Antithyroid Drug Therapy And Cardiovascular Morbidity
Overview
The impact of thyroid function test (TFT) variability and its link to adverse health outcomes in antithyroid drug (ATD) therapy remains an unexplored area. This study aimed to investigate the relationship between TFT variability and cardiovascular morbidity during ATD therapy. Conducted at Tampere University Hospital, a retrospective cohort study followed 394 hyperthyroid patients undergoing ATD therapy from March 2016 to December 2018, over a median period of 1.5 years. The study calculated coefficients of variation (CVs) for follow-up measurements of thyroid-stimulating hormone (TSH), free thyroxine (fT4), and free triiodothyronine (fT3). Logistic regression analyses were employed to assess associations between TFT variability, baseline clinical factors, and hospital visits related to cardiovascular disease (CVD). The results showed that independent risk factors for cardiovascular morbidity included age, male gender, and fT4-CV. Conversely, baseline positive thyrotropin receptor antibodies (TRAbs) were linked to reduced cardiovascular morbidity. In patients with baseline TRAb positivity, fT4-CV was associated with cardiovascular morbidity. Notably, increased fT4 variability during ATD therapy was connected to heightened cardiovascular morbidity. While positive TRAbs correlated with lower cardiovascular morbidity compared to hyperthyroidism with negative autoantibodies, the variability of fT4 remained tied to cardiovascular morbidity even among patients with positive TRAbs.
Introduction
The trajectory of hyperthyroidism’s clinical course exhibits substantial diversity, contingent upon factors encompassing disease origin, severity, patient attributes, and treatment responses. The predominant cause of hyperthyroidism, affecting 70% to 80% in iodine-replete regions, is Graves’ disease (GD), an autoimmune disorder typified by antibodies targeting the thyrotropin receptor. Concurrently, toxic multinodular goiter and solitary toxic adenoma, defined by nodular autonomy, stand as prevalent etiologies.
Hyperthyroidism treatment options encompass long-term antithyroid drug (ATD) therapy, radioactive iodine (RAI) treatment, and surgical intervention. In Europe, antithyroid drug therapy represents the first-line approach for GD, with RAI or surgery advised if hyperthyroidism reoccurs. Hyperthyroidism recurrence after ATD therapy (40% to 60% rate) amplifies the risk of associated complications, notably cardiovascular issues, given the influential role of thyroid hormones in cardiovascular regulation. Cardiovascular manifestations during hyperthyroidism are frequent, and elevated cardiovascular morbidity persists for years post achieving normal thyroid function.
Recent evidence underscores the significance of promptly addressing hyperthyroidism and considering early definitive treatments like RAI or surgery, especially if ATD-induced remission appears unlikely, as inadequate biochemical hyperthyroidism control heightens mortality risk. Swift restoration of euthyroidism enhances patient survival, regardless of the chosen treatment method. However, anticipating antithyroid drug therapy response is complex; a noteworthy proportion of patients ultimately necessitate lifelong thyroid hormone replacement, irrespective of the primary treatment choice.
Diverse studies have identified factors predicting hyperthyroidism recurrence post antithyroid drug therapy for GD, yet the response to ongoing antithyroid drug therapy, particularly in terms of thyroid hormone variability’s clinical implications, remains understudied. This study aims to assess the correlation between thyroid function test (TFT) variability and cardiovascular morbidity risk following antithyroid drug therapy initiation. Additionally, it seeks to uncover baseline clinical factors linked to elevated TFT variability during antithyroid drug treatment.
Method
In this retrospective analysis, the study encompassed patients newly diagnosed with hyperthyroidism and treated using first-line antithyroid drug (ATD) therapy at Tampere University Hospital’s endocrinology outpatient clinic. The study spanned from March 2016 to December 2018, with follow-up continuing until March 2019. The hospital predominantly manages hyperthyroid patients within the region, guided by regional clinical directives that direct primary care providers to refer adult patients (age ≥ 15 years) with new hyperthyroid diagnoses to Tampere University Hospital’s endocrinology clinic. The total population under observation was 442,960 by the end of 2018.
