Disease Presentation and Remission Rate in Graves Disease Treated With Antithyroid Drugs

Is Gender Really a Factor?

Talia Diker-Cohen, MD, PhD; Hadar Duskin-Bitan, MD; Ilan Shimon, MD; Dania Hirsch, MD; Amit Akirov, MD; Gloria Tsvetov, MD; Eyal Robenshtok, MD


Endocr Pract. 2019;25(1):43-50. 

In This Article


The optimal treatment algorithm for Graves disease is controversial, with three optional treatments, including ATDs, radioiodine, or surgery. While the most common treatment of Graves disease in the United States and Europe today is with ATDs,[29–31] the choice of treatment for each patients is tailored according to severity of disease, presence of GO, and the probability of remission with ATDs. Our study demonstrates that men and women with Graves disease have similar disease presentation and response to ATDs as first-line treatment. ATD treatment not only resulted in comparable remission rates and recurrences in both genders, but men experienced fewer adverse effects and discontinuation of treatment. This is in contrast to the common notion of men having low success rates with ATD therapy.[8]

The 2016 ATA guidelines on diagnosis and management of hyperthyroidism state that there is a lower likelihood of remission after ATD therapy in men,[11] with four studies as reference.[13–16] However, a PubMed search of this topic results in a list of at least 18 papers,[13–30] the majority showing that male gender is not a negative predictor for remission in ATD-treated patients, with one notable exception [11] (Table 4). Most of these studies were designed to assess predictors for remission, considering gender as one among many other variables. Only two other studies thus far were specifically designed to explore the role of gender in the natural history and response to treatment in Graves disease;[13,31] the first, by Allahabadia et al,[13] is the most-cited report, which lent credence for consideration of male gender as a negative predictor for response to ATDs. In that study, 92 males of 536 patients with Graves hyperthyroidism had both more-severe biochemical disease at presentation compared to females (statistically significant higher free T4 and free T3 levels), as well as lower remission rates after a course of ATDs (19.6% vs. 40%; P<.01) or one dose of radioiodine (47% vs. 74%; P<.0001). The authors concluded that as males have low remission rates to ATD, they should be offered definitive treatment with radioiodine or surgery soon after presentation. However, as men had more-severe disease compared to women in this study, the worse outcome of males could be attributed to the baseline disease severity rather than the gender difference in ATDs response. In contrast to this study, patients in our cohort had comparable baseline characteristics, and therefore, there is less concern for bias. Also, the ATA definition of remission is euthyroidism at 12 months after therapy,[11] whereas, as shown in Table 4, many studies have used other definitions. In the study by Allahabadia et al,[13] remission was defined as euthyroidism 6 months after ATD discontinuation, and the mean follow-up of patients was only 9 months after the end of therapy. As the definition of remission in that study was much shorter than recommended in the ATA guidelines, some of the cases defined as remissions would probably be classified as treatment failures after longer follow-up. In this context, another major strength of our work is the definition of remission as euthyroidism 12 months after the end of drug therapy. The second study specifically designed to explore the role of gender in Graves disease, by Magri et al,[31] retrospectively evaluated the response to ATDs in 294 patients (66 men). The response of men to ATDs was less favorable compared to women due to a higher rate of relapse. It should be noted that males had larger goiter size at presentation, a parameter that was later found to be associated with relapse in that cohort. Also, looking carefully at the data presented in the paper, a similar rate of males and females achieved biochemical normalization and were considered in remission after 18 months of ATD treatment (75.7% and 79.4%, respectively); most of the recurrences occurred during the first year after discontinuation of ATD (defined in the paper as early relapse, 48% males, 30.9% females; P = .024). As discussed above, according to the ATA hyperthyroidism guidelines, relapse is considered only after 1 year of euthyroidism, and again these cases, in which most continued to definitive therapy, would probably be classified as treatment failures after longer follow-up. There was no statistically significant difference between men and women in the rate of late relapse by the definition in Magri's work (between 1 and 5 years after ATD discontinuation), which is consistent with the definition and the results in our cohort.

