- Email: [email protected]
Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?
Clinical Radiology xxx (2016) e1ee6
Contents lists available at ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism? P.D. Maldjian*, T. Chen Department of Radiology, University Hospital, Rutgers New Jersey Medical School, Newark, NJ, USA
art icl e i nformat ion Article history: Received 18 May 2016 Received in revised form 22 June 2016 Accepted 8 July 2016
AIM: To determine if visual assessment of the attenuation of morphologically normal appearing thyroid glands on unenhanced computed tomography (CT) of the chest is useful for identifying patients with decreased thyroid function. MATERIALS AND METHODS: This was a retrospective study of 765 patients who underwent both unenhanced CT of the chest and thyroid function tests performed within 1 year of the CT examination. Attenuation of the thyroid gland was visually assessed in each patient relative to the attenuation of the surrounding muscles to categorise the gland as “low attenuation” (attenuation similar to surrounding muscles) or “high attenuation” (attenuation greater than surrounding muscles). Thyroid attenuation was quantitatively measured in each case to determine the validity of the visual assessment. Results of thyroid function tests were used to classify thyroid function as hypothyroid, euthyroid, or hyperthyroid. Data were analysed to determine the relationship between visual assessment of thyroid attenuation and status of thyroid function. RESULTS: Thyroid glands of low attenuation were present in 4.2% (32/765) of the patients. Nearly half (47%) of the patients with low-attenuation thyroids had hypofunctioning thyroid glands. Compared to patients with high-attenuation thyroids, patients with low-attenuation thyroids were significantly more likely to have decreased thyroid function (clinical and subclinical hypothyroidism) and significantly less likely to be euthyroid (p<0.0001). Quantitative measurement of thyroid attenuation confirmed the validity of the visual assessment. CONCLUSION: Low attenuation of an otherwise normal-appearing thyroid gland on unenhanced CT of the chest is strongly associated with decreased thyroid function. Ó 2016 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction The thyroid gland is known to have somewhat higher attenuation than the soft tissues on unenhanced computed * Guarantor and correspondent: P.D. Maldjian, Department of Radiology, Rutgers New Jersey Medical School, University Hospital, 150 Bergen St, UH C-320, Newark, NJ 07103-2406, USA. Tel.: þ1 (973) 972 5188; fax: þ1 (973) 972 7429. E-mail address: [email protected] (P.D. Maldjian).
tomography (CT) due to its increased iodine content. Hypothyroidism has been shown to be associated with decreased iodine content within the thyroid gland. Previous investigations have demonstrated some correlation between CT attenuation of the thyroid gland on unenhanced CT and iodine content in patients with known thyroid disease.1e4 Anecdotally, the present authors have encountered instances on unenhanced CT of the chest in hypothyroid patients where otherwise normal-appearing thyroid glands
http://dx.doi.org/10.1016/j.crad.2016.07.008 0009-9260/Ó 2016 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
e2
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
demonstrated attenuation similar to that of the surrounding muscles. The purpose of this investigation was to determine if visual assessment of the relative attenuation of the thyroid gland compared to the surrounding muscles is useful in suggesting hypothyroidism in patients with otherwise normal-appearing thyroid glands on unenhanced CT.
Materials and methods This was a retrospective evaluation approved by the Institutional Review Board. Due to the retrospective nature of the study, the Institutional Review Board waived the requirement for informed consent. The radiology data base and electronic medical record system were searched to identify all adult patients (patients >18 years of age) who had both unenhanced chest CT performed between January 2012 and June 2015 and thyroid function tests performed within 1 year of the CT examination (986 patients). The CT images in these patients were reviewed to evaluate the appearance of the thyroid gland. If patients had more than one CT examination and/or more than one set of thyroid function tests, the CT examination and thyroid function results that were closest together temporally were used for analysis. Exclusion criteria were inability to visualise the thyroid gland on the CT images (prior thyroidectomy or exclusion of the thyroid gland from the scan range), gross abnormality of the thyroid (thyroid enlargement or presence of any thyroid nodules >1 cm in diameter) and any evidence of prior thyroid surgery (partial thyroidectomy or visible surgical clips in the thyroid bed). After application of these exclusion criteria, 765 patients were identified that comprised the study population. The CT images of the thyroid were visually assessed to categorise the appearance of the thyroid as either “low attenuation” or “high attenuation”. A thyroid gland of “low attenuation” was defined as attenuation at the same level or less than the adjacent muscles, such that it was visually difficult to discern the thyroid gland from the surrounding muscles based on its attenuation alone. “High attenuation” of the thyroid was defined as attenuation visibly greater than the surrounding muscles, such that the thyroid was easily discernible from the adjacent muscles. No distinction was made between attenuation of the thyroid slightly greater or significantly greater than the surrounding muscles. As long as thyroid attenuation was visibly greater than surrounding musculature, such that the thyroid was easily discernible, a “high-attenuation” score was recorded for the case (Fig 1). After visual assessment of thyroid attenuation in each case, the attenuation of the thyroid gland was directly measured. A circular region of interest (ROI) approximately 1 cm in diameter was placed over a homogeneous region of one lobe of the thyroid on the axial images and the attenuation in Hounsfield units (HU) was recorded. All CT examinations were conventional unenhanced chest CT examinations performed for standard clinical indications (i.e., cough, lung nodule assessment, haemoptysis, etc.) CT examinations were performed on a 16-section multidetector CT machine (Lightspeed 16, GE Healthcare Milwaukee, Wisconsin, USA) at 120 kVp with automatic tube current
modulation, maximum of 300 mA, noise index of 23, 0.8second rotation time, 16 1.25 mm collimation and pitch of 1.375. Images were reconstructed at 1.25 mm thickness at 1.25 mm intervals using a standard soft-tissue algorithm. Images were viewed on a conventional picture archiving and communication system (PACS) viewing station at routine soft-tissue windows (window width¼400 HU, window level¼40 HU). Studies were interpreted by a board certified radiologist with >20 years of experience in thoracic imaging. The reader was blinded to the results of the thyroid function tests at the time of image interpretation. The serum levels of thyroid stimulating hormone (TSH) for the study patients were assessed to classify patient thyroid function as hypothyroid, euthyroid, or hyperthyroid. As primary thyroid disease (disease of the thyroid itself) is much more common than secondary thyroid disease (pituitary disease affecting TSH production), it was assumed that abnormal TSH values reflected primary thyroid disease.5 Serum TSH was determined by immunmoradiometric assay and the normal reference range for TSH at the authors’ facility is currently 0.45e4.50 mIU/ml. Normal reference ranges have varied slightly over the last few years, so the reference range concurrent with the date of the TSH results was used to classify thyroid function. Patients with serum TSH values within the reference range were classified as euthyroid, patients with TSH values above the reference range were classified as hypothyroid, and patients with TSH values below the reference range were classified as hyperthyroid. For patients with abnormal serum TSH, free thyroxine values (if available) were also assessed to differentiate subclinical thyroid dysfunction from more severe disease. Free thyroxine levels were determined by radioimmunoassay with a normal reference range currently assumed to be 0.82e1.77 ng/dl. As with TSH, reference ranges for free thyroxine have varied over the last few years, so the reference range concurrent with the date of the free thyroxine results was used to classify thyroid function. The data were analysed to determine if there was any relationship between grading of thyroid attenuation by visual assessment and the status of thyroid function. Results from direct measurement of thyroid attenuation were assessed to confirm the validity of subjective assessment of thyroid attenuation. For statistical analysis, differences in proportions were compared using the chi-square test with the Yates continuity correction or the Fisher exact test. Differences in mean values were compared using the Student ttest. For all statistical tests, a two-tailed p-value of <0.05 was considered statistically significant.
