Can we predict relapse in Graves’ disease? Results from a systematic review and meta-analysis

in European Journal of Endocrinology
View More View Less
  • 1 Medical University Department, Clinic for Endocrinology, Diabetes & Metabolism, Kantonsspital Aarau, Aarau, Switzerland
  • | 2 Medical Faculty of the University of Basel, Basel, Switzerland

Correspondence should be addressed to T Struja; Email: tristan.struja@gmail.com

(T Struja and H Fehlberg contributed equally to this work)

Free access

Background

Identification of pretreatment risk factors predicting relapse in patients with hyperthyroidism of Graves’ disease after stopping anti-thyroid drugs (ATD) is decisive to guide therapeutic options.

Purpose

We performed a systematic search and meta-analysis to study predictors for relapse after stopping ATD in patients with Graves’ disease.

Methods

Based on a pre-specified protocol, we searched PubMed, EMBASE and Cochrane in July 2015 for case–control, controlled and randomized-controlled trials reporting risk factors for relapse after stopping ATD. The primary endpoint was relapse of disease until follow-up. PRISMA and SIGN statements were used for reviewing the data and assessing the quality of included trials.

Results

We included 54 trials with a total of 7595 participants. Most trials were small with moderate-to-high risk for bias. Ten trials were assessed only qualitatively (2227 patients), genomic data were reported in 13 trials (2178 patients) and 31 trials (4346 patients) were assessed quantitatively. In total, there were 3696 relapses in 7595 patients (48.7%). By using random-effects meta-analysis, orbitopathy, smoking, thyroid volume measured by sonography, goiter size, fT4, tT3, TRAb and TBII were significantly associated with relapse, whereas male vs female sex, age and initial tT4 level did not show significant associations.

Conclusions

This analysis found several risk factors to predict relapse in Graves’ disease, which can be combined in a risk score. Prospective studies should evaluate the prognostic accuracy of such a score to guide treatment decisions.

Abstract

Background

Identification of pretreatment risk factors predicting relapse in patients with hyperthyroidism of Graves’ disease after stopping anti-thyroid drugs (ATD) is decisive to guide therapeutic options.

Purpose

We performed a systematic search and meta-analysis to study predictors for relapse after stopping ATD in patients with Graves’ disease.

Methods

Based on a pre-specified protocol, we searched PubMed, EMBASE and Cochrane in July 2015 for case–control, controlled and randomized-controlled trials reporting risk factors for relapse after stopping ATD. The primary endpoint was relapse of disease until follow-up. PRISMA and SIGN statements were used for reviewing the data and assessing the quality of included trials.

Results

We included 54 trials with a total of 7595 participants. Most trials were small with moderate-to-high risk for bias. Ten trials were assessed only qualitatively (2227 patients), genomic data were reported in 13 trials (2178 patients) and 31 trials (4346 patients) were assessed quantitatively. In total, there were 3696 relapses in 7595 patients (48.7%). By using random-effects meta-analysis, orbitopathy, smoking, thyroid volume measured by sonography, goiter size, fT4, tT3, TRAb and TBII were significantly associated with relapse, whereas male vs female sex, age and initial tT4 level did not show significant associations.

Conclusions

This analysis found several risk factors to predict relapse in Graves’ disease, which can be combined in a risk score. Prospective studies should evaluate the prognostic accuracy of such a score to guide treatment decisions.

Introduction

Graves’ disease is the most common cause of primary hyperthyroidism in iodine-sufficient areas. The overall prevalence of Graves’ disease varies but is reported to be around 0.5% (1). Graves’ disease is characterized by the autoimmune production of thyrotropin-related antibodies (TRAb) stimulating the thyroidal cells and causing an overproduction of thyroid hormones (2). Furthermore, up to 50% of patients develop Graves’ orbitopathy by the co-expression of TSH receptors by pre-adipocytes within the orbital cavities inducing proliferation and thus proptosis (3).

One standard treatment for Graves’ disease includes the use of anti-thyroid drugs (ATD) for a total duration of 12–18 months. ATD may also have immunomodulating effects and positive impact on the course of Graves’ disease (4). This approach is generally preferred in Europe and Asia, and it offers the possibility of disease resolution without residuals (i.e., post-interventional hypothyroidism, worsening of orbitopathy) (5). Yet, risk for relapse is high (around 50%) and patients may suffer from side effects such as exanthema, pruritus, urticaria, hepatitis and agranulocytosis, especially those treated with higher doses. Due to the high risk of relapse, an alternative first-line treatment used (mostly in the United States) is radioactive iodine (RAI) ablation, a usually ablative treatment modality (3, 6).

Several risk factors predicting relapse after stopping ATD have been reported, e.g. younger age, large goiters, smoking, male sex, severe biochemical disease and higher levels of antibodies (4). Still, there is a lack of a relapse risk prediction rule, which could be helpful when counseling patients about the best first treatment option (i.e. ATD, surgery or RAI), subjecting them to a potentially unnecessary and time-consuming ATD treatment. Herein, we performed a systematic review and meta-analysis to study the possible pretreatment risk factors of relapse of patients with hyperthyroidism due to Graves’ disease.

Methods

Objective

The objective was to study pretreatment risk factor of disease recurrence after stopping ATD in patients with a first episode of hyperthyroidism due to Graves’ disease.

Protocol and eligibility criteria

This review adheres to PRISMA guidelines (7). First, we generated a review protocol (Supplementary file, see section on Supplementary data given at the end of this article), outlining the main hypothesis, outcomes and search strategy. We included case–control, observational, randomized-controlled and controlled trials that assigned patients with a first episode of Graves’ disease to a standard treatment with any ATD. The diagnosis of Graves’ disease had to be established by a low thyrotropin (TSH), a high free or total thyroxine (f/tT4) or triiodothyronine (f/tT3), and if available, diffuse goiter and positive TRAb. Participants had to be adults of at least 16 years of age. There were no restrictions on the type and dosage of ATD used (e.g. methimazole, carbimazole, propylthiouracil, etc.). Also, we had no restrictions regarding language of publication, publication type or date and publication status. If studies were not available in a public library, we contacted the corresponding author. We also did include trials assessing the addition of T4 or T3 to standard ATD treatment as a former systematic review has shown that this does not affect the recurrence rate (6).

We excluded trials with a follow-up of <12 months, as well as case reports, reviews and trials not reporting relapse or risk factors, our primary endpoints of interest.

Outcomes

The primary outcome was relapse of hyperthyroidism according to pretreatment risk factors within a follow-up of at least 12 months after stopping ATD. We did not define any secondary outcomes.

Search strategy and study selection

We searched the electronic databases PubMed/Medline, EMBASE and the Cochrane CENTRAL Library (all until end of July 2015). We also hand-searched the references of leading articles to identify further articles.

