Effect of Addition of Thyroxine in the Treatment of Graves’ Disease: A Systematic Review

Graves’ disease is the most common cause of hyperthyroidism. Antithyroid drugs, radioiodine ablation, and surgery are the main treatments. Research has demonstrated that adding thyroxine to antithyroid therapy can improve the remission rate, and many similar studies have been conducted subsequently. The purpose of this systematic review was to investigate whether adding thyroxine to various treatments for Graves’ disease has a clinical benefit in remission/relapse rate, stable thyroid function, occurrence of Graves’ ophthalmopathy, etc. A total of 27 studies were included, and the risk of research bias was moderate to high. We discuss the role of thyroxine both in pharmacological and non-pharmacological therapeutic regimens. Overall, the available evidence does not support the indiscriminate addition of thyroxine to various treatments for Graves’ disease, especially in combination with oral antithyroid drugs. Further clinical studies are required to explore the indications of thyroxine addition in the treatment of Graves’ disease.


INTRODUCTION
Hyperthyroidism is characterized by increased thyroid hormone synthesis and secretion. The most common cause of hyperthyroidism is Graves' disease, followed by the toxic multinodular goiter. The worldwide prevalence of hyperthyroidism varies by region. A meta-analysis of studies that evaluated the prevalence and/or incidence of thyroid dysfunction in Europe showed a prevalence of about 0.8%, while the National Health and Nutrition Examination Survey in the United States showed a prevalence of 1.3% (1,2). Global variations in the epidemiology of hyperthyroidism may be due to differences in diagnostic thresholds, assay sensitivity, population selection, and iodine nutritional status (3). In areas with adequate iodine intake, most thyroid dysfunctions are due to thyroid autoimmunity. Numerous studies have revealed the effects of changes in iodine status and the epidemiological effects of iodine supplementation on thyroid dysfunction (4,5). Studies in Denmark have shown that Graves' disease accounts for 70-80% of patients with hyperthyroidism in areas with adequate iodine intake, while in areas with iodine deficiency it accounts for about 50% of cases (6). The etiology of Graves' disease is considered multifaceted because of the reduced immune tolerance and the development of autoantibodies that stimulate thyroid follicular cells by binding to the thyroid-stimulating hormone (TSH) receptor. There is evidence of the genetic susceptibility of Graves' disease (7). Among non-genetic factors, iodine nutritional status, psychological stress, gender, smoking, low selenium, immunomodulators, etc., may contribute to the progress of the disease (8)(9)(10)(11)(12)(13). Hyperthyroidism treatments include antithyroid drugs (ATD), radioactive iodine ablation, and surgery. Outside the United States, ATD are preferred as primary treatment, while surgery and radioactive iodine are utilized in patients with persistent or recurrent hyperthyroidism (14,15). There are two ways to use ATD: titration, and block & replace. With titration, the dose of ATD is titrated over time to the lowest dose needed for maintaining an euthyroid state. In the block-and-replace regimen, a higher dose of ATD is used with concurrent replacement with levothyroxine. A meta-analysis including twelve studies in 2010 investigated the effects of B&R and titration block protocols. Although there were adverse reactions, the recurrence rate was the same in the two groups, 51% in the B&R group and 54% in the titration block group (odds ratio, 0.86; 95% confidence interval, 0.68 to 1.08). Withdrawal due to side effects (16% and 9%, respectively) was significantly greater in the B&R group (16). Although the block-and-replace regimen is controversial in many studies, it is still used in many regions.
The role of thyroxine in the treatment of Graves' disease needs to be evaluated on multiple aspects. The purpose of this study was to examine the addition of thyroxine for the treatment of Graves' disease and explore its benefits and deficiencies. This study was designed as a systematic review.

Types of Study
Randomized controlled trials (RCT), non-RCT, crossover studies, retrospective studies, and comparative studies were included in this analysis. Protocols and case reports were ruled out. Abstracts and non-English language publications were equally excluded.

Type of Participants
We included patients with a clinical diagnosis of Graves' disease. There were no restrictions on the age or gender of patients.

Type of Interventions and Comparisons
We included studies that combined thyroxine in the treatment of Graves' disease, where the primary endpoint was the remission/ relapse rate of Graves' disease with and without thyroxine. Other effects of this combination therapy in the treatment of Graves' disease were also considered for demonstration in this study, including stable thyroid function, occurrence of Graves' ophthalmopathy, quality of life, operative process, postoperative complications and so on. If patients with Graves' disease developed hypothyroidism and used thyroxine, the respective studies were excluded.,

Search Strategy
We searched the following publication databases: PubMed, Embase, Cochrane Library, and Web of Science. Studies published in English before December 31, 2019, were included. Selected keywords included Graves' disease, Basedow's disease, thyroxine, levothyroxine, L-thyroxine. Keywords were customized for each scientific database. The references of all retrieved articles were examined to ensure that all relevant articles were included for data synthesis. The retrieval strategy and process are illustrated in the Supplementary Material.

Selection of Studies
Two authors (J.L. and L.B.) independently reviewed all potential studies for inclusion against the eligibility criteria. They examined the title and abstract and, where necessary, the full text of the studies to assess if they were eligible for inclusion. In cases where they were not able to reach an agreement, a third author (W.Y.) made the final decision concerning eligibility.

Data Extraction
Two authors, M.W. and Y.X., independently assessed and selected studies eligible for this systematic review based on their title and abstract. Disagreement was resolved by consensus. Two authors, J.L. and L.B., read the full texts of the remaining articles and made a final selection. Data were extracted and presented in a table, including author, year, country, study design, sample, sex ratio (male-to-female), age (years), medication, intervention process, follow-up, assessments, results, side effects.

