopenia was observed in only one patient (3%), and there was no
significant correlation between the lymphocyte count and the TV and TCR values. There
were no patients with severe COVID-19 infection findings in the patients included
in the study. The lack of correlation between thymus volume and lymphocyte count in
our study may indicate that similar lymphopenia mechanisms described also work in
individuals who have mildly recovered from COVID-19 infection and the active role
of the thymus in patients with lymphopenia. In our study, the amount of lymphocytes
in the COVID-19 infected patient group did not differ significantly with the control
group.
The thymus is quite active in the intrauterine and neonatal period and begins to decrease
partially after birth but continues to be active until puberty.[26] In the study of Varga et al, it was proved that there is a statistically significant
positive correlation between thymus size and body size in newborns.[14] In our study, we included TCR in addition to TV to provide standardization between
body dimensions and thymus volume. When we evaluated the control and patient groups
together, we found that TCR showed a negative correlation with age but studies with
larger patient groups including adults are needed to determine whether this is secondary
to the increase in heart size with increasing age. Our study was conducted on pediatric
patients under the age of 18 years, and no significant relationship was found between
TV and age, probably because the difference between age groups was limited or because
our patient and control groups were small. In our study, the MTD value was found to
be significantly lower in cases compared with the control group which was considered
to be related to gland activation. However, since there is no similar research in
the literature, further studies are needed to support our findings.
In the literature, abnormal thoracic CT findings have been described in 60 to 80%
of children infected with COVİD-19.[2]
[27]
[28] In our study, pneumonic infiltration was detected in the 52% of the cases. To determine
the severity of pneumonia involvement, we compared the volume of ground-glass opacity
and consolidation areas to the total lung parenchyma volume (PSI). While PSI did not
correlate with the TCR and lymphocyte values, it was correlated with CRP and leukocyte
count. In a study conducted by Garcia-Vidal et al with COVİD-19 cases followed-up
in hospital, the presence of a microbiologically confirmed secondary infection agent
was revealed in 7.2% of the patients.[29] In another study, Fattorini et al proved the presence of bacterial superinfection
in 11.7% of the cases.[30] It was considered that the correlation of the PSI value with leukocyte count and
CRP might be an indicator of concomitant bacterial superinfection, especially in severe
patients with lung involvement. However, in our study, there was no patient with a
microbiologically confirmed bacterial superinfection. In addition, broad-spectrum
antibiotics were routinely included in the treatment of all patients with lung infiltration.
Conclusion
Thymus gland is enlarged in COVİD-19 infected children as an indicator of the activation
of the virus. While thymus gland sizes are an indicator of the amount of lymphocytes
in the peripheral blood in healthy individuals and during the treatment of some diseases
(HIV-infected patients), we could not find a similar correlation with the susceptibility
to lymphopenia in COVİD-19 infection. Prospective studies with larger case series
are needed to elucidate the effect of the thymus gland on the milder course of COVİD-19
infection in children and to guide new treatment strategies.
