Advertisement

Journal of Endocrinological Investigation

, Volume 42, Issue 11, pp 1273–1283 | Cite as

Safety of long-term antithyroid drug treatment? A systematic review

  • F. AziziEmail author
  • R. Malboosbaf
Review

Abstract

Continued low-dose MMI treatment for longer than 12–18 months may be considered in patients not in remission. However, ATDs are not free from adverse effects. We undertook a systematic review to clarify safety of long-term ATD treatment. Medline and the Cochrane Library for trials published between 1950 and Nov 2018 were systematically searched. We included original studies containing data for long-term (> 18 months) ATD treatment. Two reviewers independently extracted data from included trials and any disagreement was adjudicated by consensus. Of 615 related articles found, 12 fulfilled the criteria. Six articles had data for adults, five for non-adults and one article had data for both groups. The sample sizes ranged between 20 and 249 individuals, and the mean duration of ATD treatment ranged between 2.1 and 14.2 years. Considering all data from 1660 patients treated with ATD for a mean duration of 5.8 years (around 10,000 patient-years), major complications occurred only in 14 patients: 7 severe agranulocytosis, 5 severe liver damage, one ANCA-associated glomerulonephritis and one vasculitis with small cutaneous ulcerations. Minor complications rates were between 2 and 36%, while more complications were in higher doses and in the children. The most reported AE was cutaneous reaction; the other adverse events were elevated liver enzymes, leukocytopenia, arthritis, arthralgia, myalgia, thrombocytopenia, fever, nausea and oral aphthous. Long-term ATD treatment is safe, especially in low dose and in adults, indicating that it should be considered as an earnest alternative treatment for GD.

Keywords

Graves’ disease Long-term therapy Continuous therapy Antithyroid drugs Methimazole Propylthiouracil 

Introduction

Graves’ disease (GD) is an autoimmune disorder in which thyrotropin receptor antibodies (TRAb) stimulate the TSH receptor, causing increased thyroid hormone production and release. Therapy for GD is not straightforward and implicates difficult decision making [1, 2]. There is no specific cure for the disease, and any treatment option can be associated with certain complications [3, 4, 5, 6].

There are three treatment options for GD: antithyroid drugs (ATD), radioactive iodine (RAI), and surgery [7]. In the United States, RAI was the most preferred treatment, although there has been an increasing trend in recent years to use ATD [8]. The latest report shows that majority of physicians in USA prefer ATD as first-line treatment for GD [7]. In Europe, Latin America and Asia, there is higher physician preference for ATD [9, 10, 11, 12, 13, 14], to avoid ablation and to assist some patients to enter prolonged remission.

The major clinical drawback of treatment with ATD is the high rate of relapse of hyperthyroidism when therapy is discontinued [15]. Nevertheless, the convenience of ATD treatment and the fact that it does not irreversibly compromise the thyroid have made long-term low-dose ATD treatment a rational alternative for control of GD [16]. The 2016 American Thyroid Association (ATA) guidelines recommend that “if a patient with GD becomes hyperthyroid after completing a course of MMI, consideration should be given to treatment with RAI or thyroidectomy. Continued low-dose MMI treatment for longer than 12–18 months may be considered in patients not in remission, who prefer this approach [8]”. Leger et al. [17] in a review article proposed that in children with GD continuous ATDs treatment, rather than 2-year treatment cycles, may be more effective for achieving a gradual remission, and the long-term treatment with the lowest ATDs dose resulting in euthyroidism should be offered to all patients, to increase the likelihood of remission. Recently, a systematic review and meta-analysis demonstrated that long-term ATD treatment is effective and safe, especially in adults, indicating that it should be considered as an alternative treatment for GD [18].

However, ATDs are not free from adverse effects, either minor such as rash, gastric intolerance and arthralgia in 5% of patients or major such as agranulocytosis and hepatotoxicity, sometimes life-threatening but rare (< 0.5% of cases) [16]. Since in recent years more attention has been paid to “long-term ATD treatment” as an alternative treatment for GD, in the present study we aimed to review systematically available data to clarify the safety of long-term ATD treatment.

Methods

A systematic review of the published work according to the PRISMA statement for the conduct of meta-analyses [19] was performed to evaluate safety of long-term ATD treatment.

Eligibility criteria

All original studies containing data for long-term (over 18 months) ATD treatment were eligible for inclusion. Exclusion criteria were irrelevant articles (based on screening of titles and abstracts), insufficient data available in the article, duplications of articles and no original data (review articles).

Study identification

Relevant studies were identified by searching Medline (from 1950 to Nov 2018) and the Cochrane Library database (no date restriction) using the search terms of “Graves’ disease and antithyroid drugs”, “Graves’ disease and long-term therapy”, “Graves’ disease and continuous therapy”, “long-term antithyroid drugs”, “long-term methimazole therapy”, “long-term carbimazole therapy”, “long-term propylthiouracil therapy”, “continuous antithyroid therapy”, “continuous methimazole therapy”, “continuous carbimazole therapy” and “continuous propylthiouracil therapy”. The search was done without language restriction. Search was restricted to “clinical trials” and “human” studies. Reference lists from identified articles were manually scanned to identify any other relevant studies.

All articles were reviewed independently by two reviewers (RM and FA) and any disagreement was resolved by consensus. After abstract screening and retrieval of potentially eligible studies, the full-text articles were assessed for eligibility. Duplicate studies were excluded. Prior to review, the reviewers were blinded to names of journals and authors.

Data collection and management

Two reviewers (RM and FA) independently extracted data from included trials and any disagreement was adjudicated by consensus. Published reports were obtained for every study, and standard information was extracted into a spreadsheet. The data sought included time and location of study, age and sex of individuals, number of study subjects, modality of treatment, duration of ATD treatment, complications (major, minor).

In the present study, we defined “long-term treatment” as more than 18 months’ treatment with ATD. Our defined cut point of age for adults was set at 18 years. Based on the substantial between-study clinical variability and heterogeneity of study designs, we elected that data aggregation in a meta-analysis would not be appropriate and decided not to perform pooled statistical analyses.