Results The present results (Table 1) show that a low-attenuation thyroid is an unusual finding present in only 4.2% (32/765) of the present sample. Of these 32 patients, 12 were male (38%) and 20 (62%) were female. The mean age for these patients was 59.1 years (standard deviation [SD]¼11.9 years, range¼40e96). Seven hundred and thirty-three patients (96%) had high-attenuation thyroids. Of these, 347 were
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
e3
Figure 1 Examples of low-attenuation and high-attenuation thyroid glands on unenhanced CT. (a) A 74-year-old woman with a low-attenuation thyroid gland. Note that the thyroid gland (arrows) is visibly similar in attenuation to the surrounding muscles. The patient was hypothyroid with elevated TSH and low free thyroxine levels. (b) A 65-year-old euthyroid man with a high-attenuation thyroid gland. The thyroid (arrows) is slightly increased in attenuation compared to the surrounding muscles. (c) A 64-year-old euthyroid man with a high-attenuation thyroid gland. The thyroid (arrows) is markedly increased in attenuation compared to the surrounding muscles.
male (47%) and 386 were female (53%). The mean age was 58.8 years (SD¼13.9 years, range¼20e95 years). Direct measurement of thyroid attenuation confirmed that visual assessment could differentiate low-attenuation from highattenuation thyroid glands. The mean difference in
Table 1 Visual assessment of thyroid attenuation and its relationship to thyroid function.
Total number of patients Mean thyroid attenuation SD (HU) [Range (HU)] Number of hyperthyroid patients (%) Number of euthyroid patients (%) [95% confidence interval for proportion] Number of hypothyroid patients (%) [95% confidence interval for proportion] Number of hypothyroid patients with low free thyroxine (%) [95% confidence interval for proportion]
No. of patients with low-attenuation thyroid glands on CT
No. of patients with p-Value high-attenuation thyroid glands on CT
32
733
53.413.7
94.716
<0.0001
[20.7e76.6] 1 (3.1)
[55.7e145.2] 23 (3.1)
1.0
16 (50.0)
638 (87)
<0.0001
[0.34e0.66]
[0.84e0.89]
15 (46.9)
72 (9.8)
[0.3e0.64]
[0.08e0.12]
5 (33.3)
3 (5.4)a
[0.15e0.58]
[0.02e0.15]
SD, standard deviation. a Only 56 of the 72 patients had free thyroxine levels tested.
<0.0001
0.0086
attenuation between thyroid glands visually classified as low attenuation and those visually classified as high attenuation was 41.3 HU (95% confidence interval [CI]: 35.1e47.5 HU, p<0.0001). Nearly half (47%) of the patients with low-attenuation thyroids had hypofunctioning thyroid glands. Compared to patients with high-attenuation thyroids, patients with lowattenuation thyroids were significantly more likely to be hypothyroid and significantly less likely to be euthyroid (p<0.0001). All 15 of the hypothyroid patients with low attenuation thyroid glands had free thyroxine levels tested. In 10 of these patients, the free thyroxine levels were normal, indicating subclinical hypothyroidism (high TSH with normal free thyroxine), whereas in five patients the free thyroxine was low, consistent with clinical hypothyroidism. Of the 72 hypothyroid patients with highattenuation thyroids, 56 had free thyroxine levels tested. Only three of these patients had low free thyroxine (consistent with clinical hypothyroidism), whereas the remaining 53 patients had normal free thyroxine levels (consistent with subclinical hypothyroidism). Thus, patients with low-attenuation thyroid glands were significantly more likely to have laboratory findings consistent with clinical hypothyroidism compared to patients with highattenuation thyroid glands (p¼0.0086). Within the present total sample of 765 patients, 11.4% (87/765) were hypothyroid, 85.5% (654/765) were euthyroid, and only 3.1% (24/765) were hyperthyroid. There was no difference in the frequency of hyperthyroidism between low-attenuation thyroid patients and high-attenuation thyroid patients. Only one patient with a low-attenuation thyroid was hyperthyroid. Review of this patient’s records showed that the patient was hyperthyroid due to a hyperfunctioning thyroid nodule in the right lobe producing suppression of the remainder of the gland as demonstrated on I-123 thyroid scintigraphy (Fig 2). The free thyroxine level was normal in this patient (consistent with subclinical hyperthyroidism). Of the patients with high-attenuation
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
e4
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
Figure 2 A 57-year-old hyperthyroid woman with a low-attenuation thyroid gland; the only instance of a low-attenuation thyroid in a hyperthyroid patient in the present study. (a) Unenhanced CT shows a low-attenuation thyroid gland (arrows). (b) Right anterior oblique view from I-123 thyroid scintigraphy shows a hyperfunctioning nodule in inferior right lobe of the thyroid gland (arrow) with suppression of uptake in the remainder of the thyroid.
