When SARS CoV-2 meets thyroid - short-term and long-term patients’ outcomes

Purpose: The interaction between thyroid and SARS CoV-2 is complex and not yet fully understood. The aim of this study was to identify a possible association of the previous thyroid disease and serum TSH level on the short-term and long-term outcomes of patients hospitalised for COVID-19. Methods: We retrospectively analysed the data of 1423 patients aged 18 years or older, with known previous thyroid status who were admitted for COVID-19 to Clinical Hospital Dubrava, Zagreb, Croatia, between March 2020 and June 2021. For 614 patients whose serum TSH was determined during hospitalisation we examined short-term outcomes, including the required level of oxygen supplementation, admission to the ICU and death. The MSCT image of the chest and TSH was recorded in the post-COVID outpatient clinic six to eight weeks after the discharge from the hospital and was used to estimate the long-term COVID-19 complications. Patients were divided into three groups according to measured TSH levels. Results: Those with lower/suppressed TSH (<0.4 mIU/L) had worse short-term outcomes and were more frequently admitted to the ICU (p<0.011), needed mechanical ventilation (p<0.001) or died (p=0.03). In the post-COVID period, there was an overall increase in the TSH levels when compared to TSH during hospitalisation. The greatest increase was detected among the patients with suppressed/lowered TSH level during hospitalisation, and it correlated signi�cantly with MSCT �nding of higher pulmonary involvement (p<0.001). Conclusion: Changes in the serum TSH level may be an important indicator for the course of COVID-19 and its long-term outcomes.


Introduction
The coronavirus disease 2019 (COVID-19) pandemics, caused by the severe acute respiratory syndrome coronavirus-2 (SARS CoV-2), still presents a threat to global health.It is of utmost importance to better understand pathophysiology of the disease and timely recognize phenotypes of vulnerable patients who are at the highest risk of immediate and long-term unfavourable outcomes.
The pathogenesis of COVID-19 involves SARS CoV-2 entering the respiratory system.When the angiotensin-converting enzyme 2 (ACE2) on the surface of pneumocytes binds to the spike protein of the virus, it mediates the entry of the virus into cells [1,2].Many organs, including endocrine glands, have been found to express ACE2 leading to the potential extrapulmonary and multiorgan spread of SARS CoV-2 [3,4].
A statement from the European Society of Endocrinology was published in March 2020 to raise awareness of the whole endocrine community about the role and responsibilities of endocrinologists worldwide during the current COVID-19 outbreak.People with diabetes were considered among those in high-risk categories for serious illness, together with people with obesity, malnutrition, and adrenal insu ciency [5].In an update published in April 2021 an emerging endocrine phenotype of COVID-19 was described and the importance of thyroid aspects of COVID-19 was highlighted [6].
In our study, we aimed to investigate the impact of the pre-existing thyroid status and serum thyroid stimulating hormone (TSH) level on short-term and long-term COVID-19 outcomes.

Materials And Methods
This study was approved by the Ethics Committee of Clinical Hospital Dubrava (approval number 2021/1312-02).Due to a retrospective nature of the study, informed consents were waived.

Study participants and data
We retrospectively examined the medical records of 1423 patients aged 18 years or older, with the known previous thyroid status, who were admitted for COVID-19 to Clinical Hospital Dubrava, Zagreb, Croatia, between March 2020 and June 2021.For 614 patients whose serum TSH was determined during hospitalisation we examined short-term outcomes, and for 1020 patients who were followed in the post-COVID outpatient clinic and had TSH tested we investigated long-term outcomes.The COVID-19 diagnosis was de ned as a real-time reverse transcriptase polymerase chain reaction con rmation of infection from a nasopharyngeal swab.The patients underwent a standard and additional set of blood tests within 48 hours from the admission, including the blood glucose level, glycated haemoglobin (HbA1c), C-reactive protein (CRP), interleukin-6 (IL-6), procalcitonin (PCT) and D-dimer.According to the TSH level, patients were classi ed into three categories; the rst category with a normal TSH level ranging from 0.4-4 mIU/L, the second category with an elevated TSH level in case TSH > 4 mIU/L and the third one with a lowered/suppressed TSH level in case TSH < 0.4 mIU/L.For 72 patients we also had free thyroxine levels (FT4) available.The laboratory analysis was taken at the Clinical Department of Laboratory Diagnostics -ISO 15189:2012 accredited laboratory, in Clinical Hospital Dubrava.Clinical and demographic data and comorbidities of interest were recorded, and the Charlson Comorbidity Index was calculated using the Charlson Comorbidity Index calculator.Parameters for estimation of the severity of COVID-19 infection were recorded during hospitalisation including duration of the hospitalisation, level of oxygen supplementation needed, admission to the Intensive Care Unit (ICU) and a death rate, and were used to estimate short-term COVID outcomes.TSH was repeated and the multi-slice computed tomography (MSCT) image of the chest was recorded in the post-COVID outpatient clinic six to eight weeks after the discharge from the hospital, and they were used to estimate long-term COVID-19 complications.