Our search strategy included following terms: Graves’ disease, relapse, recurrence, goiter, hyperthyroidism, Basedow (Supplementary file for further details). The articles found were transferred into EndNote Version X7 (Thomson Reuters, Philadelphia, PA, USA) and then screened by title and abstract. Full articles were retrieved for further assessment if the information given indicated that the study included patients with their first episode of Graves’ disease, reported pretreatment risk factors and had well-defined criteria for the diagnosis of Graves’ disease and relapse. If there was any doubt about the relevance of the article from the information given in the title and abstract, the full article was consulted for clarification. All articles were reviewed by two independent authors (HF and TS). In case of disagreements, a solution was obtained by consensus and if not possible by a third reviewer (PS) (Fig. 1).

Figure 1
Figure 1

Study flow diagram.

Citation: European Journal of Endocrinology 176, 1; 10.1530/EJE-16-0725

Data collection

For all included studies, a data extraction table was generated as suggested in the Cochrane Handbook for Systematic Reviews of Interventions (8). It was then tested on two pilot studies and refined afterward. Data were extracted by two authors (HF and CD) and cross-checked by another researcher (TS). Two review authors independently assessed trial quality (HF and TS). In case of disagreements, a solution was again obtained by consensus and if not possible by a third reviewer (PS). We suspected duplicate publications if the same authors published multiple studies on the same subject within two years. In November 2015, we tried to contact 12 trial authors (9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20) for suspected double publication of data and/or for missing data respectively. Four authors could not be contacted as there were no recent email addresses (15, 17, 19, 20). Two authors refused to share their data with us (10, 16). Until the end of February 2016, we received no answers from the other authors, although we contacted them twice.

Data items

From all included trials, we extracted the following data: (1) General information: first author, year of publication and country; (2) trial characteristics: design, if eligibility criteria are reported, duration, mode of randomization, blinding (assessors and patients) and intention-to-treat analysis; (3) interventions: basis intervention with ATD (days and dose) and pretreatment risk factors predicting outcome; (4) patients: number of patients included, sex and age; (5) outcomes: primary outcome, secondary outcome and effect size.

Risk assessment of bias in individual studies

All included studies were assessed by two reviewers according to the Scottish Intercollegiate Guidelines Network (SIGN) (21) (Supplementary file). Assessment was done using the appropriate checklist accordingly (i.e. RCT, cohort or case–control study). The risk of bias was judged to be unacceptable, high or low. The following aspects of internal and external validity were also assessed and reported: (1) Are there any eligibility criteria reported? (2) Mode of randomization (e.g. random number permutation table); (3) Participants blinding? (4) Assessor blinding? (5) Intention-to-treat analysis? (6) Outcome well defined and complete? (7) Mean follow-up in months; (8) Percentage of lost to follow-up?

Summary measures, planned methods of analysis and risk of bias across studies

Only risk factors reported in at least three studies were considered. Due to low number of studies, we omitted data on the following risk factors: urinary iodine, fT4I, fT3, thyroglobulin antibodies, scintigraphy, past history and family history. As none of the authors that we contacted about original data responded, we had to impute means and standard deviations for several studies as recommended, as only median and range had been reported (22). To impute a standard error for the within-study variation, Poisson distribution was assumed. To calculate thyroid volume when only a weight was given, we assumed a density of 1 g per mL (same density as water).

For dichotomous data, risk ratios with 95% CI were calculated by a random-effects model (23). For continuous data, we computed the standardized mean difference according to Hedges (g) (24) as effect size as measurements were on different scales. To facilitate clinical interpretation, we also calculated weighted raw mean differences (RMD). As test of heterogeneity, we computed the variation in risk ratio (RR) across studies attributable to heterogeneity (I2) (25). Additionally, we imputed mean values for all missing data on the relapse groups in all 31 studies using a multivariate truncated linear regression model (26). To check for internal validity, we performed univariate meta-regression and compared the results to the ones obtained by classical meta-analysis. Significant results from the univariate meta-regression were fitted into a multivariate meta-regression (27) while omitting co-linear variables. Categorical data were transformed into percentages before integration into regression models. Analyses for subgroup and sensitivity analyses were planned post hoc due to the heterogeneity of studies. Gene data were assessed qualitatively, and if sufficient data were reported, we calculated the odds ratios (OR).

As most data were available for the risk factor of age, we used these to check for small study effects. For each of these trials, we plotted the effect by the inverse of its standard error (28). The symmetry of these ‘funnel plots’ was assessed both visually and formally with Egger’s test to see if the effect decreased with increasing sample size.

The statistical analysis was conducted using Stata software version 12.1 (Stata Corp.). All significance tests were two-sided, and P < 0.05 was considered to be statistically significant.

Results

Study selection and study characteristics

A total of 1859 articles were evaluated by the accordingly skilled members of our study group (Fig. 1 according to suggestions of the PRISMA group (7)). Of these, 375 full-text articles were assessed for eligibility and 54 trials were included in the final data analysis (9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70). There were 31 trials (of which 5 randomized-controlled and 18 controlled trials) with 2322 relapses in a total of 4346 participants (overall relapse rate 53.4%) for quantitative meta-analysis (Supplementary Tables 1 and 2). Of note, the number of patients does not add up to the grand total as data reported were both used for the genomic data table and the quantitative analysis. Trials reporting genomic data were only evaluated qualitatively. Trials were published between 1977 and 2015. Mainly propylthiouracil and methimazole were used as ATDs.

Risk of bias within studies

Trials were mostly small with diverse risk of bias (10 high, 15 medium and 29 low risk of bias; Supplementary Table 3). Of the 31 trials included in the quantitative analysis, 12 were of low risk, 14 were of medium risk and 5 were of high risk of bias.

Qualitative reporting of gene data

Gene data were available from 13 of 54 studies (24.1%) with a total of 2178 patients being genotyped. Number of subjects ranged from 30 to 451 per trial, and trials had a moderate-to-low risk of bias (one high, five medium and seven low risk of bias; Supplementary Table 3). Results are summarized in Table 1 (9, 14, 20, 32, 35, 40, 41, 42, 51, 54, 57, 60, 70). The effect size for most markers were rather modest with OR around 1–2 in general. Many were below a size, which could be deemed clinically significant (i.e. >2), but with a wide range from 0.1 to 16.1.

Out of 35 studies testing HLA types, only 4 studies reported their result in detail. Mostly, negative or non-significant associations were omitted, apparently with high risk of publication bias. De Bruin (14) showed, contradictorily to others, no increased relapse rate for DR3-positive patients.

There were also investigations on the composition of peripheral leucocytes. One study did not find a difference in peripheral T-, B- and NK-lymphocyte subsets, to sufficiently predict relapse, even not after combination of factors (66). Van Ouwerkerk (34) reported the amount of subsets of T-lymphocytes (helper and suppressor type) without any predictive value.

Table 1

Associated genes for relapse.

