Endocrine Abstracts

Background: The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic disease (Covid-19) affects thyroid function with different mechanisms: non-thyroidal illness syndrome (NTIS), direct infection of thyroid gland and cytokine storm. We provided the first description of painless atypical thyroiditis coexisting with NTIS in patients hospitalised for moderate-to-severe Covid-19 disease. We aimed to: 1) correlate thyroid dysfunction with Covid-19 disease severity; 2) follow the evolution of thyroid function over time.

Methods: Baseline (at hospital admittance) and longitudinal study of patients hospitalised for moderate-to-severe Covid-19 disease, without known history of thyroid disfunction, assessing serum thyroid function and autoantibodies, inflammatory markers and thyroid ultrasound scan (US). Patients showing at US focal hypoechoic areas suggestive for thyroiditis (thyroiditis-areas) also underwent thyroid 99mTc or I123 uptake scan.

Results: 183 Covid-19 patients were studied baseline, of whom 63 (34%) were already on steroid treatment before hospital admission, thus were not considered for TSH analysis. Decreased serum TSH positively correlated with albumin (P=0.02) and lymphocyte count (P<0.01) but not with C-reactive-protein (P=0.12) and interleukin-6 (P=0.10); TSH also progressively and inversely correlated to the need of oxygen support (P=0.02). Serum FT3 correlated positively with albumin (P<0.01) and inversely with D-dimer (P=0.02). Baseline thyroid US scan showed thyroiditis-areas in 18/65 (28%) patients, associated with reduced thyroid uptake at ^99mTc/I¹²³ scintigraphy in 14/17 (82%) cases. Thyroiditis-areas were more frequent among patients with baseline low TSH (6/10, 60%) compared with those with normal TSH (10/40, 25%, P=0.034). The patients with thyroiditis-areas also had higher baseline FT4 (P=0.018) and IL-6 (P=0.016) compared with those with normal thyroid US. Follow-up analysis was conducted in 75/183 (41%) patients; thyroid function and inflammatory markers normalized at all time-points in nearly all cases and no increase of thyroid autoantibodies positivity was observed. The thyroiditis-areas, even if often reduced in size, were still present after 6 and 12 months in 13/15 (87%) and 6/12 (50%) patients, respectively. After 9 months the thyroid uptake at ^99mTc/I¹²³ scintigraphy was still reduced in 4/6 (67%) patients, even if partially recovered (mean +28%) compared with baseline.

Conclusions: Thyroid dysfunction during moderate-to-severe Covid-19 disease is mild and transient, and thyroid hormones correlate with disease severity. Thyroiditis-areas at US occur frequently and may persist after one year, even if reduced in size; long-term consequences are unknown. The association of thyroiditis-areas with low TSH and high FT4 and IL-6 serum concentrations support the hypothesis of direct thyroid gland involvement in SARS-CoV-2 infection.