Glucocorticoid Regimens for Prevention of Graves' Ophthalmopathy Progression Following Radioiodine Treatment: Systematic Review and Meta-Analysis

Abstract

Background:

Glucocorticoid (GC) therapy has been shown to prevent Graves' ophthalmopathy (GO) progression following radioactive iodine (RAI) treatment. However, the optimal regimen is controversial, with studies from recent years suggesting the use of lower doses and shorter GC treatment courses.

Methods:

We conducted a systematic review and meta-analysis of randomized controlled trials (RCTs) and retrospective controlled trials comparing GC regimens versus placebo, no treatment, or other GC regimens.

Results:

Eight trials evaluating 850 patients fulfilled inclusion criteria. In patients with preexisting GO, standard dose prednisone (0.4–0.5 mg/kg tapered over 3 months) was very effective for prevention of GO progression (OR 0.14 [CI 0.06–0.35], p<0.01) in patients with mild to moderate GO. Two studies evaluated low-dose prednisone (0.2–0.3 mg/kg for 4–6 weeks) in patients with mild GO or risk factors, but were limited by not including patients with preexisting GO in the control groups. Therefore, the two low-dose groups were evaluated using indirect comparisons with control groups matched for age and clinical activity score, showing excellent efficacy versus no treatment or placebo (OR 0.20 [CI 0.07–0.60], p=0.004) and no significant difference compared with standard dose (OR 1.7 [CI 0.52–5.52], p=0.47). In patients without preexisting GO, steroid prophylaxis had no beneficial effect (OR 1.87 [CI 0.81–4.3]), though there were insufficient data regarding patients with risk factors for GO development. GC prophylaxis had no impact on hyperthyroidism resolution (OR 1.05 [CI 0.69–1.58]), and GC side effects were common but mild.

Conclusions:

Current evidence supports a three-tier approach for prevention of GO progression following RAI. Standard dose prednisone is the best validated regimen and should be used in patients with mild to moderate GO who have high risk of progression, while low dose prednisone can be used in patients with mild GO, and in patients without preexisting GO who have risk factors and are selected for GC prophylaxis. Patients without preexisting GO and without risk factors should not be treated with GC prophylaxis.

Introduction

Radioactive iodine (RAI) treatment is an effective and widely used treatment for Graves' hyperthyroidism, together with antithyroid drugs and surgery. However, its use is limited by the potential risk of development or worsening of Graves' ophthalmopathy (GO) following therapy. While several reports did not demonstrate progression of GO following RAI (1,2), many more found a significant association between RAI and the development or the worsening of preexisting GO, including one systematic review (3 –7). GO progression after RAI is more likely to occur in patients who already have GO prior to therapy, smoke, have severe hyperthyroidism, high levels of thyrotropin receptor antibody (TRAb), short duration of disease, or whose post-RAI hypothyroidism is not promptly corrected by levothyroxine replacement therapy (8,9).

The efficacy of corticosteroids in prevention of RAI-induced GO activation was well demonstrated in two randomized controlled trials (RCTs) by Bartalena et al., using prednisone tapered over three months (3,10). Since then, several studies evaluated lower doses of prednisone with shorter treatment duration (11), and some studies found steroid prophylaxis to be ineffective (9,12). According to a treatment consensus of the European Group on Graves' Orbitopathy (EUGOGO) published in 2008, and the management guidelines of the American Thyroid Association (ATA) in association with the American Association of Clinical Endocrinologists (AACE), prednisone prophylaxis 0.4–0.5 mg/kg is recommended in patients recognized to be at risk for GO activation, tapered over three months (13,14). In practice, there is wide variation in when and how glucocorticoids (GCs) are used to prevent GO progression as far as who should be treated, the optimal GC dose, and the duration of treatment (15). To address this challenge, we performed a systematic review and meta-analysis to determine the best validated and effective GC regimen for GO prophylaxis, with an emphasis on lower doses of prednisone, stratified by GO status prior to RAI treatment.

Methods

Data sources

We searched PubMed (January 1966 to January 2014), CENTRAL (The Cochrane Library, up to January 2013), and conference proceedings in endocrinology, ophthalmology, and nuclear medicine. The terms “Graves' ophthalmopathy” and similar, “radioiodine” or “radioactive iodine” and similar, and “steroids” or specific steroid names were crossed. We scanned references of all included trials and reviews identified for additional studies.