thyroids, 23 were hyperthyroid, but only one had elevated levels of free thyroxine. Thus laboratory findings were consistent with subclinical hyperthyroidism in almost all of the hyperthyroid patients. One patient who was initially euthyroid with a highattenuation thyroid on CT developed hypothyroidism with a low-attenuation thyroid on a follow-up study. Review of this patient’s medical records showed that she presented with small cell lung carcinoma and developed hypothyroidism after radiation treatment and chemotherapy (Fig 3). Of the five patients with clinical hypothyroidism and low-attenuation thyroid glands, two patients underwent chest CT before thyroid function tests, while three underwent CT after the thyroid function tests. Only one of these patients was receiving treatment with thyroid hormone at the time of the CT examination. Three of the remaining four patients were prescribed levothyroxine at some time after CT was performed. Of the 10 patients with subclinical
hypothyroidism and low-attenuation thyroid glands, six underwent chest CT before the thyroid function tests, whereas four underwent CT after the thyroid function tests. None of these patients was receiving treatment with thyroid hormone at the time their CT examination was performed, although seven of these 10 patients were prescribed levothyroxine at some time after their CT examinations. As chest CT was performed before thyroid function tests in eight of these 15 hypothyroid patients, there was potential to suspect hypothyroidism in these eight patients based on the CT appearance of the thyroid before the thyroid function test results were known.
Discussion The thyroid gland contains 25% of the iodine in the body stored within the follicular cells as thyroglobulin. Hence the thyroid is normally high in attenuation on unenhanced CT.
Figure 3 A 65-year-old woman with small cell lung cancer whose thyroid gland changed in attenuation on subsequent CT. (a) Unenhanced CT image before therapy shows a high-attenuation thyroid gland (arrows). The patient was euthyroid at this time. (b) Eighteen months later after radiation treatment and chemotherapy, the patient developed clinically significant hypothyroidism. Unenhanced CT at this time shows that the thyroid (arrows) has become low in attenuation. (Note the small calcification in the posterior left lobe of the thyroid indicating that this image is at the same level as Fig. 3a.) Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
Low attenuation of the thyroid on CT, however, has only been sporadically addressed in the literature. In 1979, Kaneko et al.1 showed that the average attenuation of the thyroid gland on CT in a small series of five hypothyroid patients was significantly lower than that of 44 normal patients. In 1983, Iida et al.2 studied thyroid attenuation in 67 untreated patients with various thyroid disorders (Grave’s disease, Hashimoto’s thyroiditis, simple goitre, thyroid adenoma, thyroid cyst, and thyroid cancer) compared to 24 normal patients. The study reported that the mean attenuation of diseased thyroid glands was significantly lower than that of normal thyroid glands. In that investigation, tissue iodine concentrations in 14 surgical specimens were directly measured and compared to the CT attenuation values of the corresponding thyroid regions in the preoperative studies. A strong correlation was demonstrated suggesting that the CT attenuation value of the thyroid is primarily determined by the iodine content of the tissue.2 In 1994, Kamijo3 evaluated CT attenuation of the thyroid in 155 patients with chronic thyroiditis compared to 95 normal control patients. The investigation showed that CT attenuation of diseased thyroid glands was significantly lower than normal thyroid glands and also showed that the serum TSH concentration was highest (indicating hypothyroidism) in the group of patients with lowest thyroid attenuation.3 In 2000, Imanishi et al.4 correlated results of CT attenuation of the thyroid in 36 patients with thyroid disease with the histology of resected specimens. The investigators found that reduced attenuation of the thyroid in patients with thyroid disease was secondary to decreased iodine concentration in the thyroid follicles, decrease in colloid content, and increase in follicular cells and interstitial structures.4 In 2010, Han et al.6 reported the results of a study on thyroid CT attenuation and thyroid function tests in 56 patients with diffuse hypermetabolic thyroid activity on combined positron-emission tomography (PET)/CT compared to 56 control patients. Patients with hypermetabolic thyroid activity on PET/CT demonstrated significantly decreased CT attenuation compared to controls. The investigators also assessed thyroid attenuation on CT using a subjective visual scale and found that 56% of patients with elevated TSH (13/23, all with hypermetabolic thyroid activity on PET/CT) demonstrated low attenuation. The authors concluded that decreased thyroid attenuation on CT in the study patients implied reduction in colloid volume and replacement with hypermetabolic tissue such as follicular cells, inflammatory cells (lymphocytes), or other stromal tissue.6 In a more recent investigation, Pandey et al.7 evaluated the relationship between serum TSH and thyroid attenuation on unenhanced CT of the cervical spine. The investigators found that patients with low TSH levels and patients with high TSH levels demonstrated decreased thyroid attenuation on CT.7 The present study differs from these prior investigations in that the present study was undertaken to determine if visual assessment of the attenuation of an otherwise normal-appearing thyroid gland in a population not known
e5
to have thyroid disease was useful as an indicator of thyroid dysfunction. Direct quantitative measurement of thyroid attenuation in the present investigation confirms that rapid visual assessment of thyroid attenuation on CT images is sufficient to identify patients with low-attenuation thyroid glands with increased prevalence of thyroid dysfunction. The present findings indicate that low attenuation of the thyroid on CT in such patients is strongly associated with hypothyroidism. Sixteen percent of patients with low-attenuation thyroid glands in the present study (5/32) had laboratory findings consistent with clinical hypothyroidism. Hypothyroidism is the most common hormone deficiency, but due to the wide variation in severity and the non-specific nature of the symptoms, it can often go undiagnosed. Common clinical manifestations of hypothyroidism include fatigue, weight gain, muscle weakness, cognitive impairment, and depression. If untreated, life-threatening complications, such as heart failure, can occur. Clinical hypothyroidism is easily treated with levothyroxine sodium (L-thyroxine), an orally administered synthetic thyroid hormone. Treatment has been shown to alleviate the symptoms and to improve the quality of life in overtly hypothyroid