Hypothyroidism May Be Associated with Post-COVID-19 Pulmonary Fibrosis

Dear Editor:

Post-COVID-19 pulmonary fibrosis (PCPF) is a debilitating consequence of COVID-19 infection. Patients may present with symptoms similar to acute COVID-19 infection, including dry cough, shortness of breath, fatigue, nail clubbing, weight loss, and decreased exercise tolerance. The effect of hypothyroidism on a patient's risk for developing PCPF has not been determined. ¹ Other risk factors for PCPF such as asthma, smoking history, and diabetes mellitus have been identified and reviewed; however, data specifically relating hypothyroidism to a major chronic consequence of COVID-19 are unavailable. ²

Previously, low thyroid function has been linked to an increased risk of general chronic fibrosis in the heart, liver, and lungs. ³ Given the high prevalence rate of hypothyroidism and COVID-19 infection, it is important for physicians to be aware of the potential association of hypothyroidism and PCPF. In consideration of prior identified relationships between hypothyroidism and fibrotic disease, hypothyroidism as a risk factor for PCPF is evaluated.

The study design was a retrospective case–control study. Ethical approval was obtained from the institutional review board (IRB) at St. Luke's University Health Network (IRB #SLIR 2021-62). Research was completed in accordance with the Declaration of Helsinki as revised in 2013. The IRB waived the need for patient consent.

There were 730 patients at St. Luke's University Health Network (SLUHN) previously diagnosed with COVID-19 who also had a computed tomography (CT) scan 90 days or more after their first COVID-19 diagnosis. Patients included in the PCPF group were required to have both a CT scan before diagnosis of COVID-19 with no parenchymal abnormalities and a post-COVID-19 CT scan that demonstrated radiological changes. Pertinent terms used to assess fibrotic changes included bronchiectasis, honeycombing, fibrosis, interstitial lung disease, reticulation, scarring, thickening, and crazy paving.

Patients who received radiation or had another coexisting fibrotic disease were placed into the general COVID-19 group. Supplementary Figure S1 shows the process by which the two comparison groups were compiled. Demographic characteristics were collected for the remaining 120 patients who fit our criteria of PCPF, within a time frame of April 2, 2020, to January 28, 2022. Comorbidities, such as hypothyroidism, were established by reviewing patient electronic medical records for active problems diagnosed by primary care provider or specialist.

Similar information was collected for 7810 patients of the St. Luke's Health Network who were previously admitted with COVID-19 within that time frame, and did not fulfill our criteria for PCPF. Data cleaning and statistical analysis were done using R version 4.2.2. ⁴ p-Values were calculated using chi-square tests for categorical variables and Wilcoxon's rank sum test for the non-normally distributed continuous variables. Table 1 summarizes the bivariate results of the two groups.

Multivariable selection was done based on a significant bivariate relationship with fibrosis (as defined by a p-value of 0.05). The selected variables included asthma, chronic obstructive pulmonary disease (COPD), hypothyroidism, hypercholesterolemia, diabetes, hypertension, and mechanical ventilation during admission. As COVID-19 vaccinations were not widely available for a significant portion of the study period, vaccination was included despite an insignificant bivariate association with fibrosis. Multivariable modeling was performed using logistic regression to calculate adjusted odds ratios (aORs).

Collinearity was assessed with variance inflation factor, overall model fit was assessed using the Hosmer–Lemeshow test along with a likelihood ratio test comparing the fully specified model with an intercept-only model, and assessment of hypothyroidism was done using a likelihood ratio test comparing the fully specified model with that without hypothyroidism.

The patients who developed PCPF (120) differed from the overall COVID-19 inpatient population (7810) in that they were more likely to have been diagnosed with hypothyroidism (35.0% vs. 22.4%), hypercholesterolemia (77.5% vs. 64.2%), diabetes (44.2% vs. 29.9%), hypertension (74.2% vs. 51.6%), asthma (24.7% vs. 11.3%), and COPD (31.5% vs. 12.7%) (Table 1). They were also more likely to have required a ventilator during admission (11.2% vs. 5.6%). They resembled the overall COVID-19 inpatient population in age, sex, race, and percentage vaccinated against COVID-19 (p-value >0.05).