Gene locusSNP polymorphism of gene and locationReceptor or transcriptAlleleEffect size (95% confidence interval) for risk of relapseReported by
Ch 6p21.31N/AHLA-DR3N/ARR 0.29 (P 0.016)/OR 0.1 (0.0; 0.7)de Bruin (1988), The Netherlands
HLA-Cw7RR 2.05 (P 0.009)/OR 5.0 (1.4; 17.7)
Ch 6p21.31N/AHLA-A1N/AOR 0.5 (0.2; 1.3)Allanic (1983), France
HLA-B8OR 1.1 (0.4; 3.1)
HLA-DR3OR 1.8 (0.6; 4.9)
Ch 6p21.31N/AHLA-A1N/AOR 2.2 (0.8; 6.1)Irvine (1977), UK
HLA-B8OR 3.3 (1.2; 9.6)
Ch 6p21.31N/AHLA-B12N/ARR 0.23 (P 0.02)/OR 0.2 (0.0; 1.1)Schifferdecker (1991), Germany
HLA-DR5RR 9.85 (P 0.03)/OR 9.9 (1.2; 438.0)
HLA-DR7RR 0.27 (P 0.05)/OR 0.3 (0.1; 1.3)
Ch 6p21.31N/AHLA-DR3N/AOR 1.2 (0.7; 2.4)Schleusener (1987), Germany
Ch 6p21.31N/AHLA-DR3N/AOR 0.8 (0.5; 1.4)Schleusener (1989), Germany
Ch 6p21.31N/AHLA-DR3N/AOR 2.2 (0.4; 12.3)Wilson (1986), Scotland
Ch 20q12–q13.25′ UTRCD40CCOR 1.2 (0.5; 2.9)Kim (2007), South Korea
CTOR 0.6 (0.3; 1.4)
TTOR 3.7 (0.5; 166.2)
Ch 20q12–q13.2Intron 1 (rs745307)CD 40C/T vs C/COR 8.0 (1.0; 62.7)Wang (2013), Taiwan
Intron 1 (rs11569309)C/T vs T/TOR 8.1 (1.0; 63.7)
Intron 8 (rs3765457)A/G vs A/AOR 2.6 (1.1; 6.3)
Ch 2p33Position +49A/G in exon 1 (rs231775)CTLA4G/G vs A/G and A/AOR 2.0 (1.2; 3.3)Wang (2013), Taiwan
Ch 2p33CT60 polymorphism in 3′UTRCTLA4AAOR 0.7 (0.0; 42.2)Kim (2007), South Korea
AGOR 1.1 (0.4; 3.5)
GGOR 0.9 (0.3; 2.5)
Ch 2p33Promoter at position −318CTLA4CCOR 1.1 (0.4; 2.7)Kim (2007), South Korea
CTOR 0.9 (0.4; 2.2)
TTOR 0.4 (0.0; 31.9)
Ch 2p33Position +49 in exon 1CTLA4AAOR 1.9 (0.2; 96.2)Kim (2007), South Korea
AGOR 1.5 (0.4; 5.5)
GGOR 0.5 (0.1; 1.6)
Ch 2p33Position +49 in exon 1CTLA4AAOR 1.2 (0.2; 5.2)Wang (2004), Taiwan
AGOR 0.3 (0.1; 0.7)
GGOR 4.0 (1.9; 8.6)
Ch 2p33Position 49 in exon 1CTLA4AAOR 0.3 (0.1; 0.8)Tanrikulu (2006), Turkey
AGOR 6.2 (2.4; 16.6)
GGOR 3.6 (0.8; 5.7)
Ch 8q24Exon 33ThyroglobulinCCOR 1.5 (0.6; 3.6)Tanrikulu (2006), Turkey
CTOR 1.4 (0.6; 3.4)
TTOR 2.1 (0.7; 6.5)
Ch 20q13.2–q13.3T393C polymorphismG alpha subunit of G protein (GNAS1)CCOR 1.3 (0.7; 2.7)Glowacka (2009), Germany
CTOR 0.5 (0.3; 0.9)
TTOR 2.4 (1.0; 6.2)
Ch 9p24.1Position 8923 in PD-L1 intron 4Programmed cell death-1 ligandAAOR 0.4 (0.1; 1.4)Hayashi (2008), Japan
ACOR 1.5 (0.3; 5.0)
CCOR 16.1 (0.8; 954.2)

Quantitative synthesis of results

Random-effects meta-analysis did show a significant predictive value for orbitopathy RR 1.16 (1.08; 1.25), smoking RR 1.13 (1.02; 1.25), thyroid volume assessed by sonography g 0.54 (0.18; 0.90)/RMD 5.46 mL (1.92; 9.00), goiter size according to WHO grades g 0.18 (0.07; 0.30), fT4 g 0.16 (0.05; 0.28)/RMD 3.98 pM (0.53; 7.44), tT3 g 0.25 (0.04; 0.46)/RMD 0.59 nM (0.10; 1.09), TRAb g 0.33 (0.04; 0.62)/RMD 16.93 U/L (1.94; 31.93), TBII g 0.36 (0.09; 0.63)/RMD 8.27% (2.61; 13.92), TSAb g 0.26 (0.06; 0.45)/RMD 127.39% (18.01; 236.77), whereas male vs female sex RR 0.91 (0.82; 1.00), age g −0.14 (−0.35; 0.07)/RMD −0.80 year (−2.54; 0.94), and initial tT4 level g −0.04 (−0.31; 0.23)/RMD 2.11 nM (−13.84; 18.06) did not show significant associations (Table 2).

Table 2

Summary of results.

Pretreatment risk factors for relapse
Classical meta-analysisUnivariable meta-regressionMultivariable meta-regression
Risk factorRR or SMD/Hedges’ g (95% CI)Weighted raw mean difference (95% CI)Heterogeneity (I2) (%)% Risk (95% CI)P (adj. for mult. testing)% Risk (95% CI)P (adj. for mult. testing)Miscellaneous
Male vs female sexRR 0.91 (0.82; 1.00)N/A40.30178.75 (100.80; 256.71)<0.00113.77 (−38.78; 66.33)0.59Residual variance due to heterogeneity (I2)