Risk of bias assessment

Two reviewers (RM and FA) independently assessed methodological quality of the included trials using the Newcastle–Ottawa Quality Assessment Scale [20] and any disagreement was resolved by consensus. This tool appraises the selection of the study groups, the comparability of the groups and the ascertainment of exposure or outcome. Potential publication bias was assessed with funnel plots of the Arcsin prevalence versus its standard error [21]. Agreement between reviewers was estimated with Cohen’s test [22].

Results

Study selection and characteristics

The literature search yielded 615 articles, of which 61 were reviewed in full text (Fig. 1). Table 1 summarizes characteristics of the included studies; of the 12 articles that met the inclusion criteria, three of articles compared long-term ATD vs. RAI and the nine others reported data for long-term ATD treatment. Two articles were prospective cohort, four articles were randomized controlled trial and remaining six articles were retrospective. Six articles had data for adults, five for non-adults and one article had data for both groups.
Fig. 1

Flow chart for the systematic review

Table 1

Characteristics of the studies included in the systematic review

First author [ref. no]

Location

Year published

Study design

Method

Number of participants (F/M)

Age range (years)

Mean duration of treatment (years)

Reported parameters

Yasuda et al. [23]

Japan

2017

Retrospective

Long-term MMI (low dose vs. high dose)

56 (45/11)

2–15

6.9

Adverse events

Orgiazzi [16]

France

2015

Retrospective

Long-term ATD vs. RAI

238

18–60

6.3

Adverse events, TFT, GOa activity and severity, QoLb and body weight

Azizi et al. [24]

Iran

2012

Randomized Controlled Trial

Long-term ATD vs. RAI

132 (103/29)

40–65

14.2

Adverse events, TFT, lipid profiles, BMD and cardiac echocardiography and neuropsychologic tests

Leger et al. [25]

France

2012

Prospective cohort

Long-term ATD

154 (118/36)

8–18

4

Adverse events, TFT, pubertal stages, goiter and BMI

Laurberg et al. [26]

Denmark

2011

Retrospective

Long-term ATD

108 (84/24)

48–62

6.7

Adverse events

Sato et al. [27]

Japan

2011

Retrospective

Long-term ATD (MMI vs. PTU)

133 (114/19)

9–16

5

Adverse events

Chen et al. [28]

China

2009

Randomized Controlled Trial

Long-term ATD vs. RAI

460 (314/146)

8–64

8.2

Adverse events

Kaguelidou et al. [29]

France

2008

Prospective cohort

Long-term ATD

154 (118/36)

8–18

2.1

Adverse events

Mazza et al. [1]

Italy

2008

Retrospective

Long-term ATD

384 (313/71)

20–84

2.7

Adverse events

Azizi et al. [30]

Iran

2005

Randomized controlled trial

Long-term ATD vs. RAI

85 (69/16)

40–60

10.1

Adverse events, TFT, lipid profiles, BMD, cardiac echocardiography and neuropsychologic tests

Barrio et al. [31]

Spain

2005

Retrospective

Long-term ATD

20 (18/2)

8–18

6.3

Adverse events

Mashio et al. [32]

Japan

1997

Randomized controlled trial

Long-term ATD

112 (87/25)

20–60

2.3

Adverse events

aGraves orbitopathy

bQuality of life

The sample sizes ranged between 20 and 249 individuals, and the mean duration of ATD treatment ranged between 2.1 and 14.2 years.

Study quality

The risk of bias of included studies was low, mainly due to lack of blinding when assessing the outcomes (Table 2). Formal statistical testing with funnel plots showed no evidence of publication bias. Inter-reviewer’s agreement was “excellent” (Cohen’s test κ = 0.96).
Table 2

Risk of bias assessment in selected papers for this meta-analysis

First author (year)

Representativeness of the exposed cohort

Selection of the non exposed cohort

Ascertainment of exposure

Demonstration that outcome of interest was not present at start of study

Comparability of cohorts on the basis of the design or analysis

Assessment of outcome

Was follow-up long enough for outcomes to occur?

Adequacy of follow-up of cohorts

Yasuda (2017)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events

Record linkage

Yes

Complete follow-up

Orgiazzi (2015)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events, TFT, GOa activity and severity, QoLb and body weight and complications

Record linkage

Yes

Complete follow-up

Azizi (2012)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events, TFT, lipid profiles, BMD and cardiac echocardiography and neuropsychologic tests smoking, lipid profiles, BMD and echocardiography

Record linkage

Yes

Lost to follow-up unlikely to introduce biasc

Leger (2012)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events, TFT, pubertal stages, goiter and BMI

Record linkage

Yes

Lost to follow-up unlikely to introduce bias

Laurberg (2011)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events

Record linkage

Yes

Complete follow-up

Sato (2011)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events

Record linkage

Yes

Lost to follow-up unlikely to introduce bias

Chen (2009)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events

Record linkage

Yes

Lost to follow-up unlikely to introduce bias

Kaguelidou (2008)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events

Record linkage

Yes

Lost to follow-up unlikely to introduce bias

Mazza (2008)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events

Record linkage

Yes

Lost to follow-up unlikely to introduce bias

Azizi (2005)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events, TFT, lipid profiles, BMD, cardiac echocardiography and neuropsychologic tests

Record linkage

Yes

Lost to follow-up unlikely to introduce bias

Barrio (2005)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events

Record linkage

Yes

Complete follow-up

Mashio (1997)

Truly representative

Same community

Secure record

Yes

Study controls for adverse events

Record linkage

Yes

Lost to follow-up unlikely to introduce bias

aGraves orbitopathy

bQuality of life

cLost to follow-up is known to introduce bias. However, it has been proposed that for values < 20% the amount of bias could be negligible (Ref. [19])

Systematic review

Three of articles compared long-term ATD vs. RAI in a randomized controlled trial and the nine others reported data for long-term ATD treatment, with different study designs (6 retrospectives, 2 prospective cohorts and one randomized controlled trial).