Study selection

We included all RCTs, as well as retrospective controlled studies comparing steroid regimens with controls, including placebo, no intervention, or a different steroid regimen for prevention of GO progression. We included trials regardless of publication status and language. Studies in which GO progression was not a primary outcome were included only if they were RCTs, as the risk for selection bias with this design is small. Two reviewers independently inspected each reference identified by the search and applied inclusion criteria.

Data extraction and quality assessment

Two reviewers independently extracted data from the included trials. In the case of any disagreement between the two reviewers, a third reviewer extracted the data. We contacted the authors of trials for missing data, when necessary. Two independent reviewers assessed the trials for methodological quality. We individually assessed the following components: allocation concealment, generation of the allocation sequence, and blinding. We graded allocation concealment and generation as adequate, unclear, or inadequate (16). We also collected data on exclusions after randomization and whether the primary analysis was performed according to the intention-to-treat principle or per protocol.

Definition of outcomes

The primary outcome was new or worsening GO following RAI treatment, stratified according to preexisting GO status. Secondary outcomes included resolution of hyperthyroidism and steroid-related adverse events.

Data synthesis and analysis

We analyzed data by calculating the odds ratio (OR) for each trial, with uncertainty expressed using 95% confidence intervals (CI) (Review Manager (RevMan) v5.2 for Windows; The Cochrane Collaboration, Oxford, United Kingdom). We utilized a fixed-effect model using the Mantel–Haenszel method for pooling trial results throughout the review, unless statistically significant heterogeneity was found, in which case we chose a random-effects model using the DerSimonian and Laird method. We assessed the heterogeneity of trial results by inspecting graphical presentations and calculating a chi-square test of heterogeneity and a I ² measure of inconsistency (17). We explored potential sources of heterogeneity through predefined subgroup analyses.

Results

Description of the studies

The literature search identified 19 manuscripts of GC treatment to prevent GO progression following RAI treatment for Graves' hyperthyroidism. Reasons for exclusion are detailed in Figure 1. Eight studies that were performed between 1986 and 2010 fulfilled inclusion criteria, including five RCTs and three retrospective controlled studies, evaluating 850 patients (Table 1) (3,9 –12,18 –20). GCs used were oral prednisone in five studies, oral betamethasone in two studies, oral prednisolone in one study, and intravenous methylprednisolone in one study.

FIG. 1.

Selection process for included studies.

Table 1.

Included Studies

Study	Type	Pt (no.)	Treatment (initial dose)	Initiation	Duration	Follow-up	Mean age (years)	F (%)	Preexisting GO (%)	Smokers (%)	CAS (mean)	Primary outcome	Allocation generation, concealment	Binding	Sponsored y/n
Bartalena 1998 (3)	RCT	145	PO prednisone (0.4–0.5 mg/kg)	2–3 days after RAI	3 months	12 months	42	81%	52%	60%	1.2 (SD=1.2)	Hyperthyroidism resolution, GO progression	B,B	N	N
		150	No treatment				44	81%	48%	55%	1.1 (SD=1.1)
Bartalena 1989 (10)	RCT	26	PO prednisone (0.4–0.5 mg/kg)	2–3 days after RAI	3 months	18 months	Both groups—49	81%	81%	NA	NA	GO progression	A,B	N	N
		26	No treatment					77%	62%
Vannucchi 2009 (9)	Retrospective	21	PO prednisone (M=36.5 mg)^*	1–2 days after RAI	8–10 weeks^*	21 months	47	76%	36%	52%	1 (SE=0.3)	GO progression			N
		9	IV methylprednisolone (500 mg twice/week)	7 days after RAI	1 month		53.2	78%	55%	33%	2.1 (SE=0.4)
		83	No treatment				45.2	80%	28%	37%	0.6 (SE=0.1)
Lai 2010 (11)	Retrospective	28	PO prednisone: low dose (0.22±0.03 mg/kg)	1 day after RAI	6 weeks	6 months	51	64%	79%	30%	1 (SD=0.9)	GO progression			N
		28	PO prednisone: high dose (0.36±0.05 mg/kg)				47	78%	71%	30%	1.3 (SD=1)
		35	No treatment				NA	NA	NA
Jensen 2005 (12)	Retrospective	96	PO prednisolone (25 mg/day)	2 days before RAI	1 month	12 months	44.6	85%	67%	42%	NA	Hyperthyroidism resolution			N
		111	No treatment				51.3	77%	0%	41%
Vaisman 1997 (20)	RCT	25	PO prednisone (0.5 mg/kg)	The day of RAI	10–12 months	18 months	36^*	Total: 80%	100%	NA	NA “mild to moderate”	GO progression	A,A^*	N	N
		24	No treatment						100%
Gamstedt 1991 (18)	RCT	20	PO bethamethasone (6 mg)	3 weeks before RAI	7 weeks	12 months	55	85%	0%	NA	NA		B,B	Y	N
		20	Placebo				62	85%	0%
Gamstedt 1986 (19)	RCT	20	PO bethamethasone (6 mg)	3 weeks before RAI	7 weeks	1 year	57	70%	0%				B,B	N	N
		23	No treatment				60	78%	0%