Upon multivariable analysis, the relationships with hypothyroidism (aOR 1.92 [CI 1.23–3.34], p 0.004), asthma (aOR 1.94 [CI 1.153.16], p 0.01), COPD (aOR 2.08 [CI 1.27–3.34], p 0.003), hypertension (aOR 2.49 [CI 1.43–4.50], p 0.002), and mechanical ventilation (aOR 3.46 [CI 1.636.64], p 0) remained statistically significant, but those of diabetes (aOR 1.19 [CI 0.75–1.88], p 0.468) and hypercholesterolemia (aOR 1.64 [ CI 0.91–3.00], p 0.088) were found to be statistically insignificant (Table 2).

Of our 120 patients who ruled in for PCPF, 41 patients required supplemental oxygen after their COVID-19 diagnosis, including 25 patients who continued to require supplemental oxygen as of March 2023 or at the time of death. The rate of supplemental oxygen use after COVID-19 infection was not significantly associated with hypothyroidism in the PCPF population, for both patients without supplemental oxygen preinfection (OR 1.72 [CI 0.79–3.71], p 0.17) and patients requiring supplemental oxygen preinfection (OR 2.38 [CI 0.97–5.8], p 0.057).

From the onset of the COVID-19 pandemic, glucocorticoids were a mainstay of treatment at SLUHN for inpatients of all severity levels. The use of glucocorticoid for outpatients was at the discretion of providers. Blanket use of glucocorticoids for SLUHN patients diagnosed with COVID-19 prevented further examination of potential associations between such treatment and PCPF.

PCPF is a serious complication that increases probability of lung damage, mechanical ventilation, and death. ¹ When comparing our identified risk factors with other published studies, we have discovered contrasting results. A prior meta-analysis on PCPF established a protective nature for diabetes and hypertension. ² Given that millions of people have already contracted COVID-19 and thousands more continue to contract it daily, physicians must be aware of PCPF and the impacts it can have on their patients. Given the high prevalence of hypothyroidism, it is important that endocrinologists recognize this as a potential factor that could place their patients at a higher risk for PCPF.

Thyroid hormone's role in metabolic and developmental processes is well established, and it is now well known that thyroid hormone plays a major role in gene expression. ⁵ One gene of particular interest is the transcription factor purine-rich box1 (PU.1), which plays a major role in regulating fibroblast profibrotic gene expression programs. ⁶ Prior investigation shows that PU.1 expression is normally silent in fibroblasts, and loss of epigenetic control where PU.1 is thereby induced results in a profibrotic phenotypic shift for affected fibroblasts. ⁷

Abnormal overexpression of IL-9 as a result of PU.1 binding to Th9 cells was associated with lung fibrosis in mice studies. ⁶ In analysis of bronchoalveolar lavage fluid of humans, variation in the levels of IL-9/IL-9R between different types of interstitial lung disease suggests differential pathogenesis. ⁸ The exact relationship between PU.1 and thyroid hormone in lung fibrosis is unclear; however, the two have been connected through several other components, such as miR-29c, in the development of hepatic fibrosis. ⁹ It may be a similar case that PU.1 and thyroid hormone are linked through a multitude of shared interactions that could predispose patients to developing PCPF.

Elsewhere in the literature, it has been suggested that inadequate thyroid hormone may decrease the size and number of Type II pneumocytes and decrease the ability of the lung to reabsorb fluid, which ultimately inhibits recovery from acute respiratory distress syndrome. ¹⁰ Lastly, the interstitial deposition of hydrophilic mucopolysaccharides seen in hypothyroidism also promotes a profibrotic state, in the context of myxedema. ³ Although interaction between thyroid hormone and PU.1 serves as a possible explanation for PCPF development, the mechanisms by which hypothyroidism increases the risk for PCPF are currently speculative and deserve further research. It may be beneficial to explore the strength of association between these biological factors to better inform targeted therapeutic interventions.

Limitations of the study should be acknowledged. Available laboratory values of pre-COVID-19 and post-COVID-19 thyroid hormones (thyrotropin, thyroxine, and triiodothyronine) were collected for patients diagnosed with hypothyroidism in both groups, however, the missing data rate was significant (43–95%), impacting our ability to characterize this relationship. Although the diagnosis of hypothyroidism was found to be significantly associated with PCPF, this relationship was not confirmed using limited available laboratory values.

As suggested by our analysis, hypothyroidism is associated with an increased risk of PCPF anywhere from 1.23 to 3.34 times more than patients without hypothyroidism. The relationship between other notable factors, such as asthma, and PCPF have been explored in the literature and reassessed here. More research is needed to confirm these findings.