Computed tomographic evaluation of the thymus—does obesity affect thymic fatty involution in a healthy young adult population?

Abstract

Objective:

To determine a relationship between increased body mass index (BMI) and fatty involution of the thymus in subjects aged between 20 and 30 years.

Methods:

CT images of 94 patients aged between 20 and 30 years were reviewed. Quantitative thymic mean attenuation was recorded and qualitative thymic attenuation was assigned to 1 of 4 possible grades. BMI and subcutaneous fat thickness were documented. Correlations between thymic attenuation, and BMI and subcutaneous fat thickness were assessed using linear regression models. Differences in thymic attenuation in overweight vs normal weight patients were assessed using t-test and Pearson Χ² analysis.

Results:

Low mean thymic attenuation values were associated with higher patient BMI (p = 0.024). Normal weight patients had a mean quantitative thymic attenuation of 15.5 Hounsfield unit and overweight patients had a mean quantitative thymic attenuation of −16.4 Hounsfield unit (p = 0.0218). There was a significant association between increasing subcutaneous fat thickness and reduced mean quantitative thymic attenuation (p < 0.0001). There was also a significant difference in subcutaneous fat thickness when comparing qualitatively assessed thymic Grade 0 with grades 2 and 3 (p = 0.027 and 0.001 respectively); and Grade 1 with Grade 3 (p = 0.001).

Conclusion:

In patients between 20 and 30 years old, the degree of thymic fatty infiltration is related to BMI.

Advances in knowledge:

Multidetector CT can assess fatty involution of the thymus gland. This retrospective study demonstrates a relationship between BMI and thymus gland fatty involution. Subjects with increased subcutaneous fat have decreased mean thymus gland attenuation.

Introduction

The thymus is a primary lymphoid organ of the immune system that promotes the development of T-cells. As the thymus undergoes normal involution with age, the resultant dysregulation of cell-mediated immunity is believed to contribute to the morbidity and mortality associated with progressive ageing, such as increased susceptibility to infections and cancer, and decreased immune response to vaccinations.^{1, 2} As well as fatty involution occurring as part of the normal ageing process, other studies have demonstrated accelerated fatty involution of the thymic gland in disease states such as tumorigenesis, graft vs host disease and viral infections.^{2, 3}

In murine models, obesity has also been shown to induce defects in thymic function, and while the potential role that obesity and its effect on the thymus has to play in human subjects has yet to be defined, the increased risks of infection and some types of cancer in this population are well established.^{4, 5} The impact of obesity on health is varied and ranges from increased mortality in the younger population to debilitating chronic illness or reduced quality of life.^{6, 7} Its effect can have many manifestations across various organs and systems such as evidenced by associations with cardiovascular disease, diabetes, liver disease and obstructive sleep apnea.⁸ The resultant effect on the economy through direct and indirect medical costs is substantial. In the USA, the estimated cost is up to $215 billion per year.⁹

The normal thymus is readily distinguishable on cross-sectional imaging in young patients typically into the third or early fourth decades. Baron et al in the 1980s documented varying morphologies, sizes and densities of the normal thymic gland on CT and 100% of subjects under the age of 30 had identifiable thymus.¹⁰ From the fourth decade onwards, as subject age increases, the thymus typically undergoes fatty involution until thymic tissue is no longer appreciated.^{11, 12} The use of MRI for imaging of the thymus is increasing in popularity, particularly in the setting of more definitive tissue characterization of suspected thymus pathology.^{13, 14}

The potential impact of increasing patient adiposity on thymic appearance has not been defined. Obesity-related fatty infiltration of other solid viscera such as the liver and pancreas are well described phenomena, which are detectable on imaging and which may have clinical sequelae.^15–17 Given the previously demonstrated relationship between obesity and thymic dysfunction, it is possible that a similar radiological phenotype may be detectable in the thymus. The only previous study to report the relationship of thymic morphology to patient weight did not evaluate young patients, and only assessed patients over 34 years old, a group in whom thymic involution is well established and in whom the normal thymus may be difficult to detect.¹⁸ The primary aim of this study was to assess the relationship between body mass index (BMI), subcutaneous fat thickness and thymic morphology on CT in adult patients younger than 30 years old.