The choice of an appropriate mode of therapy in Graves disease should take into account the rates of remission and recurrence. Several studies looked for predictors for treatment success either before the start of treatment or during therapy; the most-established adverse predictors of long-term remission after ATDs include young age, smoking, severe biochemical disease (especially high free T4) at disease presentation, large goiter, or high TSH-receptor antibodies.[4,11,13–15,28,29,32] Others have suggested multifactorial models for predicting remission, with consideration of clinical and laboratory parameters as well as genetic polymorphisms.[4] Yet, most of these studies used various definitions of remission and recurrence and should be interpreted considering this limitation. In our cohort, neither the clinical parameters (gender, menopausal status, autoimmune comorbidities, smoking, or presence of ophthalmopathy) nor the laboratory values (thyroid function tests, antithyroid antibodies) at presentation were found to be predictors for remission. It should be noted that TSH receptor antibodies were measured only at diagnosis and not before a decision to discontinue ATDs as the 2016 ATA guidelines recommend, and in roughly 30% of the cohort. This is because in our country, TSH receptor antibody tests are not reimbursed, and patients need to pay for the test. This might explain why TSH receptor antibodies did not predict remission after ATD therapy. Also of note, there is a growing body of evidence suggesting that estrogen may play a role in the pathogenesis of autoimmunity [12,33,34]. This has led us to question whether there may be differences between pre- and postmenopausal women in regards to the response to ATD treatment, yet menopause did not predict higher remission rates. Nevertheless, we think that the role of sex hormones should be further evaluated in this context.

A statistically significant association was detected between remission and the overall ATD treatment duration, as patients who achieved remission were treated for shorter periods of time compared to patients without remission; yet, it is somewhat intuitive to assume that treatment was ended earlier in those who achieved earlier normalization of thyroid function. Nonetheless, treatment duration per se in our study was longer than that recommended in the ATA guidelines. As this was a retrospective study, treatment duration was not predetermined, and the results rather reflect the common practice in our clinic; the similarities in treatment duration in both sexes further attest to this interpretation of local practice of the treating endocrinologists and not to confounding reasons or biases that led to longer-than-expected duration of therapy. The lack of widely available thyrotropin receptor antibody testing as recommended per the ATA guidelines may be one of the reasons for prolonged ATD therapy. It should nevertheless be pointed out that the recommendation for 12 to 18 months of treatment duration is based on one meta-analysis of old data [35] (four trials included, the latest of which was published in 1999) with enrollment of less than 80 patients in each study arm, and mostly comparison of ≥12 months of therapy to <6 months of therapy (not 12 to 18 months vs. longer duration). Moreover, some papers have been published since that suggest benefit for longer duration of therapy,[27] especially in children.[36] It would be interesting to see in current practice whether length of treatment per the ATA guidelines is indeed implemented in other centers and if new data still attest to short duration of treatment.

Also, more adverse events during ATD therapy predicted lower rates of remission. This could be explained by earlier drug withdrawal upon the occurrence of an adverse event, resulting in higher rates of treatment failure.

A 2011 international survey on the management of Graves disease showed that 85.7% of endocrinologists in Europe versus only 40.5% of the endocrinologists in North America would select ATDs as the primary treatment modality for an index case of uncomplicated Graves disease.[37] In our study, ATDs were used as first-line treatment in almost 100% of patients, resulting in a cohort with less selection bias. First-line treatment with ATDs was not an inclusion criterion but rather reflects the common practice in our institute.

Our study has several limitations. First, the retrospective design has inherent bias with the potential for missing data and reliance on the common practice of treating physicians. Second, the length of treatment with ATDs was dependent on individual physician practices and patient preferences, potentially extending treatment beyond time of remission. This could theoretically result in overestimation of no-remission cases due to our definition of ATD treatment duration longer than 18 months at time of data cut-point as failure to remit. Third, noncompliance was considered per the physician's report in the medical files that were reviewed and not by number of prescriptions or other systematic method. Moreover, though noncompliant cases to ATD therapy were excluded, one cannot fully detect partial nonadherence to treatment. Additionally, definition of goiter was based on clinical experience and physical examination rather than imaging studies, and we did not have accurate estimation of the thyroid volume. In the same line, GO was defined if clinically assessed or based on coding; however, data were not sufficient to accurately assess severity and activity in most cases. Lastly, thyrotropin receptor antibody results are expressed in a qualitative manner and not as precise levels.