0.00%
Age (years)g −0.14 (−0.35; 0.07)−0.80 (−2.54; 0.94)85.301.62 (0.04; 3.20)0.05N/AN/A
Graves’ orbitopathy (y/n)RR 1.16 (1.08; 1.25)*N/A0.0089.31 (56.67; 121.95)<0.0016.65 (−18.62; 31.93)0.54Proportion of between-study variance explained (adjusted R2)
Goiter (WHO grades)g 0.18 (0.07; 0.30)*N/A0.00−64.85 (−90.33; −39.37)0.00213.48 (5.47; 21.50)0.002100%
Smoking (y/n)RR 1.13 (1.02; 1.25)*N/A7.8051.19 (−176.54; 278.92)0.21N/AN/A
Thyroid vol. (mL)g 0.54 (0.18; 0.90)*5.46 (1.92; 9.00)75.600.74 (0.03; 1.46)0.04N/AN/A
fT4 (pmol/L)g 0.16 (0.05; 0.28)*3.98 (0.53; 7.44)2.000.69 (0.08; 1.31)0.03N/AN/A
tT4 (nmol/L)g −0.04 (−0.31; 0.23)2.11 (−13.84; 18.06)83.50−0.47 (−0.65; −0.28)<0.0010.02 (−0.11; 0.15)0.74
tT3 (nmol/L)g 0.25 (0.04; 0.46)*0.59 (0.10; 1.09)66.70−1.87 (−5.73; 1.99)0.33N/AN/A
TRAb (U/L)g 0.33 (0.04; 0.62)*16.93 (1.94; 31.93)75.200.41 (0.29; 0.53)<0.0010.09 (−0.04; 0.22)0.17
TBII (% activity)g 0.36 (0.09; 0.63)*8.27 (2.61; 13.92)75.800.08 (−0.43; 0.59)0.74N/AN/A
TSAb (% activity)g 0.26 (0.06; 0.45)*127.39 (18.01; 236.77)46.40−0.07 (−0.08; −0.06)<0.001−0.01 (−0.05; 0.03)0.68

P < 0.05; TRAb (thyrotropin receptor antibodies), TBII (thyrotropin-binding inhibiting immunoglobulins), TSAb (thyroid-stimulating antibody), f/tT3/4 (free/total triiodothyronine/thyroxine), thyroid volume (assessed by sonography).

As we computed standardized mean difference according to Hedges, values are comparable across different units of measurement. This gives a crude idea of effect size between different factors. A large drawback of this method is its difficult interpretation. Generally, effect sizes with g values around 0.1 are viewed as small, around 0.3 as medium and above 0.5 as strong (24).

For better clinical comparison, we calculated raw mean differences. Thus, an increase of the units depicted above results in a 1% increase in relapse risk (e.g. thyroid volume +1% relapse risk for every 5.46 mL increase).

Additionally, we depicted the two risk factors age and goiter stratified by the subgroups relapse vs remission for age and WHO grades for goiter (Fig. 2, additional figures in Supplementary file).

Figure 2
Figure 2

(A) Relapse risk according to age at diagnosis. (B) Relapse risk according to WHO goiter grade.

Citation: European Journal of Endocrinology 176, 1; 10.1530/EJE-16-0725

In univariable meta-regression values for TBII, tT3, fT4, sonographic thyroid volume and smoking status became non-significant, whereas tT4 turned significant. As this might be due to imputation, we refitted the models with the original data, but observed no difference.

To further check for internal validity, we refitted models with values for remission. These models showed similar results except for the gender of participants were there was a collinear increase of remission and relapse with both sexes. This might be explained by the low statistical power of meta-regression and the small effect of sex on relapse risk. Also, age and tT3 showed contradictory effect sizes, which we interpreted were due to small effect size and scarcity of data respectively. These results and measurement of heterogeneity by I 2 fitted well with the classical meta-analysis, except the ones noted previously.

In multivariate meta-regression, we first fitted a model with all significant covariates from univariate regression. As these results could be influenced by imputation, we refitted the models with the original data, but observed no difference again. Therefore, we concluded that influential, biased studies attributed to these observed effects.

As data were handled as percentages, Table 2 shows a percentage increase or decrease for every unit of risk factor respectively. As an example, it can be depicted that a male with Graves’ orbitopathy and goiter grade II has a 47% increased risk for relapse (cumulative risk = 14% + 7% + (2 × 13%) = 47%) compared to a female patient without orbitopathy and no goiter.

Risk of bias across studies

As there was strong evidence for heterogeneity for most risk factors, we computed a funnel plot (Supplementary Fig. 2) for age as a risk factor showing slight asymmetry. We chose age as most data were available on this risk factor. Egger’s test for no small study effects did show a significant result, thus rejecting publication bias across studies (P = 0.006). Of course, these results may also be influenced by small study effects, poor design of the included studies and true heterogeneity considering the genetic differences in the studied populations (28).

Discussion

Summary of evidence

This is the first systematic review and meta-analysis focusing on pretreatment risk factors predicting relapse before the first treatment course with ATD in patients with hyperthyroidism due to Graves’ disease. The 54 included trials had moderate-to-low trial quality and risk of bias. The pooled analysis showed a good prediction of relapse with several risk factors, but in most trials, these factors by themselves had no enough power. Risk factors were similar in randomized trials compared to case-controlled and controlled trials.

Graves’ disease is an autoimmune disease caused by the stimulation of antibodies directed against the thyrotropin receptor in most patients (4). The risk for disease occurrence depends by part on environmental factors such as smoking and psychological factors such as stressful life events, and by part on genetic factors as previously demonstrated in twin studies (4, 71). Although environmental factors for relapse could be altered, genetic factors could not be directly influenced. But it would be possible to modify the immune system’s reaction by immunosuppressive drugs (72) until the environmental co-factors subside.

Our analysis has two main clinical benefits. First, it offers clinicians an up-to-date qualitative overview of genetic factors and other risk factor associated with the recurrence of Graves’ disease. Second, the quantitative analysis also allows to calculate the risk for relapse in an individual patient. Such numbers may be used when discussing the therapeutic options with patients, i.e. whether medical therapy or more definite treatment options (radio-iodine and surgery) should be used. Still, the effect sizes for most risk factors assessed in our analysis were small and statistically non-significant. Thus, calculation of relapse risk based on this analysis should be used with caution and only in conjunction with good clinical judgment. Also, a prospective study should validate our findings before more widespread use in clinical practice.

Strengths and limitations

The strengths of our study are the predefined study protocol adhering to PRISMA guidelines for systematic reviews, the inclusion from studies all around the globe without language restrictions and thus reflecting different ethnicities and genetic backgrounds (2) and the relevance of this topic in every day practice. As one might argue that the response to ATD depends on genetic background, our robust findings of the included trials from different countries suggest otherwise. This increases the clinical applicability and external validity. Although we do not provide insights into new risk factor and there have already been various systematic reviews assessing the different treatment options for Graves’ disease (1, 5, 6, 73), we provide the very first systematic review on pretreatment risk factors so far.

We are aware of several limitations in regard to this analysis. First, retrieved data were very heterogeneous and authors of eligible studies could either not be contacted or refused to share their date with us. Thus, we had to rely on several imputation methods, which could have influenced our results. For our data set, we assumed that data were missing at random. A complete-case analysis would introduce bias and lower the power of the analysis. By using multiple imputation, we assume that bias is minimal as the approach is generally considered to be conservative. Still, imputation has limitations particularly if data are not normally distributed.