Methods and materials

This retrospective study was approved by the institution’s ethics review board.

Subjects and clinical data

All consecutive 20–30-year-old patients who underwent contrast enhanced CT of the chest between August 1 2013 and July 31 2014 were identified using a filtered search function on the local Picture Archiving and Communication System (PACS) (McKesson Solutions, San Francisco, CA). This resulted in 157 chest CT examinations with intravenous contrast. The exclusion criteria included any patient with a diagnosis of cancer, diseases related to the thymus such as myasthenia gravis, and other systemic, infectious or autoimmune inflammatory diseases such as Graves disease, inflammatory bowel disease, cystic fibrosis, human immunodeficiency virus, cirrhosis, hepatitis B and C, and vasculitis. On review of electronic medical records, 46 patients were excluded based on the above exclusion criteria. A further 17 patients were excluded upon subsequent CT analysis and inability to accurately evaluate the thymus due to the presence of streak artefact, pneumomediastinum or thoracic anatomical deformities. Once exclusions had been applied, a total of 94 patients were eligible for CT evaluation of the thymus (Figure 1). A list of indications for CTs performed is provided in Table 1.

Figure 1.

Eligible patients for inclusion in study. PACS, Picture Archiving and Communication System.

Table 1.

Summary of clinical indications for CT chests performed on subjects aged 20–30 years old

Clinical indication for CT chest	20–30 year old, n = 94
Evaluate for pulmonary embolus	48
Trauma	21
Chest X-ray finding	8
Assorted other (e.g. work-up of anaemia, neutropaenia, night sweats, possible malignancy)	17

CT data acquisition

All chest CT scans were obtained either using multidetector CT system Aquilion 64 (Toshiba Medical Systems, Tokyo, Japan) or multidetector CT system Aquilion PRIME (Toshiba Medical Systems, Europe). Examinations were all acquired on inspiration, 35 s following the administration of 60 ml of contrast at 3 ml s⁻¹ as per CT chest protocol or following bolus tracking at 5 ml s⁻¹ as per CT pulmonary angiogram (CTPA) protocol using isohexol (Omnipaque 350; GE Healthcare, Princeton, NJ). Imaging parameters included kVp of 120 with automatic dose modulation resulting in mAs varying from 80 to 250 mAs, pitch of 1, with slice thickness of 5 mm.

Assessment of thymic attenuation

Image analysis was performed on the institutional Picture Archiving and Communication System by a radiologist (KH) with 4 years’ experience, who evaluated the thymus on axial soft tissue windows. The average attenuation of the thymus was assessed and assigned a grade on a scale of 0 to 3 (Figure 2).^{18, 19} Grade 0 was assigned, when there was complete fatty replacement of the thymus and it was no longer distinguishable from the mediastinal fat; Grade 1 represented a predominately fatty thymus and had attenuation values or Hounsfield unit (HU) of less than 10; Grade 2 represented a half fatty, half soft-tissue attenuation thymus (HU 10–30) and Grade 3 represented a thymus with attenuation values approaching that of muscle (HU >30).^{18, 19} To quantitatively measure thymic attenuation a circular region of interest (ROI) was placed and attenuation values were obtained in the axial plane in three locations, at the superior, middle and inferior aspect of the thymus gland, and an average value in HU was obtained. The ROI was placed to encompass the width of the thymus gland at that level with care taken not to include the adjacent mediastinal fat at the gland margin (Figure 3). The average ROI surface area was 0.50 cm² with a range of 0.16–0.66 cm². Note was made of regions which may skew obtained attenuation values such as streak artefact and every effort was made to avoid inclusion of these areas in the ROI.

Figure 2.