patients.5 Thirty-one percent of patients with low-attenuation thyroid glands in the present study (10/32) had laboratory findings consistent with subclinical hypothyroidism. Subclinical hypothyroidism is a laboratory diagnosis characterised by an elevated TSH level with a normal free thyroxine level. The prevalence of subclinical hypothyroidism is estimated to be 3.1e8.5% in the general adult population. It is usually caused by chronic autoimmune thyroiditis (Hashimoto’s thyroiditis). It is often asymptomatic, but clinical manifestations can be similar to overt hypothyroidism. Prospective studies of patients with subclinical hypothyroidism have reported increased risk of coronary heart disease events, heart failure, and cardiovascular mortality with increased odds of myocardial infarction to as much as 3.1 times as high as in euthyroid controls.8 Subclinical hypothyroidism is also associated with an increased risk of progression to overt clinical hypothyroidism with an average yearly progression rate of 2e6%. To reduce the rate of these complications, although current data are inconclusive, there is a consensus to treat adults with subclinical hypothyroidism if TSH levels are >10 mIU/ml and they are symptomatic or test positive for anti-thyroid antibodies.8,9 As illustrated in Figs 2 and 3, two especially noteworthy cases were encountered during the course of the present investigation. Only one out of 24 hyperthyroid patients in this study had a low-attenuation thyroid. This patient was hyperthyroid secondary to a hyperfunctioning nodule that produced suppression of iodine metabolism and presumably reduction of iodine content in the remainder of the thyroid (Fig 2); further supporting the hypothesis that iodine concentration is the key contributor to the attenuation of the thyroid gland on unenhanced CT.4 In a second patient, the thyroid gland decreased in attenuation when she progressed from a euthyroid to a hypothyroid state (Fig 3). This suggests that visibly apparent reduction of
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
e6
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
thyroid attenuation in a patient over time might be an indication of impaired thyroid function. Further investigation is necessary to determine the validity of that conjecture. The present study has several limitations. There was no histological analysis of tissue to confirm the cause of low attenuation of the thyroid on CT. Given the results of prior investigations, decreased iodine concentration is the most likely explanation. There may be selection bias in the present study as a retrospective analysis was performed on patients who had both thyroid function tests and CT examinations of the chest. As thyroid function tests would be obtained in patients clinically suspected to have underlying thyroid disease, patients with thyroid disease might be overrepresented in the present sample compared to the general population. It has been estimated that the prevalence of hypothyroidism in the United States population is 4.6% (0.3% clinical and 4.3% subclinical).10 In the present sample, 11.4% (87/765) of the patients were hypothyroid suggesting that thyroid disease may be overrepresented in the study patients compared to the general population; however, this bias would only increase the overall prevalence of thyroid disease in the sample and would not affect the conclusions. The use of specific medications was not assessed in the present patients, which can be a confounding variable. For example, amiodarone is a widely used anti-arrhythmic drug with high iodine content and structural resemblance to thyroid hormones. Use of amiodarone can result in elevation of TSH levels, hypothyroidism, or thyrotoxicosis.11 Concerning use of thyroid hormone, however, as only one of the 15 patients with low-attenuation thyroids and hypothyroidism was receiving treatment with thyroid hormone at the time of chest CT, it is not thought that medications for hypothyroidism affected the present results. The present study was limited to conventional unenhanced CT examinations of the chest using the technical parameters specified. Other types of unenhanced CT examinations that include the thyroid, such as lung cancer screening, employ different technical parameters. Further investigation would be necessary to determine if thyroid attenuation is related to thyroid dysfunction on other types of CT studies in a fashion similar to the present findings. In conclusion, the present study shows that low attenuation of an otherwise normal-appearing thyroid gland on
unenhanced CT of the chest is strongly associated with decreased thyroid function. Visual inspection of the thyroid for this characteristic can be made easily and rapidly; therefore, if present, this finding merits inclusion in the radiology report to alert the referring clinician of the possibility of decreased thyroid function.
Author contributions 1. Guarantor of integrity of the entire study: Pierre Maldjian 2. Study concept and design: Pierre Maldjian, Ting Chen 3. Literature research: Pierre Maldjian, Ting Chen 4. Clinical studies: Pierre Maldjian, Ting Chen 5. Experimental studies/data analysis: Pierre Maldjian, Ting Chen 6. Statistical analysis: Pierre Maldjian, ting Chen 7. Manuscript preparation: Pierre Maldjian, Ting Chen 8. Manuscript editing: Pierre Maldjian, Ting Chen
References 1. Kaneko T, Matsumoto M, Fukui K, et al. Clinical evaluation of thyroid CT values in various thyroid conditions. J Comput Tomogr 1979;3:1e4. 2. Iida Y, Konishi J, Harioka T, et al. Thyroid CT number and its relationship to iodine concentration. Radiology 1983;147:793e5. 3. Kamijo K. Clinical studies on thyroid CT number in chronic thyroiditis. Endocr J 1994;41:19e23. 4. Imanishi Y, Ehara N, Shinagawa T, et al. Correlation of CT values, iodine concentration, and histologic changes in the thyroid. J Comput Assist Tomogr 2000;24:322e6. 5. Carter Y, Sippel RS, Chen H. Hypothyroidism after a cancer diagnosis: etiology, diagnosis, complications and management. Oncologist 2014;19:34e43. 6. Han YM, Kim YC, Park EY, et al. Diagnostic value of CT density in patients with diffusely increased FDG uptake in the thyroid gland on PET/CT images. AJR Am J Roentgenol 2010;195:223e8. 7. Pandey V, Reis M, Zhou Y. Correlation between computed tomography density and functional status of the thyroid gland. J Comput Assist Tomogr 2016;40:316e9. 8. Hennessey JV, Espaillat R. Diagnosis and management of subclinical hypothyroidism in elderly adults: a review of the literature. J Am Geriatr Soc 2015;63:1663e73. 9. Baumgartner C, Blum MR, Rodondi N. Subclinical hypothyroidism: summary of evidence in 2014. Swiss Med Wkly 2014;144:w14058. 10. Hollowell JG, Staehling NW, Flanders WD, et al. Serum TSH, T(4), and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Examination Survey (NHANES III). J Clin Endocrinol Metab 2002;87:489e99. 11. Padmanabhan H. Amiodarone and thyroid dysfunction. South Med J 2010;103:922e30.