Secondly, trials were mostly monocentric, heterogeneous and a lot had a rather high risk for bias. Many of the included studies were not randomized or case-controlled trials. Thirdly, there were many different laboratory assays over a wide time span of 38 years, making comparability difficult, especially for the immunological antibody assays. In addition, only two studies had blinding of participants or assessors (placebo controlled). No trial had an intention-to-treat analysis performed and adverse effects were rarely, if at all, systematically reported. Still, studies reported eligibility criteria, and mean follow-up was at least one year after the withdrawal of ATD, mainly 24 months. Notably, most relapses occur within the first two years. Still, studies reported eligibility criteria and mean follow-up was at least one year after the withdrawal of ATD, mainly 24 months. Heterogeneity of data did not allow for formal quantitative analysis (e.g. mean time to relapse). Hence, we qualitatively present these data without formally analyzing it (Supplementary Table 2). Roughly 2/3 of the relapses occurred between months 6 and 18 after stopping ATD, being in line with current literature (1).

Meta-regression is prone to over-interpretation of findings. However, in our case, results were similar to classical meta-analysis. Furthermore, by adjusting P values for multiple testing, we are certain that the results reflect true findings.

Overall, the markers identified…’ for clarity., the markers identified by our systematic review, either alone or combined, were not strong enough to predict the clinical outcome of a single patient. Ideally, fresh markers would fill this gap. In this regard, a very recent publication from the Netherlands might shed new light. The authors generated a risk score based on a multivariate analysis of 173 prospectively enrolled patients subjected to their first ATD course. They identified lower age, higher fT4, higher TBII, large goiters, PTPN22 C/T SNP and HLA types DQB1*02, DQA1*05 and DRB1*03 as independent risk factors (74). Still, these results need to be externally validated and replicated in a larger cohort first. However, this approach would lead to a new perspective and could be the first step to a form of individualized treatment in Graves’ disease.

Conclusions

In conclusion, this systematic review and meta-analysis found several pretreatment risk factors of relapse in patients with standard treatment of Graves’ disease. The occurrence of orbitopathy, smoking, larger thyroid volume and biochemically more severe disease (fT4, tT3, TRAb, TBII and TSAb levels) did show a higher risk of relapse.

Yet, the small predictive power of a single risk factor is insufficient to predict the outcome of a single patient. This calls for novel markers to be identified and tested in future trials.

Supplementary data

This is linked to the online version of the paper at http://dx.doi.org/10.1530/EJE-16-0725.

Declaration of interest

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.

Funding

This study was supported in part by the Swiss National Science Foundation (SNSF Professorship, PP00P3_150531/1) and the Research Council of the Kantonsspital Aarau (1410.000.044). Funders had no role in the design, analysis or writing of this article.

Author contribution statement

T S, B M and P S designed the study and H F and C D analyzed the data. H F and T S wrote the first draft of the manuscript and had primary responsibility for the final content. All authors read and approved the final manuscript.

Acknowledgments

The authors want to express their gratitude to Alexander Litke, MD, and Merih Guglielmetti for their translation of Russian and Turkish articles respectively.

References

  • 1

    Genovese BM, Noureldine SI, Gleeson EM, Tufano RP & Kandil E. What is the best definitive treatment for Graves’ disease? A systematic review of the existing literature. Annals of Surgical Oncology 2013 20 660667. (doi:10.1245/s10434-012-2606-x)

    • Search Google Scholar
    • Export Citation
  • 2

    Antonelli A, Ferrari SM, Corrado A, Di Domenicantonio A & Fallahi P. Autoimmune thyroid disorders. Autoimmunity Reviews 2015 14 174180. (doi:10.1016/j.autrev.2014.10.016)

    • Search Google Scholar
    • Export Citation
  • 3

    Orgiazzi J. Thyroid autoimmunity. Presse Médicale 2012 41 25. (doi:10.1016/j.lpm.2012.10.002)

  • 4

    Franklyn JA & Boelaert K. Thyrotoxicosis. Lancet 2012 379 11551166. (doi:10.1016/S0140-6736(11)60782-4)

  • 5

    Sundaresh V, Brito JP, Wang Z, Prokop LJ, Stan MN, Murad MH & Bahn RS. Comparative effectiveness of therapies for Graves’ hyperthyroidism: a systematic review and network meta-analysis. Journal of Clinical Endocrinology and Metabolism 2013 98 36713677. (doi:10.1210/jc.2013-1954)

    • Search Google Scholar
    • Export Citation
  • 6

    Abraham P, Avenell A, McGeoch SC, Clark LF & Bevan JS. Antithyroid drug regimen for treating Graves’ hyperthyroidism. Cochrane Database of Systematic Reviews 2010 CD003420. (doi:10.1002/14651858.CD003420.pub4)

    • Search Google Scholar
    • Export Citation
  • 7

    Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JPA, Clarke M, Devereaux PJ, Kleijnen J & Moher D. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Medicine 2009 6 e1000100. (doi:10.1371/journal.pmed.1000100)

    • Search Google Scholar
    • Export Citation
  • 8

    Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0, updated 2011. (available at: http://handbook.cochrane.org/). Accessed on 16 October 2015.

  • 9

    Wang PW, Liu RT, Juo SHH, Wang ST, Hu YH, Hsieh CJ, Chen MH, Chen IY & Wu CL. Cytotoxic T lymphocyte-associated molecule-4 polymorphism and relapse of Graves’ hyperthyroidism after antithyroid withdrawal. Journal of Clinical Endocrinology and Metabolism 2004 89 169173. (doi:10.1210/jc.2003-030854)

    • Search Google Scholar
    • Export Citation
  • 10

    Orunesu E, Bagnasco M, Salmaso C, Altrinetti V, Bernasconi D, Del Monte P, Pesce G, Marugo M & Mela GS. Use of an artificial neural network to predict Graves’ disease outcome within 2 years of drug withdrawal. European Journal of Clinical Investigation 2004 34 210217. (doi:10.1111/j.1365-2362.2004.01318.x)

    • Search Google Scholar
    • Export Citation
  • 11

    Sonnet E, Massart C, Gibassier J, Allannic H & Maugendre D. Longitudinal study of soluble intercellular adhesion molecule-1 (ICAM-1) in sera of patients with Graves’ disease. Journal of Endocrinological Investigation 1999 22 430435. (doi:10.1007/bf03343586)

    • Search Google Scholar
    • Export Citation
  • 12

    Vitti P, Rago T, Chiovato L, Pallini S, Santini F, Fiore E, Rocchi R, Martino E & Pinchera A. Clinical features of patients with Graves’ disease undergoing remission after antithyroid drug treatment. Thyroid 1997 7 369375. (doi:10.1089/thy.1997.7.369)

    • Search Google Scholar
    • Export Citation
  • 13

    Massart C, Guilhem I, Gibassier J, Allannic H & Nicol M. Comparison of thyroperoxidase and microsomal antibody assays in sera from patients with Graves disease. Clinical Chemistry 1991 37 17771780.