Examples of the different grades of thymic density on CT chests (arrows). (a) An example of Grade 0, where the thymus is no longer distinguishable from background fat. (b) Is an example of Grade 1, a predominately fatty thymus. (c) Is an example of Grade 2. The thymus is half fatty, half soft-tissue in attenuation. (d) is an example of Grade 3, demonstrating a thymus with attenuation values approaching that of muscle.

Figure 3.

Example of how a ROI circle was placed on the thymus (dashed circle). The ROI was placed to encompass the width of the thymus gland with care taken not to include adjacent non-thymic structures. (b) An example of Grade 1, predominantly fatty thymus, with interdigitating fat within the thymus included in the ROI. ROI, region of interest.

Once thymic tissue was identified the predominant shape of the gland was characterized as either “triangular” or “quadrilateral”. In addition, the contour of the gland in the axial plane was recorded as either “concave”, “straight” or “convex”.¹⁹ If the thymus varied in shape or contour the dominant lobe was recorded.

Patient characteristics

Patient characteristics recorded in the electronic medical record at the time of CT imaging were documented. BMI was recorded on 47 patients and was either obtained on the same day as the CT or self-reported by the patient on a later date. Self-reported height and weight has been shown to have a high degree of correlation with measured height and weight and calculated BMI.²⁰ In all patients the thickness of the anterior chest wall subcutaneous fat was measured in the midline at the level of the xiphisternal junction as an additional indicator of patient body composition.²¹ This site was chosen for measurement as it was easily identifiable and as all subjects were imaged supine, without constraints about their chest, and was the least likely region to be affected by subject positioning or clothing.

Statistical analysis

Statistical analyses were performed using Stata 12.1 (Stata Statistical Software: Release 12. StataCorp LP: College Station, TX). The difference in age between patients was assessed using the student t-test. Correlations for quantitative data such as mean thymic attenuation against subject BMI and subcutaneous thickness were assessed using linear regression models. Regression analysis was also used to evaluate a potential relationship between BMI and subcutaneous thickness of the chest wall. Differences between subcutaneous fat thickness between various thymic grades was assessed using a one-way analysis of variance test for each of the variables. Differences in mean thymic attenuation and thymic grade in overweight vs normal weight patients, taking BMI > 25 as cut-off for normal weight, were assessed using t-test and Pearson Χ² analysis. p-values were regarded as statistically significant if less than 0.05.

Results

Mean age was 25.0 years ± 2.9 (standard deviation). The mean age was significantly different for the male and female groups with a mean female age of 25.6 ± 2.8 and a mean male age of 24.3 ± 3.0 (p = 0.03).

Of the 94 in whom evaluation of the thymus was performed on CT, a BMI was recorded on 47 patients, 27 females and 20 males. The mean BMI was 26.0 ± 6.1 (standard deviation) with females having an average BMI of 26.5 ± 7.3 and males of 25.3 ± 4.0. There was no significant difference in mean BMI between male and female groups (p = 0.52). Mean thoracic wall subcutaneous fat thickness was 11.7 mm ± 9.8. The mean subcutaneous thickness in females was significantly higher compared to males at 15.9 mm ± 9.9 vs 7.3 mm ±7.5 respectively (p < 0.001).

Results of qualitative thymic grading demonstrated that 16 out of 94 (16.7%) had complete fatty replacement of the thymus (Grade 0), 27.1% had a predominantly fat attenuation thymus (Grade 1), 21.9% had half fatty, half soft tissue attenuation (Grade 2) and 32.3% had soft tissue attenuation (Grade 3). Associations between thymic grade and patient characteristics are shown in the Table 2.

Table 2.

