High Prevalence of Iatrogenic Hyperthyroidism in Elderly Patients with Atrial Fibrillation in an Anticoagulation Clinic

Abstract

In elderly patients with established atrial fibrillation (AF) who are receiving thyroid replacement, regular testing for thyroid function is often not performed, placing the patient at risk for iatrogenic hyperthyroidism. Of 215 patients followed in an anticoagulation clinic, 41 were receiving thyroid replacement and 15 of these were found to have hyperthyroidism. Eight had documented AF coincident with abnormal thyroid function. In addition, only 22 patients on thyroid replacement had an annual TSH. In conclusion, iatrogenic hyperthyroidism may frequently be missed in AF patients because of inadequate monitoring of serum TSH.

Thyroid replacement is common in elderly patients with AF followed in an anticoagulation clinic. Laboratory evidence of hyperthyroidism occurred in 37%, usually in patients with higher doses of thyroid replacement, and often associated with AF. The frequency of iatrogenic hyperthyroidism may be underestimated in patients with AF since many patients who receive thyroid replacement therapy are not monitored regularly with serum TSH.

Introduction

Atrial fibrillation (AF) is associated with hyperthyroidism^1–2, and thyroid function tests are routinely obtained in the initial evaluation of the patient with AF. However, little information is available concerning the frequency of thyroid function testing in patients with established AF who also receive thyroid replacement. A recent review of drug-induced AF does not include thyroid replacement as a medication associated with AF³.

Two patients with a history of AF and iatrogenic hyperthyroidism were evaluated in an outpatient Cardiology clinic for recurrent AF. A review of the literature found only one case report of this condition⁴. The purpose of this manuscript is to ascertain the frequency of thyroid replacement therapy in patients with established AF and to determine the prevalence of iatrogenic hyperthyroidism in these patients. In addition, since this condition is potentially preventable with regular monitoring using tests, such as thyroid stimulating hormone (TSH) level, the frequency of appropriate follow-up thyroid testing as recommended by the American Association of Clinical Endocrinologists (AACE) medical guidelines for the evaluation and treatment of hyperthyroidism and hypothyroidism was evaluated.

Case Presentations

Case 1

An 80-year-old hypertensive Caucasian female with a history of Hashimoto’s thyroiditis and prior partial thyroidectomy received 100mcg of levothyroxine for years prior to presenting to the Emergency Department in May 2009 after several hours of palpitations. There was a history of paroxysmal AF. An ECG confirmed AF with a ventricular response of 130–160 bpm. She denied any drug, tobacco, or alcohol abuse. A complete blood count and comprehensive metabolic panel were normal. A serum TSH was 0.04 mU/L (normal range 0.4–4.0 mU/L). Diltiazem was prescribed and AF converted to sinus rhythm. One month later she presented for outpatient follow-up in the Cardiology clinic. Medications included levothyroxine 100 mcg po daily, diltiazem 120mg po daily, vitamin E, glucosamine, and calcium. On physical exam the rhythm was regular, and an ECG confirmed normal sinus rhythm. Laboratory results at this time revealed a TSH of 0.023 mU/L and a FT4 of 1.48 mg/dL (normal range 0.86–1.94 mg/dL). A diagnosis of paroxysmal AF due to iatrogenic subclinical hyperthyroidism was made. The dose of levothyroxine was reduced to 88 mcg/day and the patient has not had an episode of symptomatic AF in a follow-up period of three months

Case 2

A 47-year-old Caucasian female had AF after the delivery of her first child in 1994. A diagnosis of postpartum thyroiditis was made based on an elevated T4 (13.9ug/dl) and a low TSH (0.12mIU/ml). Antithyroid antibodies were elevated (160u/ml). Propylthiouracil and later I ¹³¹ ablation therapy was prescribed. In 1995 tachycardia developed which terminated with adenosine infusion, otherwise palpitations were infrequent. An admission occurred for AF on March 31, 2009. Medications included digoxin, progesterone, calcium and Armour thyroid 240 mg daily. IV diltiazem and flecainide were prescribed and AF converted to sinus rhythm. A TSH was 0.013uIU/ml. Because palpitations felt regular at onset and because of the history of SVT responsive to adenosine, electrophysiologic studies were performed on April 30, 2009, which revealed no inducible SVT. Frequent palpitations continued and a second opinion was obtained. Upon evaluation in the outpatient Cardiology clinic, the patient admitted to occasional alcohol use (5–10 beers over weekends), but denied any association of alcohol use with AF. She denied any tobacco or illicit drug use. On physical examination, BP was 150/80 mmHg, pulse was 80 bpm and regular. The weight was 78kg. Medications included ASA 81 mg po daily, digoxin 0.375 mg po daily, flecainide 50mg po every 12 hrs, paroxetine 20 mg po daily, progesterone 200 mg po at bedtime, and dessicated thyroid 240 mg po daily. Laboratory studies included a TSH of 0.04 mU/L and a FT4 1.40 ng/dL. The patient was reluctant to decrease the thyroid dosage because of the favorable effects of weight loss. She subsequently reduced the dosage and has not had a clinical recurrence after three months.