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
Contents lists available at ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism? P.D. Maldjian*, T. Chen Department of Radiology, University Hospital, Rutgers New Jersey Medical School, Newark, NJ, USA
art icl e i nformat ion Article history: Received 18 May 2016 Received in revised form 22 June 2016 Accepted 8 July 2016
AIM: To determine if visual assessment of the attenuation of morphologically normal appearing thyroid glands on unenhanced computed tomography (CT) of the chest is useful for identifying patients with decreased thyroid function. MATERIALS AND METHODS: This was a retrospective study of 765 patients who underwent both unenhanced CT of the chest and thyroid function tests performed within 1 year of the CT examination. Attenuation of the thyroid gland was visually assessed in each patient relative to the attenuation of the surrounding muscles to categorise the gland as “low attenuation” (attenuation similar to surrounding muscles) or “high attenuation” (attenuation greater than surrounding muscles). Thyroid attenuation was quantitatively measured in each case to determine the validity of the visual assessment. Results of thyroid function tests were used to classify thyroid function as hypothyroid, euthyroid, or hyperthyroid. Data were analysed to determine the relationship between visual assessment of thyroid attenuation and status of thyroid function. RESULTS: Thyroid glands of low attenuation were present in 4.2% (32/765) of the patients. Nearly half (47%) of the patients with low-attenuation thyroids had hypofunctioning thyroid glands. Compared to patients with high-attenuation thyroids, patients with low-attenuation thyroids were significantly more likely to have decreased thyroid function (clinical and subclinical hypothyroidism) and significantly less likely to be euthyroid (p<0.0001). Quantitative measurement of thyroid attenuation confirmed the validity of the visual assessment. CONCLUSION: Low attenuation of an otherwise normal-appearing thyroid gland on unenhanced CT of the chest is strongly associated with decreased thyroid function. Ó 2016 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction The thyroid gland is known to have somewhat higher attenuation than the soft tissues on unenhanced computed * Guarantor and correspondent: P.D. Maldjian, Department of Radiology, Rutgers New Jersey Medical School, University Hospital, 150 Bergen St, UH C-320, Newark, NJ 07103-2406, USA. Tel.: þ1 (973) 972 5188; fax: þ1 (973) 972 7429. E-mail address: [email protected] (P.D. Maldjian).
tomography (CT) due to its increased iodine content. Hypothyroidism has been shown to be associated with decreased iodine content within the thyroid gland. Previous investigations have demonstrated some correlation between CT attenuation of the thyroid gland on unenhanced CT and iodine content in patients with known thyroid disease.1e4 Anecdotally, the present authors have encountered instances on unenhanced CT of the chest in hypothyroid patients where otherwise normal-appearing thyroid glands
http://dx.doi.org/10.1016/j.crad.2016.07.008 0009-9260/Ó 2016 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
e2
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
demonstrated attenuation similar to that of the surrounding muscles. The purpose of this investigation was to determine if visual assessment of the relative attenuation of the thyroid gland compared to the surrounding muscles is useful in suggesting hypothyroidism in patients with otherwise normal-appearing thyroid glands on unenhanced CT.
Materials and methods This was a retrospective evaluation approved by the Institutional Review Board. Due to the retrospective nature of the study, the Institutional Review Board waived the requirement for informed consent. The radiology data base and electronic medical record system were searched to identify all adult patients (patients >18 years of age) who had both unenhanced chest CT performed between January 2012 and June 2015 and thyroid function tests performed within 1 year of the CT examination (986 patients). The CT images in these patients were reviewed to evaluate the appearance of the thyroid gland. If patients had more than one CT examination and/or more than one set of thyroid function tests, the CT examination and thyroid function results that were closest together temporally were used for analysis. Exclusion criteria were inability to visualise the thyroid gland on the CT images (prior thyroidectomy or exclusion of the thyroid gland from the scan range), gross abnormality of the thyroid (thyroid enlargement or presence of any thyroid nodules >1 cm in diameter) and any evidence of prior thyroid surgery (partial thyroidectomy or visible surgical clips in the thyroid bed). After application of these exclusion criteria, 765 patients were identified that comprised the study population. The CT images of the thyroid were visually assessed to categorise the appearance of the thyroid as either “low attenuation” or “high attenuation”. A thyroid gland of “low attenuation” was defined as attenuation at the same level or less than the adjacent muscles, such that it was visually difficult to discern the thyroid gland from the surrounding muscles based on its attenuation alone. “High attenuation” of the thyroid was defined as attenuation visibly greater than the surrounding muscles, such that the thyroid was easily discernible from the adjacent muscles. No distinction was made between attenuation of the thyroid slightly greater or significantly greater than the surrounding muscles. As long as thyroid attenuation was visibly greater than surrounding musculature, such that the thyroid was easily discernible, a “high-attenuation” score was recorded for the case (Fig 1). After visual assessment of thyroid attenuation in each case, the attenuation of the thyroid gland was directly measured. A circular region of interest (ROI) approximately 1 cm in diameter was placed over a homogeneous region of one lobe of the thyroid on the axial images and the attenuation in Hounsfield units (HU) was recorded. All CT examinations were conventional unenhanced chest CT examinations performed for standard clinical indications (i.e., cough, lung nodule assessment, haemoptysis, etc.) CT examinations were performed on a 16-section multidetector CT machine (Lightspeed 16, GE Healthcare Milwaukee, Wisconsin, USA) at 120 kVp with automatic tube current
modulation, maximum of 300 mA, noise index of 23, 0.8second rotation time, 16 1.25 mm collimation and pitch of 1.375. Images were reconstructed at 1.25 mm thickness at 1.25 mm intervals using a standard soft-tissue algorithm. Images were viewed on a conventional picture archiving and communication system (PACS) viewing station at routine soft-tissue windows (window width¼400 HU, window level¼40 HU). Studies were interpreted by a board certified radiologist with >20 years of experience in thoracic imaging. The reader was blinded to the results of the thyroid function tests at the time of image interpretation. The serum levels of thyroid stimulating hormone (TSH) for the study patients were assessed to classify patient thyroid function as hypothyroid, euthyroid, or hyperthyroid. As primary thyroid disease (disease of the thyroid itself) is much more common than secondary thyroid disease (pituitary disease affecting TSH production), it was assumed that abnormal TSH values reflected primary thyroid disease.5 Serum TSH was determined by immunmoradiometric assay and the normal reference range for TSH at the authors’ facility is currently 0.45e4.50 mIU/ml. Normal reference ranges have varied slightly over the last few years, so the reference range concurrent with the date of the TSH results was used to classify thyroid function. Patients with serum TSH values within the reference range were classified as euthyroid, patients with TSH values above the reference range were classified as hypothyroid, and patients with TSH values below the reference range were classified as hyperthyroid. For patients with abnormal serum TSH, free thyroxine values (if available) were also assessed to differentiate subclinical thyroid dysfunction from more severe disease. Free thyroxine levels were determined by radioimmunoassay with a normal reference range currently assumed to be 0.82e1.77 ng/dl. As with TSH, reference ranges for free thyroxine have varied over the last few years, so the reference range concurrent with the date of the free thyroxine results was used to classify thyroid function. The data were analysed to determine if there was any relationship between grading of thyroid attenuation by visual assessment and the status of thyroid function. Results from direct measurement of thyroid attenuation were assessed to confirm the validity of subjective assessment of thyroid attenuation. For statistical analysis, differences in proportions were compared using the chi-square test with the Yates continuity correction or the Fisher exact test. Differences in mean values were compared using the Student ttest. For all statistical tests, a two-tailed p-value of <0.05 was considered statistically significant.