Long-term ATD vs. RAI

Azizi F et al. [30] in an RCT demonstrated long-term continuous treatment of hyperthyroidism with MMI is safe. 85 Patients older than 40 years of age who relapsed after an 18-month course of treatment with methimazole were randomized into two groups: long-term continuous treatment MMI or radioactive iodine. During 10 years of successive MMI treatment except for minor allergic symptoms, no serious complications occurred. The mean maintenance dose of MMI was 4.9 ± 1.3 mg/day.

Chen et al. [28] conducted an RCT on 460 Chinese patients with hyperthyroidism. Patients over 8 years of age without any previous treatment were enrolled in a 9-year prospective, randomized, open-label blinded endpoint study and randomly assigned to receive either RAI or ATD. Among 230 patients treated with ATD, adverse events occurred in 28 patients (12%): 16 cases of cutaneous reactions, five severe liver damage and seven severe agranulocytosis. The incidence of ophthalmopathy in both therapy groups was not significantly different.

Azizi et al. [24] in an RCT concluded that long-term MMI treatment was superior to RAI therapy when cost, mood, cognition, cardiac function, lipid profiles and occurrence of thyroid dysfunction were compared. During a mean of 14.2 years continuous MMI treatment, with the exception of minor allergic symptoms, no serious adverse events occurred. All patients received maintenance doses of 2.5–10 mg daily MMI from the third month afterward.

Long-term ATD treatment

Mashio et al. [32] in an RCT show that there is no difference in the clinical and immunological course or in the long-term remission rate of Graves’ hyperthyroidism when the MMI is initiated with either a small single daily dose (group S) (15 mg) or the conventional regimen (group C) (10 mg 3 times daily); however, adverse effects occurred less frequently in the small single daily dose regimen. One hundred and twelve patients with untreated Graves’ hyperthyroidism were randomly divided into two groups. The MMI doses were gradually reduced to a maintenance level (5 mg/day) in both groups with treatment duration of around 28 months. Minor adverse reactions to MMI were less frequent in the small single daily dose than in the conventional regimen (13% vs. 24%). MMI was changed to propylthiouracil (PTU) in six patients (4 in group C, 2 in group S), due to adverse reactions.

Barrio et al. [31] reported a retrospective study of 20 children and adolescents with Graves’ disease over a period of 20 years (around 6 years of MMI treatment and around 14 years of follow-up). Minor side effects were observed in six patients (30%). The most common toxic effect was skin rash (six patients); other minor reactions included arthralgia and myalgia, and granulocytopenia and oral aphthous; none of them required stopping medication.

Mazza et al. [1] in a retrospective study showed that long-term treatment with low doses of MMI is safe. In 384 patients with median treatment duration of 33 months and maintenance dose of 2.5–5 mg/day, no case of agranulocytosis was recorded. Eight cases (2%) of urticaria were observed in the first 2 months of treatment; in 6 of them, MMI was substituted with propylthiouracil or radioiodine. In the other two patients, urticaria ceased with the use of antihistaminic agents.

Kaguelidou et al. [29] in a prospective cohort of 154 children with median treatment duration of 25 months reported that nine patients (6%) presented adverse reactions on ATD (urticaria: n = 6; arthralgia: n = 2; leukocytopenia: n = 1; thrombocytopenia: n = 1), with a change of drug treatment being necessary for five patients and radioactive iodine being administered to one patient. The median maintenance dose of carbimazole was 0.34 mg/kg/day.

Sato et al. [27] in a retrospective study concluded that PTU (especially in high dose) may not be suitable for initial use in children and adolescents with GD; lower initial dosages of MMI may be favorable for children and adolescents with mild or moderate GD to avoid adverse effects. The subjects were 133 children and adolescents with newly diagnosed GD and the mean duration of ATD treatment was 5 ± 2.1 years. The subjects were divided into two groups by type of ATD: group M (MMI) and group P (PTU). Furthermore, the subjects were divided into four subgroups by initial dose of ATD: group M1 (MMI: < 0.75 mg/kg), group M2 (MMI: ≥ 0.75 mg/kg), group P1 (PTU: < 7.5 mg/kg), and group P2 (PTU: ≥ 7.5 mg/kg). No serious adverse reaction was observed. Minor adverse effects occurred in 16 patients of group M (25.0%) and 22 in group P (31.9%), with no significant difference between the two groups. Among the subgroups, 10 episodes (skin eruption = 4, liver dysfunction = 3, neutropenia = 1, arthritis = 1, mild fever = 1) in seven patients of group M1, 11 (skin eruption = 9, liver dysfunction = 1, mild fever = 1) in nine patients of group M2, three (skin eruption = 1, liver dysfunction = 2) in two patients of group P1, and 24 (skin eruption = 5, liver dysfunction = 11, neutropenia = 3, arthritis = 2, urticaria = 1, itching = 1, nausea = 1) in 20 patients of group P2 were observed. The incidence of liver dysfunction in group P was significantly higher than that in group M (18.8% vs. 6.3%, p < 0.05). The frequency of change in ATD due to adverse effects in group P was significantly higher than that in group M (27.5% vs. 9.4%, p < 0.01). The incidence of minor adverse effects in group P2 (44.4%) was significantly higher than that in group M1 (20.6%, p < 0.05) and group P1 (8.3%, p < 0.01).

Laurberg et al. [26] in a retrospective study in a group of 108 patients suffering from Graves’ disease with median duration of ATD treatment of 80 months concluded that prolonged therapy with low-dose ATD keeps the majority of patients with severe Graves’ orbitopathy and hyperthyroidism stable and euthyroid. The median maintenance dose for MMI was 5 mg/day and for PTU was 200 mg/day. Seven of the 108 patients had an early shift of ATD from MMI to PTU. This was mostly due to suspicion of cutaneous reactions occurring within the first months of therapy (n = 5). One patient had slightly elevated circulating liver enzymes and one had moderate arthralgia, both with no clear relation to methimazole therapy. The only serious side effect to ATD therapy observed in the present study occurred in a 48-year-old woman. She had a cutaneous rash after 3 weeks of methimazole therapy and this was changed to PTU. After 6 years of uncomplicated PTU therapy, she developed vasculitis with small cutaneous ulcerations confined to the nose and the ear lobes.