Methods

All patients with a diagnosis of paroxysmal, persistent, or permanent AF followed in the Anticoagulation Clinic at the University of Missouri Hospital and Clinics during 2009 were included in this study. To determine the frequency of thyroid replacement, the medication profile was reviewed from years 2007–2009. The duration, dose, and type of thyroid replacement therapy were determined by chart review. All thyroid function tests (TFTs), including TSH and FT4 from 2007–2009 were reviewed. The cardiology clinic notes from the same months that the TSH and FT4 were drawn were also reviewed to determine if the patient was in normal sinus or AF based on ECG interpretation. The medication profile was also reviewed to see if patients were taking amiodarone simultaneously with thyroid replacement therapy. Excess thyroid replacement was defined as a TSH < 0.4mU/L or a high FT4 of >1.94ng/dL. The AACE guidelines specifically recommend that after a patient is started on thyroid replacement therapy, the level of thyroid stimulating hormone, TSH, is followed every six weeks until it is within normal range. After the TSH is normal, a follow-up TSH level is obtained at six months and then annually thereafter although the frequency of subsequent evaluation should be individualized⁵. Chart review was performed to determine the frequency of thyroid function tests in patients on thyroid replacement.

Statistical comparisons between groups with abnormal and normal thyroid function tests (TFTs) were performed with the studentt test utilizing Microsoft Excel. A p value <0.05 was considered significant. The project was approved by the University of Missouri Institutional Review Board.

Results

The study included 215 patients with 122 (56.7%) males and 93 (43.3%) females. The average age was 73.6 ± 10.6 years. Forty-one patients (19%) received thyroid replacement therapy during the two-year evaluation period.

The number of patients who received thyroid hormone replacement totaled 41. The identified patients were divided into two groups (See Table 1). There were 23 patients with normal TFTs in the past two years 15 patients with abnormal TFTs defined as either a low TSH (<0.40mU/L) or a high FT4 (>1.94ng/dL). Three patients did not have evaluation of TFTs. The most common thyroid therapy was levothyroxine (n=39). Two patients received dessicated thyroid 60mg and both were in the group with normal TFTs. Of the 15 patients in the group with abnormal TFTs, eight had documented AF during the same month of the blood test abnormality. Compared to patients with normal TFTs, patients with laboratory evidence of hyperthyroidism had a significantly higher dose of thyroid replacement therapy (p<0.01) and had significantly lower TSH values (p<0.01). There was also a significant difference in the FT4 levels between the two groups (p=0.03). Five patients received concomitant amiodarone, three with normal TFTs and two with abnormal tests.

Table 1.

Patients on Thyroid Replacement (n=41) with Normal and Abnormal TFTs

	Patients with abnormal TFTs	Patients with normal TFTs	P values
N	15	23
Age (Years ± 1 SD)	77.4 ± 11.4	77.7 ± 9.86	0.466
Gender
Males	5 (33.3%)	7 (30.4%)
Females	10 (66.6%)	16 (69.5%)
Average dose of thyroid replacement
Levothyroxine (mcg)	129.1 ± 61.3	72 ± 49.3	<0.01
Dessicated thyroid (mg)	0	60
TSH (normal range 0.4–4.0 mU/L)*	0.33 ± 0.44	2.65 ± 2.00	<0.01
FT4 (normal range 0.86–1.94 mg/dL)†	1.53 ± 0.38	1.09 ± 0.74	0.03
Amiodarone	3	2
Years on Thyroid Replacement	2.36 ± 1.63	2.04 ± 0.96	0.22

^*Average of the lowest values is reported for abnormal patients and average of all values available is reported for normal patients.

^†Average of the highest values is reported for abnormal patients and average of all values availabe is reported for normal patients.

In the group of patients on thyroid replacement, a TSH was performed at least yearly in only 22 of the 41 (54%) patients. Three patients in the study on thyroid replacement did not have any assessment of thyroid function during the two-year study period (See Table 1).