Based on supplementary data provided by the authors.

Allocation grading: A, adequate; B, unclear.

NA, not available; Y, yes; N, no.

Preexisting GO was present in >50% of patients in five studies, in <50% in one study, and in none of the patients in two studies. Steroid administration was initiated before RAI in three studies, on the day of RAI treatment in one study, one to three days after RAI treatment in four studies, and a week after RAI treatment in one study (Tables 1 and 2).

Table 2.

Treatment Protocols

Corticosteroid treatment regimen	Study
Starting two to three days after RAI treatment, oral prednisone 0.4–0.5 mg/kg for one month, and then tapered off over two months	Bartalena 1998 (3) Bartalena 1989 (10)
Starting one to two days after RAI treatment, mean initial dose of prednisone 36.5 mg (8/21 patients 50 mg; 11/21 patients 25 mg; and 2/21 patients 35 mg) tapered off over 8–10 weeks	Vannucchi 2009 (9)
IV methylprednisolone: Starting seven days after RAI treatment, two weekly doses of 500 mg, then two weekly doses of 250 mg
Prednisone—standard dose: Starting one day after RAI treatment, mean initial dose of 0.36±0.05 mg/kg tapered off over six weeks	Lai 2010 (11)
Prednisone—low dose: Starting one day after RAI treatment, mean initial dose of 0.22±0.03 mg/kg tapered off over six weeks
Starting two days before RAI treatment, oral prednisolone 25 mg once daily for one month without tapering	Jensen 2005 (12)
Starting the day of RAI treatment, oral prednisone 0.5 mg/kg for four months, tapered off over 10–12 months	Vaisman 1997 (20)
Starting three weeks before RAI treatment, oral betamethasone 6 mg daily for five days, then reduced 0.5–1.5 mg weekly for a total of seven weeks	Gamstedt 1991 (18)
	Gamstedt 1986 (19)

Patients with preexisting GO

Six studies reported GO outcome following RAI treatment in patients with preexisting GO (3,10 –12,18,19). In one study, the entire patient population had preexisting GO (20), while in others, data were extracted from the treatment and control groups. The clinical activity score (CAS) before treatment was reported in three studies (Table 1) correlating with mild to moderately active GO, and in three studies, the GO severity was described in the text as mild, or mild to moderate. Overall, GCs were very effective in preventing GO progression in patients with preexisting GO (OR 0.14 [CI 0.06–0.35], 378 patients; Fig. 2). The most studied drug was prednisone, which was evaluated in three RCTs and three retrospective studies, with doses ranging from 0.2 mg/kg to 0.5 mg/kg, and duration ranging from 1 to 10–12 months. Standard dose prednisone, 0.4–0.5 mg/kg a day for three or more months, was evaluated in five studies (three RCTs and two retrospective studies) in patients with mild to moderate GO, with excellent efficacy (OR 0.16 [CI 0.06–0.37], p<0.01). Sensitivity analysis by study type showed higher effects estimates in RCTs compared with retrospective studies (Fig. 2).

FIG. 2.

Graves' ophthalmopathy (GO) progression stratified according to preexisting GO status. (a) Patients with pre-existing GO, (b) patients without pre-existing GO.

Low-dose prednisone, 0.2–0.3 mg/kg tapered over four to six weeks, was evaluated in two retrospective studies in patients with mild GO (11,12). In both studies, the intervention arm included patients with mild or mild to moderate GO, while the control group included patients without preexisting GO, making the groups incomparable due to different patient populations. Therefore, to evaluate the efficacy of low-dose prednisone, an indirect comparison was made between the intervention groups and matched control groups from the studies included in the meta-analysis, which were matched for sex, age, and disease activity (females 80% vs. 81%, age 45.5 vs. 43.6 years, and CAS 0.9 (SD=0.8) vs. 1.1 (SD=0.7) respectively, n.s.) (3,9,20). Low-dose prednisone was effective in preventing GO progression compared to no treatment or placebo (OR 0.20 [CI 0.07–0.60], p=0.004) (3,9,20) (Fig. 3), and was not significantly different from standard dose prednisone (OR 1.7 [CI 0.52–5.52], p=0.47) (3,9 –11,20).