Leger et al. [25] in a prospective multicenter cohort study assessed the effect of the duration of long-term carbimazole therapy on GD remission in children. Participants included 154 children (< 18 years) newly diagnosed with GD. The intention was to treat for three consecutive cycles of 2 years in cases of hyperthyroidism relapse after the discontinuation of treatment at the end of a cycle. The median duration of follow-up was 10.4 (9.0–12.1) years. Three serious adverse effects of ATD treatment were reported, requiring definitive treatment 2 months and 3 and 7 years after the start of ATD treatment, due to allergic reaction (n = 1), neutropenia (n = 1), and arthralgia (n = 1), respectively.

Orgiazzi [16] in a retrospective study compared treatment with prolonged low-dose MMI and treatment with RAI plus levothyroxine in 238 patients in whom GD relapsed after a conventional course of ATD therapy. The mean maintenance dose of MMI was 3.98 ± 1.7 mg/day (range 2.5–7.5). Euthyroidism was maintained in many more patients in the low-dose MMI than in the RAI. There were no reports of major adverse effects of MMI.

Yasuda et al. [23] in a retrospective study investigated the relationship between MMI doses and adverse events (AEs) in 56 GD patients with onset age of < 15 years. The mean duration of ATD treatment was 6.9 years and the mean follow-up duration was 9.1 years. Patients initially treated with MMI were divided into low-dose (< 0.7 mg/kg/day) and high-dose (≥ 0.7 mg/kg/day) MMI groups. ATD-induced AEs were identified in 20 patients (35.7%). In two of these patients, one with a rash and another with ANCA-associated glomerulonephritis, AE could not be controlled by switching to PTU, and a thyroidectomy had to be performed subsequently. Eighty percent of the AEs developed within 3 months of treatment initiation, and the longest duration from treatment initiation to AE development was 11 years. Only three of AEs occurred after first year of treatment, one case of rash at 5 years and two cases of neutropenia at 9 and 11 years of treatment. AEs were less frequent in the low-dose group compared to the high-dose group (20% vs. 50%, p = 0.031). Of the four patients in the low-dose MMI group who developed AEs, three had neutropenia and one had a rash. 16 patients in the high-dose MMI group developed AEs, 10 patients had a rash, 6 had arthritis, two had neutropenia, 2 had mild hepatotoxicity, and 1 patient had a high serum CPK level. Thus, a greater variety of clinical symptoms was observed in the high-dose MMI group compared to the low-dose group. Three female patients delivered four healthy babies, including two babies born to a mother treated with MMI during early pregnancy.

Considering all above data from 1660 patients treated with ATDs for a mean duration of 5.8 years (around 10,000 patient-years), major complications occurred only in 14 patients: 7 severe agranulocytosis, 5 severe liver damage, one ANCA-associated glomerulonephritis and one vasculitis with small cutaneous ulcerations.

Minor complications rates were between 2 and 36%, while more complications were in higher doses and in the children. The most reported AEs were cutaneous reaction (74 cases); the other adverse events were elevated liver enzymes (20 cases), leukocytopenia (11 cases), arthritis (9 cases), arthralgia (5 cases), myalgia (2 cases), thrombocytopenia (2 cases), fever (2 cases), nausea (1 case) and oral aphthous (1 case). In 19 patients, ATD was stopped because of adverse events. Only in 5 patients, AEs were reported after first year of treatment.

Discussion

To the best of our knowledge, this is the most comprehensive systematic review to date clarifying adverse events of long-term ATD treatment for Graves’ disease. This study suggests that long-term ATD treatment is safe.

Results of this study show that long-term ATD treatment has few adverse events, the majority of these adverse events occur in first months of treatment and the adverse events are more frequent in higher dose and in children. So, once we are standing at the end of a custom course of ATD treatment (12–18 months) we have passed the hazard window and afterward the adverse events will be very rare. Therefore, while we are ongoing on low-dose ATD we can keep the patient euthyroidism safely.

Therapy for Graves’ disease implicates difficult decision making. There is no perfect recommended treatment for all patients, and any therapeutic option is associated with certain limitations and complications. There are three treatment modalities for GD: antithyroid drugs (ATD), radioactive iodine (RAI), and surgery [7]. ATD can restore euthyroidism, although the disease may recur after a course of ATD. In contrast, RAI treatment as well as surgical thyroidectomy leads to hypothyroidism in most cases. Other drawbacks of these modes of treatment include the potential the following uncommon events: surgical complications and radiation exposure [16]. However, ATDs are not free from adverse effects and minor complications such as rash, gastric intolerance, and arthralgia are observed in up to 19% of patients, while major complications, such as agranulocytosis and hepatotoxicity, occur rarely (< 0.5% of cases) but could be life-threatening [33, 34].

In spite of over 70 years of experience with the use of ATD and RAI for the treatment of Graves’ disease, the rationale for choice of therapy is often vague [35, 36]. Very often the choice has to be made between prolonged treatment with ATD on the one hand and lifetime therapy with thyroid hormones for thyroid failure on the other [37]. In the United States, RAI has been the preferred therapy although in recent years there is a trend suggesting an increased use of ATD and a reduction of RAI as primary treatment [8]. A 2011 survey of clinical endocrinologists showed that 59.7% of respondents from the United States selected RAI as primary therapy for uncomplicated cases of GD, compared to 69% in a similar survey performed 20 years earlier [10]. A recent study by Brito et al. [7] revealed that the most common “initial” treatment of GD in commercially insured patients in the US today is ATD. Increased use of ATD started in 2005 at the expense of RAI use. ATD was the most commonly used treatment in this cohort (58%), followed by RAI (35%) and surgery (6%). In Europe, Latin America and Asia, there is higher physician preference for ATD [9, 10, 11, 12, 13, 14], to avoid ablation and to induce prolonged remission in some patients.