Discussion

This study revealed three important findings:

1. thyroid replacement is common in this group of elderly patients with AF;

2. iatrogenic hyperthyroidism is common, with laboratory evidence of hyperthyroidism occurring in 37% of patients over a two year period, often associated with documented AF; and

3. thyroid function testing was not performed regularly as recommended by published guidelines of the American Association of Clinical Endocrinologists (AACE)⁵.

Low serum TSH^{2, 6–8} and high free T4 values⁹ are associated with AF. The risk of AF has been reported to be five times more common than in patients with normal TSH values⁷. Even patients with a low “normal” TSH value (0.4–1.04mU/L) had a risk of AF nearly twice that of those in the high normal range (2.17–3.98mU/L)². However, these studies either did not report the number of patients on thyroid replacement hormone or excluded them from the analysis. In the Cardiovascular Heart Study involving 3,233 individuals, 271 (12%) patients who did not receive thyroid hormone at baseline subsequently received thyroid replacement in the course of the study. Only seven of these patients developed subclinical hyperthyroidism. However, in this study, the patients who became thyrotoxic after starting thyroid replacement were excluded from analysis of incident AF by thyroid status⁸. To date there is no published literature on the effect of thyroid replacement on recurrences of atrial fibrillation.

Low serum TSH values have been associated with hyperthyroidism in a large group of patients without clinical evidence of hyperthyroidism^{6, 7–8}. However, these studies did not report the frequency of excess thyroid replacement hormone patients on exogenous thyroid hormone. The reported incidence of iatrogenic hyperthyroidism in a relatively healthy population without AF or other comorbidities but receiving thyroid replacement is low. Data from the National Health and Nutrition Examination Survey relied on self-reported hyperthyroidism from 17,353 people aged ≥12 years old selected to represent the geographic and ethnic distribution of the United States¹⁰. Of the 820 people in the total population who self reported thyroid disease and who were on thyroid replacement therapy, 150 (18.3%) patients were thyrotoxic-7.3% clinical and 10.9% subclinical¹⁰. In the Colorado Health Fair study of 25,862 subjects, 1,525 of whom received thyroid replacement, 13 (0.9%) were clinically thyrotoxic whereas 316 (20.7%) were subclinically thyrotoxic¹¹. The comparatively high incidence of iatrogenic hyperthyroidism (37%) in the current study is likely due to the elderly population with established AF.

Amiodarone, frequently prescribed for patients with AF to maintain sinus rhythm, is a known cause of thyroid abnormalities³. The overall use of amiodarone in this study was low and not significantly different in patients with and without thyroid abnormalities. Amiodarone use does not account for most of the causes of iatrogenic hyperthyroidism.

In summary, thyroid replacement was common in this group of elderly patients with AF followed in an anticoagulation clinic. Iatrogenic hyperthyroidism was also common in these patients with laboratory evidence of hyperthyroidism found in 37% of patients over a two year follow-up period. These laboratory abnormalities were usually associated with periods of AF. Finally, thyroid function tests were not performed regularly in many of the patients on thyroid replacement as recommended by published guidelines of the AACE. Greater physician awareness of the association between thyroid excess and AF may improve the care of these patients and prevent AF.

Limitations

The size of our population is small and may be biased since we examined a cardiac population with AF referred to an anticoagulation clinic with a high frequency of patients with symptomatic AF. Thyroid function testing might have been performed by physicians separate form out health care facility.

Conclusion

Both atrial fibrillation and hypothyroidism are common in the elderly and may occur concomitantly. Iatrogenic hyperthyroidism also occurs frequently and may remain asymptomatic in the elderly. The American Association of Clinical Endocrinologists suggests checking serum TSH levels six weeks after changing the dose or beginning therapy with thyroid replacement, then every six months until stabilized, and annually thereafter. Some studies have suggested that less frequent monitoring intervals are acceptable in younger adults but for adults over 65 years of age it is recommended that TSH be checked at least annually because of age-related changes in the volume of distribution, drug-drug interactions affecting thyroid hormone clearance and protein binding, and the diminished physiologic tolerance for slight thyroid hormone excess which accompanies declining health status^12–14. Considering that most patients with AF are older than 65 years (84% of this patient population), a prudent recommendation would be to monitor TSH at least annually for all patients over the age of 65 with a history of AF who receive thyroid replacement.

Biography

Sandeep Kumar Krishnan, MD, former MSMA member, is at the Emory University School of Medicine in Atlanta. Mary L. Dohrmann, MD, Stephen A. Brietzke, MD, David A. Fleming, MD, MSMA member since 1998, and Greg C. Flaker, MD (above photo) are all at the University of Missouri - Columbia School of Medicine and University Hospital and Clinics.

Contact: flakerg@health.missouri.edu