Summary of patient characteristics and associations with thymus gland grade

		Thymic grade				p-valuec
Grade 0N = 16 (17%)a	Grade 1N = 26 (28%)	Grade 2N = 21 (22%)	Grade 3N = 31 (33%)
Age of patients (years)	Mean	26.5 ± 2.5b	25.7 ± 2.7	24.3 ± 2.6	24.2 ± 3.2	0.640
Thymic morphology – number (%)	Triangular	–	15 (57.7)	11 (52.38)	17 (44.8)	0.997
Quadrilateral	–	11 (32.3)	10 (47.6)	14 (45.2)
Not identifiable	16 (100)	0	0	0
Thymic contour – number (%)	Straight	–	17 (65.4)	10 (47.6)	16 (51.6)	0.890
Convex	–	3 (11.5)	7 (33.3)	9 (29.0)
Concave	–	6 (23.1)	4 (19.1)	6 (19.4)
Not identifiable	16 (100)

^aPercentage values are rounded.

^bPlus-minus values are mean ± standard deviation

^cp-value calculation was calculated using one-way analysis of variance test. p < 0.05 as defined as significant.

BMI and mean thymus attenuation/thymic grade

There was a statistically significant relationship between mean thymic attenuation and subject BMI (p-value = 0.024). Increasing BMI was associated with lower thymic attenuation values (Figure 4).

Figure 4.

Mean thymic attenuation plotted as a function of BMI (p-value = 0.024).

When the subjects were divided into overweight or normal weight groups as indicated by a BMI of ≥25, there was a significant difference in the mean thymic attenuation between the two Groups. 26 subjects had a BMI <25 and 21 had a BMI ≥25, with normal weight patients having a mean thymic attenuation of 15.5 HU (CI −3.4 to 34.3) and overweight subjects having a mean thymic attenuation of −16.4 HU (CI −36.6 to 3.8), p = 0.0218.

There appeared to be a trend between thymic grade and being overweight. 19 out of 27 (70%) of patients with thymic Grade 2 or 3 had BMIs <25, while 13 out of 20 (65%) patients with thymic Grade 0 or 1 had BMIs ≥25. However, this apparent trend did not reach statistical significance, p = 0.068.

Subcutaneous fat thickness and mean thymic attenuation/thymic grade

Using linear regression analysis and controlling for age and sex, there was a statistically significant relationship between subcutaneous fat thickness and mean thymic attenuation, with increasing chest wall subcutaneous fat, p-value < 0.0001 (Figure 5). A linear relationship was demonstrated when comparing different grades of thymic attenuation against subcutaneous fat thickness at the level of the xiphisternum, with a significant difference in subcutaneous fat thickness when comparing Grade 0 with Grade 2 and Grade 3 (p = 0.027 and 0.001 respectively) and Grade 1 with Grade 3 (p = 0.001).

Figure 5.

Mean thymic attenuation plotted as a function of subcutaneous fat thickness at the xiphisternum. For every 1 mm increase in subcutaneous fat, there was a 0.89 increase in BMI, p-value ≤ 0.001 (CI 0.6 to 1.18). BMI, body mass index; CI, confidence interval.

A statistically significant relationship was demonstrated between subject BMI and the thickness of subcutaneous fat of the anterior chest wall, measured at the level of the xiphisternal junction. For every 1 mm increase in subcutaneous fat, there was a 1.06 unit increase in BMI, p-value ≤ 0.001 (CI 0.8 to 1.4).

Thymic morphology

Triangular morphology of the thymus was slightly more common in both males and females. When the organ was distinctly visible, 19 out of 35 (54.2%) male thymuses and 24 out of 43 (56%) female thymuses exhibited a triangular shape. The most common thymic margin contour was straight (43, 54%) with the next most common contour being convex (16, 20%). There was no significant difference between thymic shape or contour and patients with a high BMI and those with normal BMI (p = 0.379 and p = 0.304 respectively).

Discussion

We assessed the relationship between BMI, subcutaneous fat thickness and thymic density in a cohort of patients between 20 and 30 years old undergoing chest CT. Overweight or obese patients were more likely to have lower attenuation of the thymus and a higher qualitatively assessed thymic grade when compared to their normal weight or underweight counterparts, suggesting that obesity places patients at risk for thymic fatty infiltration. To our knowledge this is the first study to demonstrate an association between thymic fatty infiltration and increasing adiposity in young adults.