The major clinical drawback of treatment with antithyroid drugs is the 20–70% relapse of hyperthyroidism when therapy is discontinued [15, 38]. Although some predictive factors for relapse of hyperthyroidism after treatment have been proposed, none of them are perfect to predict at baseline [39]. However, treatment with ATD might have a beneficial immunosuppressive role, either primarily, to decrease thyroid-specific autoimmunity, or secondarily, by ameliorating the hyperthyroid state, which may restore the dysregulated immune system to normal [40]. It is still a matter of debate whether this effect is due to an immunosuppressive effect of the drug itself, as suggested by some studies [41, 42] or to a reduced production of thyroid antigens which is followed by a reduced immune response [43]. Selenium appears to influence the immune system by unknown mechanisms. There is some evidence supporting a possible use of selenium as an adjuvant measure in selenium-deficient patients treated with antithyroid medications [44]; however, another recent randomized clinical trial performed in selenium-sufficient patients did not show a beneficial effect of selenium [45]. Based on these conflicting results, it appears that selenium supplementation might be offered to patients with Graves’s hyperthyroidism only if selenium deficiency is documented [46, 47]. Considering all these possible mechanisms highlights the importance of maintaining the patients in a euthyroid state for a long period to minimize autoimmune aberration and GD recurrence [40, 48]. This may occasionally require more prolonged use of methimazole (MMI), of which the lowest possible dose should be used to minimize the risk of side effects [40]. Long-term treatment has been proposed by some authors [18, 30, 49], especially in children [25, 29, 50] and patients with Graves’ orbitopathy (GO) [26, 51].

Some studies have shown that long-term MMI is a reasonable alternative approach in selected patients (i.e., in younger patients with mild stable disease on a low-dose of MMI) [30, 49]. Another study by Azizi et al. [24] reported that continuous administration of MMI is safe, without major complications and accompanied by fewer events of subclinical hypothyroidism and dyslipidemia in comparison to patients on levothyroxine substitution for radioiodine-induced hypothyroidism. A recent retrospective analysis compared long-term outcomes of patients who had relapsed after a course of ATD, treated with either RAI and levothyroxine or long-term ATD therapy [52]; patients treated with RAI more often had persistent thyroid eye disease, continuing thyroid dysfunction, and experienced more weight gain compared to those patients receiving long-term ATD treatment. Orgiazzi [16] concluded that the use of low doses of MMI is efficient and safe and offers better outcomes for GO than RAI treatment, and proposed that prolonged use of low doses of MMI may be an alternative for patients whose GD has relapsed, particularly for patients with GO or for patients who refuse a definitive treatment.

A recent meta-analysis [18] demonstrated that long-term ATD treatment is effective and safe, especially in adults, and proposed that it should be considered as an alternative treatment for Graves’ disease. Primary focus of that meta-analysis was effectiveness of long-term ATD treatment and complications were analyzed briefly and in overall (The rate of complications was 19.1%, of which only 1.5% were major complications); only original studies containing data for over 24 months treatment were included in analysis. All included studies were planned for long-term treatment duration and all patients that withdrew from ATD treatment due to side effects or poor compliance were excluded. Six articles were included and the main results were: long-term ATD treatment induced a remission rate of 57% (95% CI = 45–68%), a rate that was higher in adults than in non-adults (61% vs. 53%). The annual remission rate for each year of treatment was 16%. In our present systematic review, we focused on safety of long-term ATD treatment, with specific concern on details of adverse events such as type, time and frequency. Thus, all original studies containing data for long-term (over 18 months) ATD treatment were eligible for inclusion. Twelve articles fulfilled the criteria, data from 1660 patients treated with ATD for a mean duration of 5.8 years (around 10,000 patient-years) and the main results were: major complications occurred only in 14 patients (7 severe agranulocytosis, 5 severe liver damage, one ANCA-associated glomerulonephritis and one vasculitis with small cutaneous ulcerations). Minor complications rates were between 2 and 36%, while more complications were in higher doses and in the children.

Adverse effects (AEs) usually occur during the first 3–6 months of treatment, and major adverse events tend to be associated with high ATD doses [1, 16, 53]. Several reports on MMI-induced AEs have shown that AEs are related to the initial MMI dose [5, 54, 55, 56, 57, 58, 59]. Our systematic review shows that major complications occurred only in 14 patients and merely in 5 patients AEs were reported after first year of treatment.

In one study from Japan [34], a retrospective cohort analysis of over 50,000 GD patients, 55 developed agranulocytosis, 5 of whom had pancytopenia, for an estimated cumulative incidence of 0.3% in 100 days, with a median interval to onset of 69 days; all 50 patients with agranulocytosis alone were successfully treated with granulocyte colony stimulating factor (G-CSF), steroids, or supportive care, but one of the five patients with pancytopenia died; no predictive risk factors for the development of agranulocytosis could be identified. The second study from Japan [60] was based on a national database for adverse drug reactions, which may have included some patients reported earlier in the first study; a total of 754 GD patients who developed ATD-induced hematologic complications were reported, for an estimated incidence of 0.1–0.15%; at the onset of agranulocytosis, the average MMI dose was 25 mg/day and the average PTU dose was 217 mg/day; the average age of patients developing agranulocytosis was slightly older (45 vs. 40 years), 72% developed agranulocytosis within 60 days of starting ATD, and 85% within 90 days; thirty of the events (4%) were fatal. A recent pharmacoepidemiologic study from Taiwan [61] demonstrated that hepatotoxicity incidence rates peaked within 30 days of continuous treatment and tended to decline afterward; the rates were higher in older patients (aged over 65 years) and with higher doses (MMI ≥ 10 mg/day and PTU ≥ 100 mg/day). Similar findings were also recently reported from China [62]; the average doses of ATDs at the onset of hepatotoxicity were around 19 and 212 mg/day for MMI and PTU, respectively, indicating severe hepatotoxicity tended to occur within the first 3 months after the onset of ATD therapy.