    • Search Google Scholar
    • Export Citation
  • 14

    de Bruin TW, Bolk JH, Bussemaker JK, Stijnen T, Schreuder GM, de Vries RR, van der Heide D. Graves’ disease: immunological and immunogenetic indicators of relapse. British Medical Journal (Clinical Research Edition) 1988 296 12921295. (doi:10.1136/bmj.296.6632.1292)

    • Search Google Scholar
    • Export Citation
  • 15

    Kimball LE, Kulinskaya E, Brown B, Johnston C & Farid NR. Does smoking increase relapse rates in Graves’ disease? Journal of Endocrinological Investigation 2002 25 152157. (doi:10.1007/bf03343979)

    • Search Google Scholar
    • Export Citation
  • 16

    Laurberg P, Buchholtz Hansen PE & Iversen E. Goitre size and outcome of medical treatment of Graves’ disease. Acta Endocrinologica 1986 111 3943. (doi:10.1530/acta.0.1110039)

    • Search Google Scholar
    • Export Citation
  • 17

    Takaichi Y, Tamai H, Honda K, Nagai K, Kuma K & Nakagawa T. The significance of antithyroglobulin and antithyroidal microsomal antibodies in patients with hyperthyroidism due to Graves’ disease treated with antithyroidal drugs. Journal of Clinical Endocrinology and Metabolism 1989 68 10971100. (doi:10.1210/jcem-68-6-1097)

    • Search Google Scholar
    • Export Citation
  • 18

    Yamada T, Aizawa T, Koizumi Y, Komiya I, Ichikawa K & Hashizume K. Age-related therapeutic response to antithyroid drug in patients with hyperthyroid Graves’ disease. Journal of the American Geriatrics Society 1994 42 513516. (doi:10.1111/j.1532-5415.1994.tb04973.x)

    • Search Google Scholar
    • Export Citation
  • 19

    Khanna CM, Shankar LR, Jaggi CB, Bansal JK & Chugh P. Predictor of outcome of hyperthyroidism due to Graves disease: serum triiodothyronine/thyroxine ratio. Journal of the Association of Physicians of India 1996 44 98101.

    • Search Google Scholar
    • Export Citation
  • 20

    Wilson R, McKillop JH & Henderson N. The ability of the serum thyrotrophin receptor antibody (TRAb) index and HLA status to predict long-term remission of thyrotoxicosis following medical therapy for Graves’s disease. Clinical Endocrinology 1986 25 151156. (doi:10.1111/j.1365-2265.1986.tb01676.x)

    • Search Google Scholar
    • Export Citation
  • 21

    Scottish Intercollegiate Guidelines Network (SIGN), 2015. (available at: http://www.sign.ac.uk/index.html). Accessed on 16 October 2015.

  • 22

    Leon-Guerrero A & Frankfort-Nachmias C. Essentials of Social Statistics for a Diverse Society. Thousand Oaks, CA, USA: SAGE, 2012.

  • 23

    DerSimonian R & Laird N. Meta-analysis in clinical trials revisited. Contemporary Clinical Trials 2015 45 (Part A) 139145. (doi:10.1016/j.cct.2015.09.002)

    • Search Google Scholar
    • Export Citation
  • 24

    Borenstein M, Hedges LV, Higgins JPT & Rothstein HR. Introduction to Meta-Analysis. Chichester, UK: Wiley, 2010.

  • 25

    Higgins JPT, Thompson SG, Deeks JJ & Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003 327 557560. (doi:10.1136/bmj.327.7414.557)

  • 26

    Raghunathan TE, Lepkowski JM, van Hoewyk J & Solenberger P. A multivariate technique for multiply imputing missing values using a sequence of regression models. Survey Methodology 2001 27 8596.

    • Search Google Scholar
    • Export Citation
  • 27

    Thompson SG & Higgins JPT. How should meta-regression analyses be undertaken and interpreted? Statistics in Medicine 2002 21 15591573. (doi:10.1002/sim.1187)

    • Search Google Scholar
    • Export Citation
  • 28

    Egger M, Davey Smith G, Schneider M & Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997 315 629634. (doi:10.1136/bmj.315.7109.629)

    • Search Google Scholar
    • Export Citation
  • 29

    Yamada T, Komiya I, Miyahara Y, Komatsu M, Shima I, Inazawa T & Aizawa T. Effect of methimazole treatment for 2 years on circulating IL-4, IgE, TBII, and TSAb in patients with hyperthyroid Graves’ disease. Endocrine Journal 2006 53 783788. (doi:10.1507/endocrj.K06-054)

    • Search Google Scholar
    • Export Citation
  • 30

    Wilson R, McKillop JH & Pearson DWM. Relapse of Graves’ disease after medical therapy: predictive value of thyroidal technetium-99m uptake and serum thyroid stimulating hormone receptor antibody levels. Journal of Nuclear Medicine 1985 26 10241028.

    • Search Google Scholar
    • Export Citation
  • 31

    Wang PW, Liu RT, Tung SC, Chien WY, Lu YC, Chen CH, Kuo MC, Hsieh JR & Wang ST. Outcome of Graves’ disease after antithyroid drug treatment in Taiwan. Journal of the Formosan Medical Association 1998 97 619625.

    • Search Google Scholar
    • Export Citation
  • 32

    Wang PW, Chen IY, Juo SH, Hsi E, Liu RT & Hsieh CJ. Genotype and phenotype predictors of relapse of graves’ disease after antithyroid drug withdrawal. European Thyroid Journal 2013 1 251258. (doi:10.1159/000342621)

    • Search Google Scholar
    • Export Citation
  • 33

    Vitti P, Rago T, Mancusi F, Pallini S, Tonacchera M, Santini F, Chiovato L, Marcocci C & Pinchera A. Thyroid hypoechogenic pattern at ultrasonography as a tool for predicting recurrence of hyperthyroidism after medical treatment in patients with Graves’ disease. Acta Endocrinologica 1992 126 128131.

    • Search Google Scholar
    • Export Citation
  • 34

    van Ouwerkerk BM, Krenning EP & Docter R. Cellular and humoral immunity in patients with hyperthyroid Graves’ disease before, during and after antithyroid drug treatment. Clinical Endocrinology 1987 26 385394. (doi:10.1111/j.1365-2265.1987.tb00795.x)

    • Search Google Scholar
    • Export Citation
  • 35

    Tanrikulu S, Erbil Y, Ademoglu E, Issever H, Barbaros U, Kutluturk F, Ozarmagan S & Tezelman S. The predictive value of CTLA-4 and Tg polymorphisms in the recurrence of Graves’ disease after antithyroid withdrawal. Endocrine 2006 30 377381. (doi:10.1007/s12020-006-0017-0)

    • Search Google Scholar
    • Export Citation
  • 36

    Takamatsu J, Kuma K & Mozai T. Serum triiodothyronine to thyroxine ratio: a newly recognized predictor of the outcome of hyperthyroidism due to Graves’ disease. Journal of Clinical Endocrinology and Metabolism 1986 62 980983. (doi:10.1210/jcem-62-5-980)

    • Search Google Scholar
    • Export Citation
  • 37

    Stefanic M & Karner I. Thyroid peroxidase autoantibodies are associated with a lesser likelihood of late reversion to hyperthyroidism after successful non-ablative treatment of Graves’ disease in Croatian patients. Journal of Endocrinological Investigation 2014 37 7177. (doi:10.1007/s40618-013-0026-9)

    • Search Google Scholar
    • Export Citation
  • 38

    Soveid M, Shaabani A, Ghaedi Gh H, Jafari SM & Omrani G. Prognostic factors in the relapse of Graves’ disease following treatment with antithyroid drugs. Iranian Journal of Medical Sciences 2003 28 106110.