The prevalence of obesity has increased dramatically since the 1980s, with recent forecasts predicting a 33% increase in prevalence in the USA over the next two decades.^{22, 23} BMI is frequently used as an indicator of body fat content and subjects with a BMI ≥ 25–29 are considered overweight and a BMI ≥ 30 are considered obese.²⁴ Obesity-related fatty infiltration of other organ systems is well-described and can have significant clinical effects. For example, obesity is a risk factor in the development of non-alcohol related fatty liver disease, which in some cases can progress to cirrhosis. Such changes can be readily appreciated across multiple imaging modalities.^{25, 26}

The thymus gland is normally identifiable in young patients, with Baron et al documenting thymic tissue in 100% of patients under the age of 30.¹⁰ Notably, Baron et al’s paper was published in the 1980s’, a time when obesity was less prevalent.²² More recently Ackman et al described differences in thymic tissue attenuation between the sexes in the 20–30 age group. In 6% of patients, no thymic tissue was identified and while the authors demonstrated increased fatty involution of the thymus in young males compared to females, the relationship of thymic morphology to BMI was not explored.¹⁹ In 2016, in a group of older patients with a mean age of 59 (range 34–90), Araki et al demonstrated that patients with qualitatively assessed low thymic grade had significantly higher BMIs, than their counterparts with a high thymic grade.¹⁸ The significance of this finding in a group of older patients, in whom established thymic involution would be expected, is uncertain.

The thymus is a primary lymphoid organ of the immune system and is necessary for the production of T-cells, which play a key role in cell-mediated immunity. The thymus undergoes involution as part of the natural aging process, with loss of thymic epithelial cells and increased adipocytes within the perivascular spaces seen, resulting in reduced thymic output and function.^{27, 28} As well as fatty involution associated with normal aging, studies have demonstrated that certain disease states can also accelerate fatty involution of the thymus, such as infection, tumorigenesis and chronic inflammatory states.^{2, 3}

In addition to conveying an increased risk of certain cancers, Type 2 diabetes and cardiovascular disease, obesity can also be thought of as an immunodeficient state with studies demonstrating increased susceptibility to, and increased morbidity and mortality from acquired infections.^{4, 29} In mouse models, obesity and calorie excess have been shown to induce accelerated thymic fatty involution and conversely, calorie restrictive diets have demonstrated inhibition of thymic involution with preservation and even restoration of function.³⁰ One recent retrospective study associated a high BMI with reduced T-cell production in a large cohort of elderly patients, however, the role that obesity plays in fatty involution of the thymus in humans remains to be fully defined. Our study demonstrated varying degrees of thymic density in the young population we studied, however, a lower attenuating fatty thymic appearance was more likely to be seen in subjects with a higher BMI and increased subcutaneous fat. These findings suggest a potential role for obesity in altering thymic morphology and function. We did not find and correlation between thymic shape or contour and patient BMI.

This study has several limitations. Given the retrospective nature of our study, inherent limitations in terms of data collection exist. Imaging was performed on two different CT scanners and, although all studies were contrast enhanced, some were bolus-tracked and the rest were injected following a pre-determined amount of time as per CT thorax protocol, which meant there were some small differences in the timing contrast was administered. Participant BMI was recorded objectively or, in instances where objective measurements were not recorded, self-reported weight and height was provided upon contacting the patient. Although not ideal, self-reported height and weight has been shown to have high correlation with measured height and weight values.²⁰

Conclusion

In patients between 20 and 30 years old the degree of fatty infiltration of the thymus is related to BMI, and overweight patients are more likely to have higher levels of thymic fatty infiltration than normal or underweight patients. Our results suggest that further work is needed to fully define the possible role that obesity play in thymic dysfunction.