Ohye et al. [63] have reported that while most AEs (91.6%) occurred within the first 3 months of ATD treatment, rarely agranulocytosis may develop several months after ATD initiation [64, 65]. Yasuda et al. reported that while 80% of the AEs developed within 3 months of ATD treatment initiation, the longest duration for AE development was noted for MMI-induced agranulocytosis (11.5 years from treatment initiation). These results indicate that the possible development of agranulocytosis should be anticipated regardless of the MMI administration period [23]. He et al. [66] in a retrospective study of 64 hospitalized patients diagnosed with ATD-induced agranulocytosis between 2000 and 2015 demonstrated that agranulocytosis occurred in 81% of patients within the first 3 months after initiation of ATD therapy. Fever and sore throat were the most common symptoms. However, 11 (17.2%) patients were asymptomatic before being diagnosed with agranulocytosis and were diagnosed by once weekly or bi-weekly monitoring of the WBC and granulocyte counts. In addition, 30 (46.9%) patients delayed seeking treatment despite the appearance of symptoms. Balavoine et al. [67] in a review article showed that ANCA-associated vasculitis is a very rare but severe side effect of ATD treatment, it develops usually after months to years of treatment and the main risk factors are pediatric population, taking PTU and the duration of ATD therapy.

Abhayaratna et al. [68] conducted a systematic review and concluded that long-term treatment with low-dose methimazole in Grave’s disease is shown to be safe and effective. Leger et al. [25] in a review article proposed that in children with GD continuous ATDs treatment, rather than 2-year treatment cycles, may be more effective for achieving a gradual remission, and the long-term treatment with the lowest ATDs dose resulting in euthyroidism should be offered to all patients, to increase the likelihood of remission.

If a patient with GD becomes hyperthyroid after completing a course of MMI, the 2016 ATA guidelines recommend that “consideration should be given to treatment with RAI or thyroidectomy. Continued low-dose MMI treatment for longer than 12–18 months may be considered in patients not in remission who prefer this approach”. Howard [69] emphasized in 1967 that his approach to the therapy of thyrotoxicosis is to control the disease, for a lifetime if need be, with an ATD until spontaneous remission occurs. It has been repeatedly noted that thyrotoxicosis may be a self-limited disease and treatment with ATD is often used for limited time durations only [69, 70, 71]. Slingerland et al. [49] declared: “A controllable, self-limited disease seems best treated with a controllable and limitable therapy”.

Despite the limitations of this study, which is based on published data and thus limits the capacity to fully explore factors associated with adverse events, this systematic review benefits from a comprehensive search strategy, no limitations for time and language in our search strategy, independent reviews by two reviewers and no publication bias.

Conclusion

Long-term ATD treatment is safe, especially in low-dose and in adults, indicating that it should be considered as an earnest alternative treatment for GD.

Notes

Acknowledgements

The authors wish to acknowledge Ms. Niloofar Shiva for critical editing of English grammar and syntax of the manuscript.

Compliance with ethical statement

Conflict of interest

The authors declare that they have no conflict of interest.

Research involving human participants and/or animals

This article is a systematic review, so does not contain any studies with human participants or animals performed by any of the authors.

Informed consent

For this type of study, formal consent is not required.