FIG. 3.

GO progression stratified by drug. (a) Patients with pre-existing GO, (b) patients without pre-existing GO.

Intravenous methylprednisolone was evaluated in one retrospective study, in which five patients with preexisting GO were treated once weekly for one month (Table 2), and none had disease progression. This was compared to no treatment and to standard dose prednisone, but the sample size was insufficient to prove efficacy or noninferiority when a noninferiority limit (d) of 25% was used.

Patients with no preexisting GO

GC treatment for the prevention of GO progression in patients without GO was evaluated in seven studies, of which four were RCTs and three were retrospective studies (3,9 –12,18,19) (Table 1). In two studies, no participant had preexisting GO (18,19), while in others, these patients represented 20–68% of the study population. Prednisone was evaluated in four studies (two RCTs and two retrospective), betamethasone in two RCTs, prednisolone in one retrospective study, and intravenous methylprednisolone in one retrospective study (Table 1).

In patients without preexisting GO, standard dose prednisone (0.4–0.5 mg/kg for three months) was evaluated in two RCTs and two retrospective studies, and was ineffective (OR 1.87 [CI 0.81–4.3]). Betamethasone was evaluated in two RCTs, and was also found to be ineffective (OR 6 [CI 0.68–52.9]).

When subgroup analysis was performed according to study type, there was a clear difference between RCTs showing steroid prophylaxis to be ineffective (OR 0.8 [CI 0.26–2.46]), and retrospective studies that surprisingly found steroid prophylaxis to be harmful. However, in these three retrospective studies, the intervention and control groups were uneven, with all patients in the intervention group having one or more risk factors, while all patients in the control group had no risk factors, and thus a significantly lower risk of progression. The risk factors included smoking and/or thyrotropin receptor antibody (TRAb) levels (9,11,12). Even after carefully inspecting all included studies in the meta-analysis, none of the control groups were comparable to the intervention groups of patients with risk factors, and therefore the retrospective studies cannot guide treatment in patients without preexisting GO who have risk factors for GO progression.

Hyperthyroidism resolution

Four studies reported rates of hyperthyroidism resolution following RAI treatment with concomitant steroids compared to those treated with RAI alone (3,12,18,19). GC administration had no impact on treatment success rates (OR 1.05 [CI 0.69–1.58]). In a study by Bartalena et al., the rate of hyperthyroidism resolution was not reported, but was described as “similar” (10).

GC-related side effects

Four studies reported side effects of GC treatment, and all were described as minor (3,9 –11). These side effects included Cushingoid features, weight gain (average 2.8 kg with regular-dose prednisone and 1.8 kg with low-dose prednisone), anxiety, hyperglycemia, and insomnia. Side effects were reported in 62% with prednisone 0.4–0.5 mg/kg, in 43% with low-dose prednisone (0.2–0.3 mg/kg), and in 48% with intravenous methylprednisolone.

Discussion

To guide the optimal use of GC to prevent GO progression following RAI treatment, we performed a systematic review and meta-analysis evaluating 850 patients with Graves' disease. Our work stratified patients, for the first time to our knowledge, according to GO status prior to RAI treatment. Our results show that GCs are very effective in preventing GO progression in patients with preexisting GO, but not in patients without preexisting eye involvement (Fig. 2). The most validated regimen is prednisone 0.4–0.5 mg/kg given for three months or more in patients with preexisting eye disease, as recommended by both the ATA/ACEE management guidelines from 2011 and the EUGOGO consensus statement from 2008. While this regimen was evaluated in patients with mild or moderate GO, newer studies evaluated low-dose prednisone (0.2–0.3 mg/kg for four to six weeks) in patients with mild GO or risk factors, which was found to be an effective alternative with significantly less adverse effects. In patients without preexisting GO, steroid prophylaxis was ineffective in preventing GO development.