Results The present results (Table 1) show that a low-attenuation thyroid is an unusual finding present in only 4.2% (32/765) of the present sample. Of these 32 patients, 12 were male (38%) and 20 (62%) were female. The mean age for these patients was 59.1 years (standard deviation [SD]¼11.9 years, range¼40e96). Seven hundred and thirty-three patients (96%) had high-attenuation thyroids. Of these, 347 were
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
e3
Figure 1 Examples of low-attenuation and high-attenuation thyroid glands on unenhanced CT. (a) A 74-year-old woman with a low-attenuation thyroid gland. Note that the thyroid gland (arrows) is visibly similar in attenuation to the surrounding muscles. The patient was hypothyroid with elevated TSH and low free thyroxine levels. (b) A 65-year-old euthyroid man with a high-attenuation thyroid gland. The thyroid (arrows) is slightly increased in attenuation compared to the surrounding muscles. (c) A 64-year-old euthyroid man with a high-attenuation thyroid gland. The thyroid (arrows) is markedly increased in attenuation compared to the surrounding muscles.
male (47%) and 386 were female (53%). The mean age was 58.8 years (SD¼13.9 years, range¼20e95 years). Direct measurement of thyroid attenuation confirmed that visual assessment could differentiate low-attenuation from highattenuation thyroid glands. The mean difference in
Table 1 Visual assessment of thyroid attenuation and its relationship to thyroid function.
Total number of patients Mean thyroid attenuation SD (HU) [Range (HU)] Number of hyperthyroid patients (%) Number of euthyroid patients (%) [95% confidence interval for proportion] Number of hypothyroid patients (%) [95% confidence interval for proportion] Number of hypothyroid patients with low free thyroxine (%) [95% confidence interval for proportion]
No. of patients with low-attenuation thyroid glands on CT
No. of patients with p-Value high-attenuation thyroid glands on CT
32
733
53.413.7
94.716
<0.0001
[20.7e76.6] 1 (3.1)
[55.7e145.2] 23 (3.1)
1.0
16 (50.0)
638 (87)
<0.0001
[0.34e0.66]
[0.84e0.89]
15 (46.9)
72 (9.8)
[0.3e0.64]
[0.08e0.12]
5 (33.3)
3 (5.4)a
[0.15e0.58]
[0.02e0.15]
SD, standard deviation. a Only 56 of the 72 patients had free thyroxine levels tested.
<0.0001
0.0086
attenuation between thyroid glands visually classified as low attenuation and those visually classified as high attenuation was 41.3 HU (95% confidence interval [CI]: 35.1e47.5 HU, p<0.0001). Nearly half (47%) of the patients with low-attenuation thyroids had hypofunctioning thyroid glands. Compared to patients with high-attenuation thyroids, patients with lowattenuation thyroids were significantly more likely to be hypothyroid and significantly less likely to be euthyroid (p<0.0001). All 15 of the hypothyroid patients with low attenuation thyroid glands had free thyroxine levels tested. In 10 of these patients, the free thyroxine levels were normal, indicating subclinical hypothyroidism (high TSH with normal free thyroxine), whereas in five patients the free thyroxine was low, consistent with clinical hypothyroidism. Of the 72 hypothyroid patients with highattenuation thyroids, 56 had free thyroxine levels tested. Only three of these patients had low free thyroxine (consistent with clinical hypothyroidism), whereas the remaining 53 patients had normal free thyroxine levels (consistent with subclinical hypothyroidism). Thus, patients with low-attenuation thyroid glands were significantly more likely to have laboratory findings consistent with clinical hypothyroidism compared to patients with highattenuation thyroid glands (p¼0.0086). Within the present total sample of 765 patients, 11.4% (87/765) were hypothyroid, 85.5% (654/765) were euthyroid, and only 3.1% (24/765) were hyperthyroid. There was no difference in the frequency of hyperthyroidism between low-attenuation thyroid patients and high-attenuation thyroid patients. Only one patient with a low-attenuation thyroid was hyperthyroid. Review of this patient’s records showed that the patient was hyperthyroid due to a hyperfunctioning thyroid nodule in the right lobe producing suppression of the remainder of the gland as demonstrated on I-123 thyroid scintigraphy (Fig 2). The free thyroxine level was normal in this patient (consistent with subclinical hyperthyroidism). Of the patients with high-attenuation
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
e4
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
Figure 2 A 57-year-old hyperthyroid woman with a low-attenuation thyroid gland; the only instance of a low-attenuation thyroid in a hyperthyroid patient in the present study. (a) Unenhanced CT shows a low-attenuation thyroid gland (arrows). (b) Right anterior oblique view from I-123 thyroid scintigraphy shows a hyperfunctioning nodule in inferior right lobe of the thyroid gland (arrow) with suppression of uptake in the remainder of the thyroid.