References

  1. 1.
    Mazza E et al (2008) Long-term follow-up of patients with hyperthyroidism due to Graves’ disease treated with methimazole. Comparison of usual treatment schedule with drug discontinuation vs continuous treatment with low methimazole doses: a retrospective study. J Endocrinol Invest 31:866–872PubMedGoogle Scholar
  2. 2.
    Marino M et al (2014) An update on the medical treatment of Graves’ hyperthyroidism. J Endocrinol Invest 37:1041–1048PubMedGoogle Scholar
  3. 3.
    Rivkees SA, Dinauer C (2007) An optimal treatment for pediatric Graves’ disease is radioiodine. J Clin Endocrinol Metab 92:797–800PubMedGoogle Scholar
  4. 4.
    Lee JA et al (2007) The optimal treatment for pediatric Graves’ disease is surgery. J Clin Endocrinol Metab 92:801–803PubMedGoogle Scholar
  5. 5.
    Nakamura H et al (2007) Comparison of methimazole and propylthiouracil in patients with hyperthyroidism caused by Graves’ disease. J Clin Endocrinol Metab 92:2157–2162PubMedGoogle Scholar
  6. 6.
    Rivkees SA, Szarfman A (2010) Dissimilar hepatotoxicity profiles of propylthiouracil and methimazole in children. J Clin Endocrinol Metab 95:3260–3267PubMedGoogle Scholar
  7. 7.
    Brito JP et al (2016) Antithyroid drugs-the most common treatment for Graves’ disease in the United States: a nationwide population-based study. Thyroid 26:1144–1145PubMedGoogle Scholar
  8. 8.
    Ross DS et al (2016) American thyroid association guidelines for diagnosis and management of hyperthyroidism and other causes of thyrotoxicosis. Thyroid 26:1343–1421PubMedGoogle Scholar
  9. 9.
    Wartofsky L et al (1991) Differences and similarities in the diagnosis and treatment of Graves’ disease in Europe, Japan and the United States. Thyroid 1:129–135PubMedGoogle Scholar
  10. 10.
    Burch HB et al (2012) A 2011 survey of clinical practice patterns in the management of Graves’ disease. J Clin Endocrinol Metab 97:4549–4558PubMedGoogle Scholar
  11. 11.
    Vaidya B et al (2008) Radioiodine treatment for benign thyroid disorders: results of a nationwide survey of UK endocrinologists. Clin Endocrinol (Oxf) 68:814–820Google Scholar
  12. 12.
    Escobar-Jimenez F et al (2000) Trends in diagnostic and therapeutic criteria in Graves’ disease in the last 10 years. Postgrad Med J 76:340–344PubMedPubMedCentralGoogle Scholar
  13. 13.
    Yamashita S et al (2011) The American Thyroid Association and American Association of Clinical Endocrinologists hyperthyroidism and other causes of thyrotoxicosis guidelines: viewpoints from Japan and Korea. Thyroid 21:577–580PubMedGoogle Scholar
  14. 14.
    Moon JH, Yi KH (2013) The diagnosis and management of hyperthyroidism in Korea: consensus report of the Korean thyroid association. Endocrinol Metab (Seoul) 28:275–279Google Scholar
  15. 15.
    Anagnostis P et al (2013) Predictors of long-term remission in patients with Graves’ disease: a single center experience. Endocrine 44:448–453PubMedGoogle Scholar
  16. 16.
    Orgiazzi J (2015) Should protracted treatment with antithyroid drug (ATD) be considered as a routine strategy in patients with graves’ disease who had a relapse after a first course of ATD? Clin Thyroidol 27:302–305Google Scholar
  17. 17.
    Léger J, Carel JC (2017) Management of endocrine disease: arguments for the prolonged use of antithyroid drugs in children with Graves’ disease. Eur J Endocrinol 177:R59–R67PubMedGoogle Scholar
  18. 18.
    Azizi F, Malboosbaf R (2017) Long-term antithyroid drug treatment: a systematic review and meta-analysis. Thyroid 27:1223–1231PubMedGoogle Scholar
  19. 19.
    Liberati A et al (2009) The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ 21:339b2700.  https://doi.org/10.1371/journal.pmed.1000100 CrossRefGoogle Scholar
  20. 20.
    Wells GA (2001) The Newcastle-Ottawa scale (NOS) for assessing the quality of non randomised studies in meta-analyses. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp
  21. 21.
    Egger M et al (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315:629–634PubMedPubMedCentralGoogle Scholar
  22. 22.
    Centre for Reviews and Dissemination (2009) CRD's guidance for undertaking reviews in healthcare. York Publishing Services. www.york.ac.uk
  23. 23.
    Yasuda K et al (2017) Relationship between dose of antithyroid drugs and adverse events in pediatric patients with Graves’ disease. Clin Pediatr Endocrinol 26:1–7PubMedPubMedCentralGoogle Scholar
  24. 24.
    Azizi F et al (2012) Long-term continuous methimazole or radioiodine treatment for hyperthyroidism. Arch Iran Med 15:477–484PubMedGoogle Scholar
  25. 25.
    Le´ger J et al (2012) Positive impact of long-term antithyroid drug treatment on the outcome of children with Graves’ disease: national long-term cohort study. J Clin Endocrinol Metab 97:110–119Google Scholar
  26. 26.
    Laurberg P et al (2011) Sustained control of Graves’ hyperthyroidism during long-term low-dose antithyroid drug therapy of patients with severe Graves’ orbitopathy. Thyroid 21:951–956PubMedGoogle Scholar
  27. 27.
    Sato H et al (2011) Comparison of methimazole and propylthiouracil in the management of children and adolescents with Graves’ disease: efficacy and adverse reactions during initial treatment and long-term outcome. J Pediatr Endocrinol Metab 24:257–263PubMedGoogle Scholar
  28. 28.
    Chen DY et al (2009) Comparison of the long-term efficacy of low dose 131I versus antithyroid drugs in the treatment of hyperthyroidism. Nucl Med Commun 30:160–168PubMedGoogle Scholar
  29. 29.
    Kaguelidou F, French Childhood Graves’ Disease Study Group et al (2008) Predictors of autoimmune hyperthyroidism relapse in children after discontinuation of antithyroid drug treatment. J Clin Endocrinol Metab 93:3817–3826PubMedGoogle Scholar
  30. 30.
    Azizi F et al (2005) Effect of long-term continuous methimazole treatment of hyperthyroidism: comparison with radioiodine. Eur J Endocrinol 152:695–701PubMedGoogle Scholar
  31. 31.
    Barrio R et al (2005) Graves’ disease in children and adolescents: response to long-term treatment. Acta Paediatr 94:1583–1589PubMedGoogle Scholar
  32. 32.
    Mashio Y et al (1997) Treatment of hyperthyroidism with a small single daily dose of methimazole: a prospective long-term follow-up study. Endocr J 44:553–558PubMedGoogle Scholar
  33. 33.
    Rivkees SA et al (2010) Adverse events associated with methimazole therapy of graves’ disease in children. Int J Pediatr Endocrinol 2010:1–4Google Scholar
  34. 34.
    Watanabe N et al (2012) Antithyroid drug-induced hematopoietic damage: a retrospective cohort study of agranulocytosis and pancytopenia involving 50,385 patients with Graves’ disease. J Clin Endocrinol Metab 97:E49–53PubMedGoogle Scholar
  35. 35.
    Wartofsky L (1993) Has the use of antithyroid drugs for Graves’ disease become obsolete? Thyroid 3:335–344PubMedGoogle Scholar
  36. 36.
    Franklyn JA et al (1991) Long-term follow up of treatment of hyperthyroidism by three different methods. Clin Endocrinol 34:71–76Google Scholar
  37. 37.