Initially, a cohort was established, involving patients whose first outpatient visit to the endocrinology clinic was due to diagnosis codes E05 or H06.2*E05.9 according to the Tenth Revision of the International Classification of Diseases (ICD-10). Subsequently, 73 patients were excluded due to a lack of available thyroid function test (TFT) results over a minimum 3-month follow-up period post their initial visit. An additional 18 patients were excluded due to not requiring hyperthyroidism treatment or technical issues. Upon excluding 132 patients treated with first-line radioactive iodine (RAI) and 23 patients treated with first-line surgery, the study cohort comprised 394 patients.
Data about the study cohort and their treatment-related clinical details were sourced from the endocrinology clinic’s hospital patient record system (Endo registry), an electronic tool designed to support clinical practice. The Endo registry data encompassed patients’ initial endocrinology clinic visit information and subsequent updates about potential second-line treatments involving RAI or thyroid surgery. Cardiovascular morbidity analysis centered on hospital visits to Tampere University Hospital due to cardiovascular diseases (CVD) between the patients’ first endocrinology clinic visit and the follow-up period ending on March 31, 2019. The dataset concerning CVD-related inpatient and outpatient hospital visits was collected from the hospital’s data administration services, encompassing remote interactions as well. Cardiovascular diagnoses (ICD-10 codes I10-99) were categorized into nine subgroups, and hospital visits were included in the evaluation if the primary diagnosis or one of the first two secondary diagnoses corresponded to the predefined subgroups.
Thyroid Tests
Thyroid function tests (TFTs) were subjected to analysis at Fimlab Laboratories. Plasma concentrations of thyroid-stimulating hormone (TSH), free thyroxine (fT4), free triiodothyronine (fT3), and thyroid peroxidase antibody (TPOAb) were assessed using an electro-chemiluminescence immunoassay (Elecsys Cobas e immunoassay analyzer, ECLIA; Roche Diagnostics). Thyrotropin receptor antibodies (TRAbs) were measured through a fluoro enzyme immunoassay (EliA anti-TSH-R method, Phadia AB). For patients above 20 years, reference ranges were established: TSH (0.27–4.2 mU/L), fT4 (11.0–22.0 pmol/L), and fT3 (3.1–6.8 pmol/L). For patients aged 15 to 20, the reference ranges were TSH (0.5–4.3 mU/L), fT4 (12.6–21.0 pmol/L), and fT3 (3.9–7.7 pmol/L). TRAb measurements’ reference range was 0–1 IU/L until February 14, 2018, and subsequently adjusted to 1–2.9 IU/L. The reference range for TPOAb measurements was less than 34 kU/L.
It’s pertinent to note that, in accordance with local clinical practice, antithyroid drug (ATD) therapy is initiated concurrently with hyperthyroidism diagnosis in primary healthcare settings. The first visit to the endocrinology clinic is scheduled 3–6 weeks after antithyroid drug therapy commencement. Therefore, the thyroid function tests conducted during the initial endocrinology clinic visit reflect the preliminary response to antithyroid drug therapy, typically initiated with a daily dose of 30 mg carbimazole 4–6 weeks before the first visit. Dosage adjustments were tailored to each patient based on disease severity, administered by trained nurses using a standardized protocol guided by TFT results and the prior 4–6 weeks’ antithyroid drug dosage.
This study adhered to the principles outlined in the Declaration of Helsinki. The ethics committee of the Pirkanmaa Hospital District granted approval for the study protocol (study number R18099).
Statistical Analysis
The statistical analyses were conducted using IBM SPSS Statistics version 27.0. A p-value of <0.05 was set for statistical significance. Continuous variables were compared using the Mann–Whitney U test, while categorical variables were assessed using the χ² test. To gauge the variability of thyroid hormone measurements between visits, the coefficient of variation (CV) was employed as a statistical metric. The CV, presented as a percentage of the mean’s standard deviation (SD), signifies the dispersion of measurements around the mean. This dimensionless measure enables the comparison of measurements with different reference ranges, analogous to its application in diabetology for assessing glycemic variability.