thyroids, 23 were hyperthyroid, but only one had elevated levels of free thyroxine. Thus laboratory findings were consistent with subclinical hyperthyroidism in almost all of the hyperthyroid patients. One patient who was initially euthyroid with a highattenuation thyroid on CT developed hypothyroidism with a low-attenuation thyroid on a follow-up study. Review of this patient’s medical records showed that she presented with small cell lung carcinoma and developed hypothyroidism after radiation treatment and chemotherapy (Fig 3). Of the five patients with clinical hypothyroidism and low-attenuation thyroid glands, two patients underwent chest CT before thyroid function tests, while three underwent CT after the thyroid function tests. Only one of these patients was receiving treatment with thyroid hormone at the time of the CT examination. Three of the remaining four patients were prescribed levothyroxine at some time after CT was performed. Of the 10 patients with subclinical
hypothyroidism and low-attenuation thyroid glands, six underwent chest CT before the thyroid function tests, whereas four underwent CT after the thyroid function tests. None of these patients was receiving treatment with thyroid hormone at the time their CT examination was performed, although seven of these 10 patients were prescribed levothyroxine at some time after their CT examinations. As chest CT was performed before thyroid function tests in eight of these 15 hypothyroid patients, there was potential to suspect hypothyroidism in these eight patients based on the CT appearance of the thyroid before the thyroid function test results were known.
Discussion The thyroid gland contains 25% of the iodine in the body stored within the follicular cells as thyroglobulin. Hence the thyroid is normally high in attenuation on unenhanced CT.
Figure 3 A 65-year-old woman with small cell lung cancer whose thyroid gland changed in attenuation on subsequent CT. (a) Unenhanced CT image before therapy shows a high-attenuation thyroid gland (arrows). The patient was euthyroid at this time. (b) Eighteen months later after radiation treatment and chemotherapy, the patient developed clinically significant hypothyroidism. Unenhanced CT at this time shows that the thyroid (arrows) has become low in attenuation. (Note the small calcification in the posterior left lobe of the thyroid indicating that this image is at the same level as Fig. 3a.) Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
Low attenuation of the thyroid on CT, however, has only been sporadically addressed in the literature. In 1979, Kaneko et al.1 showed that the average attenuation of the thyroid gland on CT in a small series of five hypothyroid patients was significantly lower than that of 44 normal patients. In 1983, Iida et al.2 studied thyroid attenuation in 67 untreated patients with various thyroid disorders (Grave’s disease, Hashimoto’s thyroiditis, simple goitre, thyroid adenoma, thyroid cyst, and thyroid cancer) compared to 24 normal patients. The study reported that the mean attenuation of diseased thyroid glands was significantly lower than that of normal thyroid glands. In that investigation, tissue iodine concentrations in 14 surgical specimens were directly measured and compared to the CT attenuation values of the corresponding thyroid regions in the preoperative studies. A strong correlation was demonstrated suggesting that the CT attenuation value of the thyroid is primarily determined by the iodine content of the tissue.2 In 1994, Kamijo3 evaluated CT attenuation of the thyroid in 155 patients with chronic thyroiditis compared to 95 normal control patients. The investigation showed that CT attenuation of diseased thyroid glands was significantly lower than normal thyroid glands and also showed that the serum TSH concentration was highest (indicating hypothyroidism) in the group of patients with lowest thyroid attenuation.3 In 2000, Imanishi et al.4 correlated results of CT attenuation of the thyroid in 36 patients with thyroid disease with the histology of resected specimens. The investigators found that reduced attenuation of the thyroid in patients with thyroid disease was secondary to decreased iodine concentration in the thyroid follicles, decrease in colloid content, and increase in follicular cells and interstitial structures.4 In 2010, Han et al.6 reported the results of a study on thyroid CT attenuation and thyroid function tests in 56 patients with diffuse hypermetabolic thyroid activity on combined positron-emission tomography (PET)/CT compared to 56 control patients. Patients with hypermetabolic thyroid activity on PET/CT demonstrated significantly decreased CT attenuation compared to controls. The investigators also assessed thyroid attenuation on CT using a subjective visual scale and found that 56% of patients with elevated TSH (13/23, all with hypermetabolic thyroid activity on PET/CT) demonstrated low attenuation. The authors concluded that decreased thyroid attenuation on CT in the study patients implied reduction in colloid volume and replacement with hypermetabolic tissue such as follicular cells, inflammatory cells (lymphocytes), or other stromal tissue.6 In a more recent investigation, Pandey et al.7 evaluated the relationship between serum TSH and thyroid attenuation on unenhanced CT of the cervical spine. The investigators found that patients with low TSH levels and patients with high TSH levels demonstrated decreased thyroid attenuation on CT.7 The present study differs from these prior investigations in that the present study was undertaken to determine if visual assessment of the attenuation of an otherwise normal-appearing thyroid gland in a population not known
e5
to have thyroid disease was useful as an indicator of thyroid dysfunction. Direct quantitative measurement of thyroid attenuation in the present investigation confirms that rapid visual assessment of thyroid attenuation on CT images is sufficient to identify patients with low-attenuation thyroid glands with increased prevalence of thyroid dysfunction. The present findings indicate that low attenuation of the thyroid on CT in such patients is strongly associated with hypothyroidism. Sixteen percent of patients with low-attenuation thyroid glands in the present study (5/32) had laboratory findings consistent with clinical hypothyroidism. Hypothyroidism is the most common hormone deficiency, but due to the wide variation in severity and the non-specific nature of the symptoms, it can often go undiagnosed. Common clinical manifestations of hypothyroidism include fatigue, weight gain, muscle weakness, cognitive impairment, and depression. If untreated, life-threatening complications, such as heart failure, can occur. Clinical hypothyroidism is easily treated with levothyroxine sodium (L-thyroxine), an orally administered synthetic thyroid hormone. Treatment has been shown to alleviate the symptoms and to improve the quality of life in overtly hypothyroid patients.5 Thirty-one percent of patients with low-attenuation thyroid glands in the present study (10/32) had laboratory findings