    Astwood EB (1967) Use of antithyroid drugs. In: Irvine WJ (ed) Hyperthyroidism. Williams & Wilkins, Baltimore, pp 85–98Google Scholar
  38. 38.
    Abraham P et al (2010) Antithyroid drug regimen for treating Graves’ hyperthyroidism. Cochrane Database Syst Rev 20:CD003420Google Scholar
  39. 39.
    Masiello E et al (2018) Antithyroid drug treatment for Graves’ disease: baseline predictive models of relapse after treatment for a patient-tailored management. J Endocrinol Invest 41:1425–1432PubMedGoogle Scholar
  40. 40.
    Laurberg P (2006) Remission of Graves’ disease during anti-thyroid drug therapy. Time to reconsider the mechanism? Eur J Endocrinol 155:783–786PubMedGoogle Scholar
  41. 41.
    Montani V et al (1998) Regulation of major histocompatibility class II gene expression in FRTL-5 thyrocytes: opposite effects of interferon and methimazole. Endocrinology 139:290–302PubMedGoogle Scholar
  42. 42.
    Mozes E et al (1998) Spontaneous autoimmune disease in (NZB x NZW) F1 mice is ameliorated by treatment with methimazole. J Clin Immunol 18:106–113PubMedGoogle Scholar
  43. 43.
    Volpe R (1994) Evidence that the immunosuppressive effects of antithyroid drugs are mediated through actions on the thyroid cell, modulating thyrocyte-immunocyte signaling: a review. Thyroid 4:217–223PubMedGoogle Scholar
  44. 44.
    Calissendorff J et al (2015) A prospective investigation of graves’ disease and selenium: thyroid hormones, auto-antibodies and self-rated symptoms. Eur Thyroid J 2:93–98Google Scholar
  45. 45.
    Leo M et al (2017) Effects of selenium on short-term control of hyperthyroidism due to Graves’ disease treated with methimazole: results of a randomized clinical trial. J Endocrinol Invest 40:281–287PubMedGoogle Scholar
  46. 46.
    Marinò M et al (2017) Selenium in the treatment of thyroid diseases. Eur Thyroid J 2:113–114Google Scholar
  47. 47.
    Rayman MP (2012) Selenium and human health. Lancet 379:1256–1268PubMedGoogle Scholar
  48. 48.
    Rotondi M et al (2007) Role of chemokines in endocrine autoimmune diseases. Endocr Rev 28:492–520PubMedGoogle Scholar
  49. 49.
    Slingerland DW, Burrows BA (1979) Long-term antithyroid treatment in hyperthyroidism. JAMA 242:2408–2410PubMedGoogle Scholar
  50. 50.
    Lippe BM et al (1987) Hyperthyroidism in children treated with Long term medical therapy: twenty-five percent remission every two years. J Clin Endocrinol Metab 64:1241–1245PubMedGoogle Scholar
  51. 51.
    Elbers L et al (2011) Outcome of very long-term treatment with antithyroid drugs in Graves’ hyperthyroidism associated with Graves’ orbitopathy. Thyroid 21:279–383PubMedGoogle Scholar
  52. 52.
    Villagelin D et al (2015) Outcomes in relapsed graves’ disease patients following radioiodine or prolonged low dose of methimazole treatment. Thyroid 25:1282–1290PubMedGoogle Scholar
  53. 53.
    Hieu TT et al (2012) Cancer risk after medical exposure to radioactive iodine in benign thyroid diseases: a meta-analysis. Endocr Relat Cancer 19:645–655PubMedGoogle Scholar
  54. 54.
    Sato H et al (2012) Higher dose of methimazole causes frequent adverse effects in the management of Graves’ disease in children and adolescents. J Pediatr Endocrinol Metab 25:863–867PubMedGoogle Scholar
  55. 55.
    Sato S et al (2015) Comparison of efficacy and adverse effects between methimazole 15 mg + inorganic iodine 38 mg/day and methimazole 30 mg/day as initial therapy for Graves’ disease patients with moderate to severe hyperthyroidism. Thyroid 25:43–50PubMedGoogle Scholar
  56. 56.
    Reinwein D et al (1993) A prospective randomized trial of antithyroid drug dose in Graves’ disease therapy. European Multicenter Study Group on Antithyroid Drug Treatment. J Clin Endocrinol Metab 76:1516–1521PubMedGoogle Scholar
  57. 57.
    Takata K et al (2009) Methimazole-induced agranulocytosis in patients with Graves’ disease is more frequent with an initial dose of 30 mg daily than with 15 mg daily. Thyroid 19:559–563PubMedGoogle Scholar
  58. 58.
    Romaldim JH et al (1983) Comparison of effects of high and low dosage regimens of antithyroid drugs in the management of Graves’ hyperthyroidism. J Clin Endocrinol Metab 57:563–570Google Scholar
  59. 59.
    Wilberg JJ, Nuttall FQ (1972) Methimazole toxicity from high doses. Ann Intern Med 77:414–416Google Scholar
  60. 60.
    Nakamura H et al (2013) Analysis of 754 cases of antithyroid drug-induced agranulocytosis over 30 years in Japan. J Clin Endocrinol Metab 98:4776–4783PubMedGoogle Scholar
  61. 61.
    Wang MT et al (2014) Antithyroid drug-related hepatotoxicity in hyperthyroidism patients: a population-based cohort study. Br J Clin Pharmacol 78:619–629PubMedPubMedCentralGoogle Scholar
  62. 62.
    Yang J et al (2015) Analysis of 90 cases of antithyroid drug-induced severe hepatotoxicity over 13 years in China. Thyroid 25:278–283PubMedGoogle Scholar
  63. 63.
    Ohye H et al (2014) Antithyroid drug treatment for Graves ‘disease in children: a long-term retrospective study at a single institution. Thyroid 24:200–207PubMedGoogle Scholar
  64. 64.
    Minamitani K et al (2011) A report of three girls with antithyroid drug-induced agranulocytosis; retrospective analysis of 18 cases aged 15 years or younger reported between 1995 and 2009. Clin Pediatr Endocrinol 20:39–46PubMedPubMedCentralGoogle Scholar
  65. 65.
    Tamai H et al (1989) Methimazole-induced agranulocytosis in Japanese patients with Graves’ disease. Clin Endocrinol 30:525–530Google Scholar
  66. 66.
    He Y et al (2017) Emphasis on the early diagnosis of antithyroid drug-induced agranulocytosis: retrospective analysis over 16 years at one Chinese center. J Endocrinol Invest 40:733–740PubMedGoogle Scholar
  67. 67.
    Balavoine AS et al (2015) Antineutrophil cytoplasmic antibody-positive small-vessel vasculitis associated with antithyroid drug therapy: how significant is the clinical problem? Thyroid 25:1273–1281PubMedGoogle Scholar
  68. 68.
    Abhayaratna S, Somasundaram N (2013) The role of long term use of antithyroid drugs in Graves’ disease. Sri Lanka J Diabetes Endocrinol Metabol 3:41–44Google Scholar
  69. 69.
    Howard JE (1967) Treatment of thyrotoxicosis. JAMA 202:706–709PubMedGoogle Scholar
  70. 70.
    Sattler H (1952) Basedow’s disease. In: Marchand GW, Marchland JF (trans-ed). Grune & Stratton Inc, New York, pp 367–403Google Scholar
  71. 71.
    McLarty DG et al (1973) Remission of thyrotoxicosis during treatment with propranolol. Br Med J 2:332–334PubMedPubMedCentralGoogle Scholar

Copyright information

© Italian Society of Endocrinology (SIE) 2019

Authors and Affiliations

  1. 1.Internal Medicine and Endocrinology, Endocrine Research Center of Research Institute for Endocrine SciencesShahid Beheshti University of Medical SciencesTehranIslamic Republic of Iran
  2. 2.Internal Medicine and Endocrinology, Endocrine Research Center, Institute of Endocrinology and MetabolismIran University of Medical SciencesTehranIslamic Republic of Iran

Personalised recommendations