The need for an evidence-based approach to GC prophylaxis was well demonstrated by a survey among members of the European Thyroid Association (15). In that survey, there was a wide variation in the decision of who to treat and what regimen to use. Some respondents did not prescribe GC in any clinical scenario, others gave GCs to all patients receiving RAI therapy, while the majority only gave GCs to patients with some degree of ocular involvement or if risk factors, such as smoking, were present. This heterogeneity is understandable in light of the currently available studies included in our meta-analysis. These studies included heterogeneous patient populations (with active GO, nonactive GO, patients with risk factors only, and patients without known eye disease), and used multiple treatment regimens (Table 2). While several tools are available to evaluate for severity and activity of GO (CAS, measured proptosis, diplopia assessment, etc.), there is not a single tool that comprehensively describes GO disease status. Our goal was to stratify patients according to both severity of structural disease and disease activity, but most outcome data from included studies stratified patients only according to the presence or absence of preexisting GO, with most treated patients having some degree of active disease. Although grouping according to preexisting GO may not represent the full complexity of GO pathology, it reflects the currently available data and is effective in guiding clinical decision making.

In agreement with previously published studies and one meta-analysis, we found a high risk for GO progression following RAI in patients with preexisting GO (25%), and a much lower risk in patients without known eye involvement, with or without risk factors (6%) (3 –7). In patients with preexisting eye disease, who are at a high risk for progression, the risk was reduced from 25% to 4% by the use of prednisone or prednisolone (Table 3). However, in patients who had no preexisting GO in whom the risk of de novo GO development is relatively low, a larger sample size is required to prove the efficacy of preventive treatment. For example, to prove a 50% risk reduction in patients who have no preexisting GO, a sample size of approximately 300 patients is required. Therefore, current recommendations on preventive treatment for patients, who do not have GO but have risk factors for progression, is supported by limited evidence, and are mostly based on expert opinion and our understanding of the disease.

Table 3.

Risk of GO Progression Stratified by Drug and Preexisting GO

	No treatment	Prednisone 0.4–0.5 mg/kg for 3 months	Prednisone mean 36.5 mg for 8–10 weeks	Prednisone 0.2 mg/kg for 6 weeks	Prednisolone 25 mg/day for 1 month	IV Methylprednisolone for 1 month	Betamethasone 6 mg/day for 7 weeks
Preexisting GO	34/136 (25%)	3/120 (2.5%)	2/10 (20%)	0/22	4/64 (6%)	0/5	ND
No preexisting GO	16/293 (6%)^*	0/78	8/11 (73%)	0/6	3/32 (9%)	0/4	4/40 (10%)
Data source	5 RCTs, 3 retrospective	3 RCTs	1 retrospective	1 retrospective	1 retrospective	1 retrospective	2 RCTs

This group includes patients both with and without risk factors for GO occurrence.

RCT, randomized controlled trial.

While treatment with GCs is effective in preventing GO progression following RAI, it is not risk-free. The recommended dose of 0.4–0.5 mg/kg for three months may cause significant side effects, including elevated blood glucose levels, hypertension, weight gain, mood changes, and acne, among others. Thus, assessment of the risk–benefit ratio should be performed for each patient, considering the treatment alternatives for hyperthyroidism. For patients with preexisting GO who failed medical treatment and who are otherwise healthy, treatment with RAI with concurrent GC prophylaxis is reasonable and evidence-based. However, in patients who are at risk for GC-related side effects (patients with metabolic syndrome, mental disease, etc.), an individualized risk assessment should be performed, weighing the risk of GO progression according to GO severity and activity versus the risk for GC-related adverse effects. The use of low-dose prednisone or referral to thyroid surgery can be effective options in these cases.

Our systematic review is limited by several factors. First, most studies did not clearly define preexisting GO, which may include both active and inactive disease. While most included patients had active disease according to CAS scores and as described in the text, it is unclear whether these conclusions can be generalized to patients with inactive GO given previous studies showing low risk for progression in patients with inactive disease (21). Further studies should evaluate the value of GC prophylaxis in patients with inactive disease, and especially in patients with long duration of disease who have lower risk of activation (9). Second, studies included in our meta-analysis provided insufficient data regarding risk factors for GO occurrence and progression. Smoking and post-RAI hypothyroidism are known risk factors for GO, whereas a high TRAb autoantibody titer is a known risk factor for GO progression (22 –24). Currently, the role of GC prophylaxis in patients with risk factors who do not have preexisting GO remains unclear. Moreover, previous studies have shown that post-RAI hypothyroidism may contribute to GO progression, which was prevented or blunted by early administration of levothyroxine (21). We have no data regarding this risk factor, which may have affected the risk of GO progression or the efficacy of steroid therapy. Third, most trials evaluated multiple outcome measures, without a clear definition of a primary outcome. This, in combination with the relatively small sample sizes, increased the risk for alpha-type errors of finding false statistical differences between treatment groups. Finally, the natural history of GO is incompletely defined, and may remit or improve spontaneously, which might affect results in an unpredictable manner (25,26). The use of a meta-analysis may partially overcome the methodological difficulties mentioned above by enlarging the sample size.