This study illustrates three cases within the cohort with the same mean follow-up fT4 levels but distinct fT4-CVs. This divergence in the course of hyperthyroidism becomes evident when observing fT4 variability instead of mean fT4 values. Calculating the CVs of TSH, fT4, and fT3 for each patient involved dividing the SD of the measurements by the mean, followed by a correction factor (√[n/(n − 1)]) accounting for the impact of the number of measurements (n) on variability. All thyroid hormone measurements from 2 months prior to the first endocrinology clinic visit until March 31, 2019 were included.
Median values of TFTs during follow-up were calculated for comparison purposes. Binary logistic regression analyses were conducted to evaluate associations between baseline clinical factors, thyroid hormone variability, and cardiovascular morbidity. The analysis was reiterated considering only patients with a baseline positive TRAb measurement, focusing on GD patients to eliminate potential etiology-related effects on hormone variability or cardiovascular morbidity. Covariates for cardiovascular morbidity analysis encompassed age, gender, body mass index, smoking, pre-existing hypercholesterolemia, diabetes or hypertension, second-line treatment with RAI or surgery, baseline TRAb positivity, and either the median or CV of TSH, fT4, or fT3 values during follow-up.
Separate evaluation of baseline clinical factors in relation to median and CV of TFTs was performed to identify factors linked to TFT variability during antithyroid drug therapy.
Result
Among the 549 patients referred to the endocrinology clinic for hyperthyroidism, a majority of 72% (n = 394) underwent primary treatment with antithyroid drug (ATD) therapy, forming the basis of this analysis. Over a follow-up median duration of 1.5 years (interquartile range [IQR]: 0.8 – 2.0), 23% (n = 92) necessitated second-line treatment. Among these, 19% (n = 74) received radioactive iodine (RAI) treatment and 5% (n = 18) opted for surgery. Of those requiring a second-line treatment, 64% (n = 52) were positive for thyrotropin receptor antibodies (TRAbs), with 36% (n = 29) being TRAb negative (p = .011). Notably, the frequency of second-line treatments, whether RAI or surgery, did not significantly differ between patients with cardiovascular morbidity (30%, n = 17) and those without (22%, n = 75) (p = .121). The median follow-up duration for patients with cardiovascular morbidity was 1.6 years (IQR: 1.0–2.1) and 1.4 years (IQR: 0.8–2.0) for patients without (p = .270).
With a median age of 46 years (IQR 34–64), the majority of patients (78%, n = 308) were female. Positive TRAbs were observed in 75% (n = 264) of patients, while positive thyroid peroxidase antibodies (TPOAbs) were noted in 57% (n = 202). During the follow-up, 14% of patients (n = 57) experienced hospital visits associated with cardiovascular disease (CVD). In total, these 57 patients accounted for 340 CVD-associated hospital visits, with a median of 4 visits per patient (IQR: 2–6). Of these visits, 18% (n = 61) led to inpatient admissions. The median interval between the first endocrinology clinic visit and the initial CVD-associated hospital visit was 84 days (IQR: 14–202).
Differences emerged between patients with and without cardiovascular morbidity. Those with cardiovascular issues were typically older, more frequently male, had higher body mass index and systolic blood pressure, and were more prone to hypertension, hypercholesterolemia, and diabetes. Conversely, they displayed fewer instances of eye symptoms, family history of thyroid disease, or TRAb positivity.
A total of 3860 TSH measurements, 3851 fT4 measurements, and 3649 fT3 measurements were conducted within the study’s follow-up. The median number of TFT measurements was 9 (IQR: 5–12) among patients with cardiovascular morbidity and 9 (IQR: 6–13) among those without (p = .498). No substantial differences in medians or CVs of TSH, fT4, and fT3 measurements were observed between patients with and without cardiovascular morbidity.