Statistical Methods
The data was analysed using SPSS software (IBM SPSS Statistics for Windows, Version 25.0.Armonk, NY: IBM Corp.).Descriptive statistics is used to outline the basic features of the sample in a study (proportions for categorical data and mean with a standard deviation for normally distributed continuous variables, or median and interquartile range for variables deviating from normal distribution).Normality of distribution was checked using the Shapiro-Wilk test.Binary or categorical outcome variables were analysed with a Chi-square test or Fisher's exact test (alternative to the Chi-square for 2x2 contingency tables).
For continuous variables ANOVA was used to compare differences between three groups of patients categorised according to TSH levels.The Kruskal-Wallis non-parametric test was used where the homogeneity of variances was violated.The paired T-test or Wilcoxon's test was used to compare the difference in means of continuous measuring of variables in hospital and the post-COVID outpatient clinic.A binary logistic regression was calculated to determine what are statistically signi cant predictors for death outcome (Odds Ratios with 95% CI are shown).

Results
Data from 1423 COVID-19 positive patients were analysed, of which 186 (13.1%) had thyroid disease.Of 186 patients with thyroid disease 166 received levothyroxine therapy.Patients with a history of thyroid disease were not associated with worse short-or longterm outcomes.The serum TSH level of 614 patients was examined during hospitalisation and those patients were included in our study.Short-term (maximum oxygen supplementation, admission to the ICU, death) and long-term outcomes (post-COVID lung affection assessed by the chest MSCT) were analysed according to the TSH status during hospitalisation and post-COVID period.

Short-term outcomes and TSH
We classi ed patients into three categories according to their TSH status during hospitalisation.Most of our patients had TSH within the reference ranges (65.3%).Patients in the elevated TSH group were more frequently female and had a history of thyroid disease.Patients with lower/suppressed TSH more frequently had pneumonia and needed oxygen supplementation including High Flow Oxygen Therapy (HFOT), corticosteroid treatment (all p < 0.001) and insulin (p < 0.033).They also had worse short-term outcomes and were more frequently admitted to the ICU (p < 0.011), needed mechanical ventilation (p < 0.001) or died (p = 0.03).(Table 1 and Table 2) Corticosteroid treatment differed between the three groups; patients with lower/suppressed TSH were more frequently treated with corticosteroids (χ2 = 14.976, p = 0.001).(Fig. 1)

Long-term outcomes and TSH
Long-term outcomes were de ned as the percentage of pulmonary parenchyma changes after COVID-19, assessed by the MSCT of the chest six to eight weeks after hospitalisation.We divided patients into four groups: a) no changes, b) less than 30% of parenchyma affected, c) 30-50% of parenchyma affected, d) more than 50% of parenchyma affected.In the post-COVID period, there was an overall increase in the TSH levels when compared to TSH measured during hospitalisation.The greatest increase was detected among patients with suppressed/lowered TSH during hospitalisation, and it correlated signi cantly with MSCT nding of higher pulmonary involvement (p < 0.001, Fig. 2.)