In conclusion, current evidence supports a three-tier, risk-adjusted approach for prevention of GO progression. Standard dose prednisone (0.4–0.5 mg/kg tapered over three months) is the best validated regimen and should be used in patients with mild to moderate GO who have high risk of progression, while low-dose prednisone (0.2–0.3 mg/kg tapered over four to five weeks) can be used in patients with mild GO and in patients without preexisting GO who have risk factors and are selected for GC prophylaxis. Future research should further substantiate the efficacy of low-dose prednisone, and evaluate the role of GC prophylaxis in patients with risk factors for GO development and in patients with inactive GO.

Footnotes

Acknowledgments

We would like to thank Prof. Mario Vaisman and Dott.ssa Guia Vannucchi for kindly providing additional information regarding their studies, which were included in the meta-analysis.

Author Disclosure Statement

No competing financial interests exist.

References

Sisson

, Schipper

, Nelson

, Freitas

, Frueh

. 2008. Radioiodine therapy and Graves' ophthalmopathy. J Nucl Med, 49:923–930.

Sridama

, DeGroot

. 1989. Treatment of Graves' disease and the course of ophthalmopathy. Am J Med, 87:70–73.

Bartalena

, Marcocci

, Bogazzi

, Manetti

, Tanda

, Dell'Unto

, Bruno-Bossio

, Nardi

, Bartolomei

, Lepri

, Rossi

, Martino

, Pinchera

. 1998. Relation between therapy for hyperthyroidism and the course of Graves' ophthalmopathy. N Engl J Med, 338:73–78.

Kung

, Yau

, Cheng

. 1994. The incidence of ophthalmopathy after radioiodine therapy for Graves' disease: prognostic factors and the role of methimazole. J Clin Endocrinol Metab, 79:542–546.

Acharya

, Avenell

, Philip

, Burr

, Bevan

, Abraham

. 2008. Radioiodine therapy (RAI) for Graves' disease (GD) and the effect on ophthalmopathy: a systematic review. Clin Endocrinol (Oxf), 69:943–950.

Tallstedt

, Lundell

, Torring

, Wallin

, Ljunggren

, Blomgren

, Taube

. 1992. Occurrence of ophthalmopathy after treatment for Graves' hyperthyroidism. The Thyroid Study Group. N Engl J Med, 326:1733–1738.

Vestergaard

, Laurberg

. 1989. Radioiodine and aggravation of Graves' ophthalmopathy. Lancet, 2:47.

Bartalena

, Tanda

, Piantanida

, Lai

. 2005. Glucocorticoids and outcome of radioactive iodine therapy for Graves' hyperthyroidism. Eur J Endocrinol, 153:13–14.

Vannucchi

, Campi

, Covelli

, Dazzi

, Curro

, Simonetta

, Ratiglia

, Beck-Peccoz

, Salvi

. 2009. Graves' orbitopathy activation after radioactive iodine therapy with and without steroid prophylaxis. J Clin Endocrinol Metab, 94:3381–3386.

10.

Bartalena

, Marcocci

, Bogazzi

, Panicucci

, Lepri

, Pinchera

. 1989. Use of corticosteroids to prevent progression of Graves' ophthalmopathy after radioiodine therapy for hyperthyroidism. N Engl J Med, 321:1349–1352.

11.

Lai

, Sassi

, Compri

, Marino

, Sivelli

, Piantanida

, Tanda

, Bartalena

. 2010. Lower dose prednisone prevents radioiodine-associated exacerbation of initially mild or absent Graves' orbitopathy: a retrospective cohort study. J Clin Endocrinol Metab, 95:1333–1337.

12.

Jensen

, Bonnema

, Hegedus

. 2005. Glucocorticoids do not influence the effect of radioiodine therapy in Graves' disease. Eur J Endocrinol, 153:15–21.