consistent with subclinical hypothyroidism. Subclinical hypothyroidism is a laboratory diagnosis characterised by an elevated TSH level with a normal free thyroxine level. The prevalence of subclinical hypothyroidism is estimated to be 3.1e8.5% in the general adult population. It is usually caused by chronic autoimmune thyroiditis (Hashimoto’s thyroiditis). It is often asymptomatic, but clinical manifestations can be similar to overt hypothyroidism. Prospective studies of patients with subclinical hypothyroidism have reported increased risk of coronary heart disease events, heart failure, and cardiovascular mortality with increased odds of myocardial infarction to as much as 3.1 times as high as in euthyroid controls.8 Subclinical hypothyroidism is also associated with an increased risk of progression to overt clinical hypothyroidism with an average yearly progression rate of 2e6%. To reduce the rate of these complications, although current data are inconclusive, there is a consensus to treat adults with subclinical hypothyroidism if TSH levels are >10 mIU/ml and they are symptomatic or test positive for anti-thyroid antibodies.8,9 As illustrated in Figs 2 and 3, two especially noteworthy cases were encountered during the course of the present investigation. Only one out of 24 hyperthyroid patients in this study had a low-attenuation thyroid. This patient was hyperthyroid secondary to a hyperfunctioning nodule that produced suppression of iodine metabolism and presumably reduction of iodine content in the remainder of the thyroid (Fig 2); further supporting the hypothesis that iodine concentration is the key contributor to the attenuation of the thyroid gland on unenhanced CT.4 In a second patient, the thyroid gland decreased in attenuation when she progressed from a euthyroid to a hypothyroid state (Fig 3). This suggests that visibly apparent reduction of
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008
e6
P.D. Maldjian, T. Chen / Clinical Radiology xxx (2016) e1ee6
thyroid attenuation in a patient over time might be an indication of impaired thyroid function. Further investigation is necessary to determine the validity of that conjecture. The present study has several limitations. There was no histological analysis of tissue to confirm the cause of low attenuation of the thyroid on CT. Given the results of prior investigations, decreased iodine concentration is the most likely explanation. There may be selection bias in the present study as a retrospective analysis was performed on patients who had both thyroid function tests and CT examinations of the chest. As thyroid function tests would be obtained in patients clinically suspected to have underlying thyroid disease, patients with thyroid disease might be overrepresented in the present sample compared to the general population. It has been estimated that the prevalence of hypothyroidism in the United States population is 4.6% (0.3% clinical and 4.3% subclinical).10 In the present sample, 11.4% (87/765) of the patients were hypothyroid suggesting that thyroid disease may be overrepresented in the study patients compared to the general population; however, this bias would only increase the overall prevalence of thyroid disease in the sample and would not affect the conclusions. The use of specific medications was not assessed in the present patients, which can be a confounding variable. For example, amiodarone is a widely used anti-arrhythmic drug with high iodine content and structural resemblance to thyroid hormones. Use of amiodarone can result in elevation of TSH levels, hypothyroidism, or thyrotoxicosis.11 Concerning use of thyroid hormone, however, as only one of the 15 patients with low-attenuation thyroids and hypothyroidism was receiving treatment with thyroid hormone at the time of chest CT, it is not thought that medications for hypothyroidism affected the present results. The present study was limited to conventional unenhanced CT examinations of the chest using the technical parameters specified. Other types of unenhanced CT examinations that include the thyroid, such as lung cancer screening, employ different technical parameters. Further investigation would be necessary to determine if thyroid attenuation is related to thyroid dysfunction on other types of CT studies in a fashion similar to the present findings. In conclusion, the present study shows that low attenuation of an otherwise normal-appearing thyroid gland on
unenhanced CT of the chest is strongly associated with decreased thyroid function. Visual inspection of the thyroid for this characteristic can be made easily and rapidly; therefore, if present, this finding merits inclusion in the radiology report to alert the referring clinician of the possibility of decreased thyroid function.
Author contributions 1. Guarantor of integrity of the entire study: Pierre Maldjian 2. Study concept and design: Pierre Maldjian, Ting Chen 3. Literature research: Pierre Maldjian, Ting Chen 4. Clinical studies: Pierre Maldjian, Ting Chen 5. Experimental studies/data analysis: Pierre Maldjian, Ting Chen 6. Statistical analysis: Pierre Maldjian, ting Chen 7. Manuscript preparation: Pierre Maldjian, Ting Chen 8. Manuscript editing: Pierre Maldjian, Ting Chen
References 1. Kaneko T, Matsumoto M, Fukui K, et al. Clinical evaluation of thyroid CT values in various thyroid conditions. J Comput Tomogr 1979;3:1e4. 2. Iida Y, Konishi J, Harioka T, et al. Thyroid CT number and its relationship to iodine concentration. Radiology 1983;147:793e5. 3. Kamijo K. Clinical studies on thyroid CT number in chronic thyroiditis. Endocr J 1994;41:19e23. 4. Imanishi Y, Ehara N, Shinagawa T, et al. Correlation of CT values, iodine concentration, and histologic changes in the thyroid. J Comput Assist Tomogr 2000;24:322e6. 5. Carter Y, Sippel RS, Chen H. Hypothyroidism after a cancer diagnosis: etiology, diagnosis, complications and management. Oncologist 2014;19:34e43. 6. Han YM, Kim YC, Park EY, et al. Diagnostic value of CT density in patients with diffusely increased FDG uptake in the thyroid gland on PET/CT images. AJR Am J Roentgenol 2010;195:223e8. 7. Pandey V, Reis M, Zhou Y. Correlation between computed tomography density and functional status of the thyroid gland. J Comput Assist Tomogr 2016;40:316e9. 8. Hennessey JV, Espaillat R. Diagnosis and management of subclinical hypothyroidism in elderly adults: a review of the literature. J Am Geriatr Soc 2015;63:1663e73. 9. Baumgartner C, Blum MR, Rodondi N. Subclinical hypothyroidism: summary of evidence in 2014. Swiss Med Wkly 2014;144:w14058. 10. Hollowell JG, Staehling NW, Flanders WD, et al. Serum TSH, T(4), and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Examination Survey (NHANES III). J Clin Endocrinol Metab 2002;87:489e99. 11. Padmanabhan H. Amiodarone and thyroid dysfunction. South Med J 2010;103:922e30.
Please cite this article in press as: Maldjian PD, Chen T, Is visual assessment of thyroid attenuation on unenhanced CT of the chest useful for detecting hypothyroidism?, Clinical Radiology (2016), http://dx.doi.org/10.1016/j.crad.2016.07.008