Discussion
The interaction between SARS CoV-2 and thyroid function is bidirectional and complex, and it is not yet fully understood.There is no clear evidence that patients with the existing thyroid dysfunction are more at risk of developing more severe COVID-19 disease [7].In our study, the pre-existing thyroid disease was not associated with increased risk of COVID-19 related worse outcomes in hospitalised patients, including death, which is in agreement with the published literature so far [8,9].In a retrospective study conducted by van Gerwen M et al., hypothyroidism was not associated with an increased risk of hospitalisation, mechanical ventilation nor death [10].Similar to previous ndings, in our patient's cohort, 88.8% had no history of pre-existing thyroid disorder, and of those with previous thyroid disorder 10.5% had hypothyroidism and 0.7% reported hyperthyroidism.
In our study, lower/suppressed TSH level during hospitalisation for COVID-19 was associated with worse short-term outcomes, including admission to the ICU, mechanical ventilation and death.In addition, patients with lower TSH were more frequently treated with corticosteroids.Non-thyroidal illness (NTI) is a well-known phenomenon reported in critically ill patients and in many diseases.It is characterised by decreased free triiodothyronine (FT3) levels and low or normal TSH and high, normal or low FT4 levels [11,12].One of the shortcomings of this study was that FT3 and FT4 were not assessed at admission, owing to the retrospective nature of the study.Furthermore, TSH was tested while most patients were treated with glucocorticoids, an additional factor affecting the pituitarythyroid axis.Therefore, it is di cult to exclude the effect of corticosteroids and systemic in ammation on the pituitary-thyroid axis, leading to a possible misdiagnosis of thyroid dysfunction in severe cases of COVID-19.NTI is considered the most frequent thyroidrelated problem in the COVID-19 care, particularly in hospitalised patients and in intensive care units [7].In a retrospective study by Chen M et al. 56% of patients, hospitalised for COVID-19 infection, had signi cantly lower-than-normal values of TSH compared with healthy controls and non-COVID-19 pneumonia patients with a similar degree of severity.After recovery, the levels of all thyroid hormones returned to the normal value, with no signi cant differences in the thyroid function [13].In our patients we did not see the differences in laboratory parameters associated with severity of viral disease (such as CRP, IL-6, procalcitonin, D-dimer) and TSH level.Moreover, we did not see any difference between the duration of hospital stay according to patients' TSH level.Khoo et al. aimed to detail the acute effects of COVID-19 on the thyroid function.In 50 patients included in the study, signi cant differences in TSH were found comparing baseline vs admission and admission vs follow-up values.On the contrary, there was no signi cant difference between baseline vs follow-up values [14].We observed that patients who had lower TSH during hospitalisation, were the patients in whom TSH rose more signi cantly in the post-COVID period.
On the other hand, post-mortem studies suggest morphological and pathological changes of endocrine glands, including the thyroid, in people affected by the coronavirus family not previously diagnosed with the thyroid disease [15,16].In the study of Muller et al. patients treated in high intensity care units (HICU) due to COVID-19 showed lower TSH levels than SARS-CoV-2 negative patients admitted to the same HICU.Furthermore, serum FT4 levels showed no difference between the groups while the FT3 levels were low in both groups [9].In our patient cohort we only had FT4 levels determined for 72 patients, and in 56.9% of cases it was normal, while elevated FT4 was detected in 41.7% of patients.Unfortunately, FT3 was not determined.Therefore, due to the small sample size and the absence of FT3, we cannot interpret the ndings and differentiate between NTI syndrome or hyperthyroidism due to other causes.
The cases of subacute thyroiditis with a possibility of clinically manifest hyperthyroidism, as well as the cases of autoimmune thyroiditis or Graves' disease triggered by the "cytokine storm" induced by SARS-CoV-2 infection, have been described so far [17][18][19][20].
As mentioned earlier, we did not detect the correlation between cytokine levels (speci cally IL-6) and TSH levels.
There is limited knowledge of thyroid function in patients who have recovered from COVID-19 and medical literature lacks data about the potential thyroid in uence on non-speci c symptoms often described during the post-COVID period which negatively in uence patients' quality of life and recovery.Due to the retrospective nature of the study, we could not investigate non-speci c symptoms that patients had reported.Therefore, we aimed to investigate if there is an association between objective changes in lung parenchyma (seen on chest MSCT after hospitalisation) and serum TSH.It is known that the T3 receptor is present in the lung in the alveolar type II cells that are believed to be involved in the recovery after lung injury [21].Furthermore, studies in rats showed that hypothyroid rats are less able to clear uid from their lungs, and uid clearance can be ameliorated by liothyronine administration [21,23].In our study, those patients that had lower TSH during hospitalisation, also had worse long-term outcomes and more pulmonary parenchyma affected by COVID.As already mentioned, those were the patients with a more signi cant increase of TSH in the post-COVID period.It would be interesting to see if those are the patients that had low FT3, and more research is needed to answer this question.Currently, a phase 2 clinical trial (NCT 04115514) is being conducted, investigating liothyronine as a treatment for ARDS in humans, including that associated with COVID-19 [24].
One of the main shortcomings of this study, owing to the retrospective analysis, was that FT3 and FT4 were not assessed at admission.In addition, the serum TSH was tested while patients were receiving glucocorticoids.Furthermore, the patients included in this study were hospitalised patients with moderate to severe disease.Thus, we lack data from patients with milder disease.Nevertheless, to our knowledge, this is the rst study investigating the serum TSH and the COVID-19 associated outcomes of patients in a population drawn from the rst and largest COVID-19 hospital in Croatia, with a follow-up to assess post-COVID pulmonary involvement.The results of our research may indicate that lower TSH might be a predictor of worse outcomes.
In conclusion, although sick euthyroidism is the most frequent thyroid-related problem encountered in COVID-19 patients, especially those with more severe disease, other thyroid disorders have also been reported.Physicians should be aware of possible connections between SARS CoV-2 and the thyroid.Further prospective studies including FT3 and FT4 are needed to investigate the impact of COVID-19 on hypothalamic-pituitary-thyroid axis and vice versa, as well as the impact of thyroid changes on outcomes of patients with COVID-