    • Search Google Scholar
    • Export Citation
  • 39

    Shahbazian H, Saiedinia S & Aleali AM. Outcome of graves disease after anti-thyroid drug treatment in south west of Iran. Pakistan Journal of Medical Sciences 2012 28 445449.

    • Search Google Scholar
    • Export Citation
  • 40

    Schleusener H, Schwander J, Fischer C, Holle R, Holl G, Badenhoop K, Hensen J, Finke R, Bogner U & Mayr WR et al. Prospective multicentre study on the prediction of relapse after antithyroid drug treatment in patients with Graves’ disease. Acta Endocrinologica 1989 120 689701. (doi:10.1530/acta.0.1200689)

    • Search Google Scholar
    • Export Citation
  • 41

    Schleusener H, Schwander J, Holl G, Badenhoop K, Hensen J, Finke R, Schernthaner G, Mayr WR & Kotulla P. Do HLA-DR-typing and measurement of TSH-receptor antibodies help in the prediction of the clinical course of Graves’ thyrotoxicosis after antithyroid drug treatment? Acta Endocrinologica: Supplementum 1987 281 318324.

    • Search Google Scholar
    • Export Citation
  • 42

    Schifferdecker E, Kuhnl P, Schoffling K, Manfras B, Holzberger G, Spielmann W & Bohm BO. Immunogenetic markers in patients with Graves’ disease. Klinische Wochenschrift 1991 69 256260. (doi:10.1007/BF01666851)

    • Search Google Scholar
    • Export Citation
  • 43

    Scazziga BR. Drug therapy of Basedow’s disease. Long term functional results (302 cases). Schweizerische Medizinische Wochenschrift 1977 107 423427.

    • Search Google Scholar
    • Export Citation
  • 44

    Romaldini JH, Bromberg N, Werner RS, Tanaka LM, Rodrigues HF, Werner MC, Farah CS & Reis LC. Comparison of effects of high and low dosage regimens of antithyroid drugs in the management of Graves’ hyperthyroidism. Journal of Clinical Endocrinology and Metabolism 1983 57 563570. (doi:10.1210/jcem-57-3-563)

    • Search Google Scholar
    • Export Citation
  • 45

    Prakash R. Prediction of remission in Graves’ disease treated with long-term carbimazole therapy: evaluation of technetium-99m thyroid uptake and TSH concentrations as prognostic indicators. European Journal of Nuclear Medicine 1996 23 118122. (doi:10.1007/BF01731833)

    • Search Google Scholar
    • Export Citation
  • 46

    Park SM, Cho YY, Joung JY, Sohn SY, Kim SW & Chung JH. Excessive iodine intake does not increase the recurrence rate of Graves’ disease after withdrawal of the antithyroid drug in an iodine-replete area. European Thyroid Journal 2015 4 3642. (doi:10.1159/000375261)

    • Search Google Scholar
    • Export Citation
  • 47

    Orhan Y, Azezli A, Aral F, Tascioglu C, Molvalilar S & Sencer E. The frequency of remission and relapse in Graves’ disease was investigated. Istanbul Üniversitesi Tıp Fakültesi Mecmuası 1992 55 525530.

    • Search Google Scholar
    • Export Citation
  • 48

    Murakami M, Koizumi Y, Aizawa T, Yamada T, Takahashi Y, Watanabe T & Kamoi K. Studies of thyroid function and immune parameters in patients with hyperthyroid Graves’ disease in remission. Journal of Clinical Endocrinology and Metabolism 1988 66 103108. (doi:10.1210/jcem-66-1-103)

    • Search Google Scholar
    • Export Citation
  • 49

    Lucas A, Salinas I, Rius F, Pizarro E, Granada ML, Foz M & Sanmarti A. Medical therapy of Graves’ disease: does thyroxine prevent recurrence of hyperthyroidism? Journal of Clinical Endocrinology and Metabolism 1997 82 24102413. (doi:10.1210/jcem.82.8.4118)

    • Search Google Scholar
    • Export Citation
  • 50

    Komiya I, Yamada T, Sato A, Kouki T, Nishimori T & Takasu N. Remission and recurrence of hyperthyroid Graves’ disease during and after methimazole treatment when assessed by IgE and interleukin 13. Journal of Clinical Endocrinology and Metabolism 2001 86 35403544. (doi:10.1210/jcem.86.8.7734)

    • Search Google Scholar
    • Export Citation
  • 51

    Kim KW, Park YJ, Kim TY, Park DJ, Park KS & Cho BY. Susceptible alleles of the CD40 and CTLA-4 genes are not associated with the relapse after antithyroid withdrawal in Graves’ disease. Thyroid 2007 17 12291234. (doi:10.1089/thy.2007.0011)

    • Search Google Scholar
    • Export Citation
  • 52

    Kim TY, Park YJ, Park DJ, Chung H-K, Kim WB, Kohn LD & Cho BY. Epitope heterogeneity of thyroid-stimulating antibodies predicts long-term outcome in Graves’ patients treated with antithyroid drugs. Journal of Clinical Endocrinology and Metabolism 2003 88 117124. (doi:10.1210/jc.2002-020389)

    • Search Google Scholar
    • Export Citation
  • 53

    Jonas M, Ambroziak U, Bednarczuk T & Nauman J. Predicting a relapse of Graves’ hyperthyroidism in adults during the early phase of treatment with anti-thyroid drugs. Endokrynologia Polska 2006 57 596604.