Univariable analyses for the complete study cohort demonstrated associations between cardiovascular morbidity and factors including age, male gender, hypercholesterolemia, diabetes, hypertension, and baseline TRAbs. In multivariable analyses, significant associations were observed between cardiovascular morbidity and age, male gender, positive TRAbs, and fT4-CV. Similarly, when evaluating cardiovascular morbidity due to arrhythmias separately, age, male gender, positive TRAbs, and potentially fT4-CV (p = .06) showed associations.
Furthermore, analyzing fT4 median in multivariable analyses revealed that age, male gender, and positive TRAbs were significantly linked to cardiovascular morbidity, while fT4 median itself was not. For arrhythmias, age, male gender, fT4 median, and positive TRAbs exhibited associations. When multivariable analyses were conducted specifically on patients with positive TRAbs, age and fT4-CV displayed links to cardiovascular morbidity, while fT4 median did not. Notably, the medians or CVs of TSH or fT3 measurements were not linked to cardiovascular morbidity in multivariable analyses, encompassing the entire cohort or only patients with positive TRAbs.
Exclusion of patients with isolated hypertension (n = 2) or coronary artery disease (n = 3) as the sole cardiovascular diagnosis maintained the unchanged association of age, male gender, fT4-CV, and TRAbs with cardiovascular morbidity. Additionally, male patients and those with diabetes exhibited slightly higher fT4 medians compared to female patients and those without diabetes (p = .048 and .003, respectively). Patients with a family history of thyroid disease had slightly lower fT4 medians compared to those without (p = .028). Furthermore, patients with positive TRAbs or TPOAbs showed higher fT4-CV in contrast to those with negative TRAbs or TPOAbs (p = .002 and .024, respectively).
Conclusion
In this study, we uncovered a significant link between higher variability of free thyroxine (fT4) levels and the initiation of antithyroid drug (ATD) therapy for patients with autoantibody-related hyperthyroidism, compared to those with hyperthyroidism lacking autoantibodies. Independent risk factors for cardiovascular morbidity were identified as age, male gender, and fT4 variability, whereas thyrotropin receptor antibody (TRAb) positivity seemed to provide protection. However, even among patients with positive TRAbs, the variability of fT4 remained associated with cardiovascular morbidity. This suggests that fT4 variability might be a risk factor for cardiovascular complications across all hyperthyroid patients.
Previous studies have highlighted the significance of maintaining effective control over thyroid hormone levels during antithyroid drug therapy. This ensures optimal outcomes and avoids both persistent hyperthyroidism and overtreatment-induced hypothyroidism, which can each lead to cardiovascular abnormalities. The conventional reliance on median or mean thyroid hormone values alone may not fully capture the dynamic biochemical burden on the cardiovascular system, as thyroid hormones can oscillate between hyperthyroid and hypothyroid states during treatments.
Among patients in this study, 23% received second-line treatment within a relatively brief follow-up of 1.5 years. This finding likely underscores the challenge of hyperthyroidism management during antithyroid drug therapy rather than indicating relapse after complete antithyroid drug treatment. A significant thyroid hormone variability during antithyroid drug therapy could reflect inadequate response to treatment or arise due to the need for various treatment options, which can induce temporary hypothyroidism.
Interestingly, positive TRAbs appeared to be associated with reduced cardiovascular morbidity, potentially due to the underlying etiology of thyrotoxicosis. Notably, this aligns with the broader literature, where fT4 variability, age, and negative TRAbs have been independently tied to cardiovascular morbidity. This study primarily focused on cardiovascular complications related to arrhythmias, especially atrial fibrillation, which is a known risk associated with hyperthyroidism.
Although the study contributes valuable insights, certain limitations exist. The relatively short follow-up period and limited study cohort size might limit generalizability. Comprehensive information on pre-existing cardiovascular morbidity was lacking, and primary healthcare-related data were not available. Therefore, while causality cannot be inferred from this study, it underscores the necessity of effective thyroid hormone control in antithyroid drug therapy and early definitive treatment for hyperthyroidism when needed. Further research is essential to validate these findings and better comprehend the clinical implications of thyroid hormone variability in connection with various health outcomes, including quality of life and mortality.
Oncology Related Tools
Other
Latest Research