13.

Bartalena

, Baldeschi

, Dickinson

, Eckstein

, Kendall-Taylor

, Marcocci

, Mourits

, Perros

, Boboridis

, Boschi

, Curro

, Daumerie

, Kahaly

, Krassas

, Lane

, Lazarus

, Marino

, Nardi

, Neoh

, Orgiazzi

, Pearce

, Pinchera

, Pitz

, Salvi

, Sivelli

, Stahl

, von Arx

, Wiersinga

. 2008. Consensus statement of the European Group on Graves' Orbitopathy (EUGOGO) on management of Graves' orbitopathy. Thyroid, 18:333–346.

14.

Bahn

, Burch

, Cooper

, Garber

, Greenlee

, Klein

, Laurberg

, McDougall

, Montori

, Rivkees

, Ross

, Sosa

, Stan

. 2011. Hyperthyroidism and other causes of thyrotoxicosis: management guidelines of the American Thyroid Association and American Association of Clinical Endocrinologists. Endocr Pract, 17:456–520.

15.

Perros

, Baldeschi

, Boboridis

, Dickinson

, Hullo

, Kahaly

, Kendall-Taylor

, Krassas

, Lane

, Lazarus

, Marcocci

, Marino

, Mourits

, Nardi

, Orgiazzi

, Pinchera

, Pitz

, Prummel

, Wiersinga

. 2006. A questionnaire survey on the management of Graves' orbitopathy in Europe. Eur J Endocrinol, 155:207–211.

16.

Cochrane Handbook for Systematic Reviews of Interventions 5.1.0. Available at: www.cochrane.org/resources/handbook (accessed January 2014 ).

17.

Higgins

, Thompson

, Deeks

, Altman

. 2003. Measuring inconsistency in meta-analyses. BMJ, 327:557–560.

18.

Gamstedt

, Karlsson

. 1991. Pretreatment with betamethasone of patients with Graves' disease given radioiodine therapy: thyroid autoantibody responses and outcome of therapy. J Clin Endocrinol Metab, 73:125–131.

19.

Gamstedt

, Wadman

, Karlsson

. 1986. Methimazole, but not betamethasone, prevents 131I treatment-induced rises in thyrotropin receptor autoantibodies in hyperthyroid Graves' disease. J Clin Endocrinol Metab, 62:773–777.

20.

Vaisman

. 1997. Emprego profilático de corticóide na prevençäo da progressä da oftalmopatia de Graves após tratamento do hipertireoidismo com iodo radioativo [Corticosteroids therapy to prevent progression of Graves' ophthalmopathy after treatment of hyperthyroidism with radioiodine.]. Folha méd, 115:165–167.

21.

Perros

, Kendall-Taylor

, Neoh

, Frewin

, Dickinson

. 2005. A prospective study of the effects of radioiodine therapy for hyperthyroidism in patients with minimally active Graves' ophthalmopathy. J Clin Endocrinol Metab, 90:5321–5323.

22.

Traisk

, Tallstedt

, Abraham-Nordling

, Andersson

, Berg

, Calissendorff

, Hallengren

, Hedner

, Lantz

, Nystrom

, Ponjavic

, Taube

, Torring

, Wallin

, Asman

, Lundell

. 2009. Thyroid-associated ophthalmopathy after treatment for Graves' hyperthyroidism with antithyroid drugs or iodine-131. J Clin Endocrinol Metab, 94:3700–3707.

23.

Tallstedt

, Lundell

, Blomgren

, Bring

. 1994. Does early administration of thyroxine reduce the development of Graves' ophthalmopathy after radioiodine treatment?. Eur J Endocrinol, 130:494–497.

24.

Ponto

, Zang

, Kahaly

. 2010. The tale of radioiodine and Graves' orbitopathy. Thyroid, 20:785–793.

25.

Tanda

, Piantanida

, Liparulo

, Veronesi

, Lai

, Sassi

, Pariani

, Gallo

, Azzolini

, Ferrario

, Bartalena

. 2013. Prevalence and natural history of Graves' orbitopathy in a large series of patients with newly diagnosed Graves' hyperthyroidism seen at a single center. J Clin Endocrinol Metab, 98:1443–1449.

26.

Piantanida

, Tanda

, Lai

, Sassi

, Bartalena

. 2013. Prevalence and natural history of Graves' orbitopathy in the XXI century. J Endocrinol Invest, 36:444–449.