    • Search Google Scholar
    • Export Citation
  • 54

    Irvine WJ, Gray RS, Morris PJ & Ting A. Correlation of HLA and thyroid antibodies with clinical course of thyrotoxicosis treated with antithyroid drugs. Lancet 1977 2 898900. (doi:10.1016/S0140-6736(77)90833-9)

    • Search Google Scholar
    • Export Citation
  • 55

    Hou X, Li Y, Li J, Wang W, Fan C, Wang H, Zhang H, Shan Z & Teng W. Development of thyroid dysfunction and autoantibodies in Graves’ multiplex families: an eight-year follow-up study in Chinese Han pedigrees. Thyroid 2011 21 13531358. (doi:10.1089/thy.2011.0035)

    • Search Google Scholar
    • Export Citation
  • 56

    Hoermann R, Quadbeck B, Roggenbuck U, Szabolcs I, Pfeilschifter J, Meng W, Reschke K, Hackenberg K, Dettmann J & Prehn B et al. Relapse of Graves’ disease after successful outcome of antithyroid drug therapy: Results of a prospective randomized study on the use of levothyroxine. Thyroid 2002 12 11191128. (doi:10.1089/105072502321085225)

    • Search Google Scholar
    • Export Citation
  • 57

    Hayashi M, Kouki T, Takasu N, Sunagawa S & Komiya I. Association of an A/C single nucleotide polymorphism in programmed cell death-ligand 1 gene with Graves’ disease in Japanese patients. European Journal of Endocrinology 2008 158 817822. (doi:10.1530/EJE-07-0649)

    • Search Google Scholar
    • Export Citation
  • 58

    Hansen PEB. The recurrence-rate following thyrostatic treatment of patients with hyperthyroidism and diffuse goitre (Graves’ disease). The relationship to the concentrations of inorganic iodine in the plasma. Ugeskrift for Laeger 1983 145 980982.

    • Search Google Scholar
    • Export Citation
  • 59

    Guilhem I, Massart C, Poirier JY & Maugendre D. Differential evolution of thyroid peroxidase and thyrotropin receptor antibodies in Graves’ disease: thyroid peroxidase antibody activity reverts to pretreatment level after carbimazole withdrawal. Thyroid 2006 16 10411045. (doi:10.1089/thy.2006.16.1041)

    • Search Google Scholar
    • Export Citation
  • 60

    Glowacka D, Loesch C, Johnson KT, Mann K, Esser J, Morgenthaler NG, Siffert W, Schmid KW & Eckstein AK. The T393C polymorphism of the Galphas gene (GNAS1) is associated with the course of Graves’ disease. Hormone and Metabolic Research 2009 41 430435. (doi:10.1055/s-0029-1220902)

    • Search Google Scholar
    • Export Citation
  • 61

    Glinoer D, Nayer P de, Bex M. Effects of l-thyroxine administration, TSH-receptor antibodies and smoking on the risk of recurrence in Graves’ hyperthyroidism treated with antithyroid drugs: a double-blind prospective randomized study. European Journal of Endocrinology/European Federation of Endocrine Societies 2001 144 475483. (doi:10.1530/eje.0.1440475)

    • Search Google Scholar
    • Export Citation
  • 62

    Edmonds CJ & Tellez M. Treatment of Graves’ disease by carbimazole: high dose with thyroxine compared to titration dose. European Journal of Endocrinology/European Federation of Endocrine Societies 1994 131 120124. (doi:10.1530/eje.0.1310120)

    • Search Google Scholar
    • Export Citation
  • 63

    Eckstein AK, Lax H, Losch C, Glowacka D, Plicht M, Mann K, Esser J & Morgenthaler NG. Patients with severe Graves’ ophthalmopathy have a higher risk of relapsing hyperthyroidism and are unlikely to remain in remission. Clinical Endocrinology 2007 67 607612. (doi:10.1111/j.1365-2265.2007.02933.x)

    • Search Google Scholar
    • Export Citation
  • 64

    Duprey J & Colliard M. Results of treatment of Graves’ disease by antithyroid drugs. Report of 90 cases (author’s transl). Annales d’endocrinologie 1981 42 233249.

    • Search Google Scholar
    • Export Citation
  • 65

    Dauksiene D, Dauksa A & Mickuviene N. Independent pretreatment predictors of Graves’ disease outcome. Medicina 2013 49 427434.

  • 66

    Corrales JJ, Lopez A, Ciudad J & Orfao A. The distribution of the major peripheral blood T, B and NK cell subsets does not predict the clinical outcome of Graves’ disease patients after methimazole therapy. Journal of Biological Regulators and Homeostatic Agents 2001 14 193199.

    • Search Google Scholar
    • Export Citation
  • 67

    Choo YK, Yoo WS, Kim DW & Chung HK. Hypothyroidism during antithyroid drug treatment with methimazole is a favorable prognostic indicator in patients with Graves’ disease. Thyroid 2010 20 949954. (doi:10.1089/thy.2009.0126)

    • Search Google Scholar
    • Export Citation
  • 68

    Cappelli C, Gandossi E, Castellano M, Pizzocaro C, Agosti B, Delbarba A, Pirola I, Martino E de, Rosei EA. Prognostic value of thyrotropin receptor antibodies (TRAb) in Graves’ disease: a 120 months prospective study. Endocrine Journal 2007 54 713720. (doi:10.1507/endocrj.K06-069)

    • Search Google Scholar
    • Export Citation
  • 69

    Bo Youn C, Min Ho S, Ka Hee Y, Hong Kyo L, Chang-Soon K & Hun Ki M. Evaluation of serum basal thyrotrophin levels and thyrotrophin receptor antibody activities as prognostic markers for discontinuation of antithyroid drug treatment in patients with Graves’ disease. Clinical Endocrinology 1992 36 585590. (doi:10.1111/j.1365-2265.1992.tb02269.x)

    • Search Google Scholar
    • Export Citation
  • 70

    Allannic H, Fauchet R & Lorcy Y. A prospective study of the relationship between relapse of hyperthyroid Graves’ disease after antithyroid drugs and HLA haplotype. Journal of Clinical Endocrinology and Metabolism 1983 57 719722. (doi:10.1210/jcem-57-4-719)

    • Search Google Scholar
    • Export Citation
  • 71

    Winsa B, Adami HO, Bergstrom R, Gamstedt A, Dahlberg PA, Adamson U, Jansson R & Karlsson A. Stressful life events and Graves’ disease. Lancet 1991 338 14751479. (doi:10.1016/0140-6736(91)92298-G)

    • Search Google Scholar
    • Export Citation
  • 72

    Struja TM, Guebelin L, Kutz A, Fehlberg H, Mueller B & Schuetz P. Does immunosuppressive therapy improve outcomes in Graves’ disease? A systematic review and meta-analysis. Thyroid 2016 26 634640. (doi:10.1089/thy.2015.0647)

    • Search Google Scholar
    • Export Citation
  • 73

    Nygaard B. Hyperthyroidism (primary). BMJ Clinical Evidence 2010 2010 0605.

  • 74

    Vos XG, Endert E, Zwinderman AH, Tijssen JGP & Wiersinga WM. Predicting the risk of recurrence before the start of antithyroid drug therapy in patients with Graves’ hyperthyroidism. Journal of Clinical Endocrinology and Metabolism 2016 101 13811389. (doi:10.1210/jc.2015-3644)

    • Search Google Scholar
    • Export Citation

 

     European Society of Endocrinology

Sept 2018 onwards Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 410 410 129
PDF Downloads 323 323 112