Assessment of Risk of Bias
The included studies were evaluated for the risk of bias using the Cochrane Collaboration's tool for randomized trials. The following aspects of the study were evaluated: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data addressed (intention-to-treat analysis), selective reporting, other bias. The non-RCT studies were assessed using the MINORS score, which serves as a methodological index for non-randomized studies. It has 12 domains for which non-comparative studies use the first 8 domains. Each domain is scored out of 2 with ideal scores being at least 16 for non-comparative studies and 24 for comparative studies.

RESULTS
The search resulted in 6,209 potential articles from PubMed (n = 1,777), Cochrane Library (n = 95), Embase (n = 3,672), and Web of Science (n = 665). In total, 6,139 articles were excluded based on the double-blind screening of the title and abstract, including 904 duplicates. The remaining 70 full text articles were assessed, and 27 articles were found eligible for this review (21 RCT and 6 non-RCT). Figure 1 shows the flowchart of the search and selection process. Overall, the risk of bias of randomized studies was moderate to high, as shown in Figure 2. The MINORS score of non-RCT studies were assessed in Figure 3. Research of Graves' disease related to pharmacotherapy is shown in Tables 1-3, while non-pharmacological research is shown in Table 4.

Effects of Thyroxine Addition on the Relapse or Remission Rates of Graves' Disease
Thyroxine in the "Block-and-Replace" Regimen The "block-and-replace" (B&R) regimen is well known to endocrinologists, but it is not routinely recommended in clinical practice. The B&R regimen completely blocks the thyroid hormone synthesis with high-dose ATD, while supplementing exogenous thyroxine to replace normal thyroid function. As per the 2016 American Thyroid Association guidelines, B&R therapy should be avoided or, if necessary, should be used after addressing compliance, in patients whose condition was inadequately controlled with one dose of methimazole (MMI) and who then become hypothyroid after a minimal dose increase (44). Nine of the clinical studies screened were on the B&R regimen, seven of which were RCT (17,(19)(20)(21)(22)(23)(24)35) and all used titration therapy as the control group.  (18,20,23). In the study by Perozim et al. (18), 28 patients first received 60 mg of MMI for 4-6 weeks and then received 40 mg of MMI and 75 mg of levothyroxine daily for 18-24 months. After the study was completed, patients were retrospectively divided into a remission group and a non-remission group and were followed up for 2 years. Owing to the low number of cases and the low level of methodological evidence, the study conclusions are debatable. The studies of Jorde et al. (20) and Grebe et al. (23) do not support the B&R regimen reducing the long-term relapse rate of patients with Graves' disease, but the results show that it can reduce the relapse rate in the short term. It is worth noting that compared with the 12-24 months medication cycle of other clinical studies, these two studies discontinued the drug after 6 months of medication and started a 3-year follow-up. It is not clear whether this short-term clinical benefit is related to the duration of medication. It is possible that a too short medication time may have a more obvious negative impact on the clinical effect of the titration therapy.   (19,21,22,24). There were differences in the timing of thyroxine addition in these four studies. Edmonds and Tellez (19)     little evidence that the B&R regimen could stabilize patients' thyroid function (25). Therefore, the included studies showed that thyroxine addition during the B&R regimen or after stopping the ATD did not show more benefits than the titration regimen in the long-term recurrence rate of Graves' disease.

Thyroxine Addition in the Titration Regimen
The titration regimen is widely used in clinical practice. Throughout the course of the treatment, the dose of ATD is adjusted according to the condition and thyroid hormone concentrations, which can generally be divided into initial, decrement, and maintenance phases. There were 11 RCTs (26, 28-37) among the included studies focused on adding thyroxine to titration therapy.
Among the 11 studies, only the study by Hashizume et al. showed a positive effect (26). The researchers believed that the addition of exogenous thyroxine can inhibit TSH secretion and reduce TSH receptor antibody (TRAb) levels and ultimately reduce the recurrence rate of Graves' disease. In this study, 109 patients received MMI 30 mg daily for six months until their thyroid function was normal. Then, the experimental group and the control group were administered thyroxine or placebo while receiving the maintenance dose of MMI. After 1 year of combination therapy, MMI was discontinued, but the patients continued with either thyroxine or placebo. The results showed that 1 patient (1.7%) in the thyroxine group and 17 patients (34.7%) in the placebo group had relapsed hyperthyroidism within 3 years after MMI discontinuation.
These results have not been confirmed in subsequent studies, and all the remaining nine studies drew contrary conclusions. Similar to the study by Hashizume et al. (26), in the study by Hoermann et al. (34), thyroxine was added after completion of the titration therapy (34). In this study, patients who successfully underwent ATD treatment for 12-15 months were stratified for risk factors and randomly assigned to receive TSH inhibitor doses of levothyroxine for 2 years or no treatment. The study by Tamai et al. not only compared the difference in the efficacy of titration therapy with or without the addition of thyroxine but also divided patients into two subgroups based on normal or suppressed serum TSH concentrations (28). During the two-year observation period, there was no significant difference in the recurrence rates between the three treatment options. Liu et al. and Yang et al. added thyroxine during the initial or maintenance phases of the titration therapy and discontinued ATD and thyroxine after the treatment (36,37).  (33) did not emphasize what dose of thyroxine was needed to inhibit the serum TSH concentration during treatment, which was quite different from Hashizume et al. (26). Therefore, the duration of the thyroxine treatment and whether it inhibits the serum TSH concentration had no effect on the remission or relapse rates of Graves' disease.

Thyroxine Addition in Special Populations
The treatment of Graves' disease in special groups, such as children and pregnant women, has its special features. In the study by Hashizume et al., 78 patients with Graves' disease received MMI for 1-3 years before pregnancy and thyroid function remained normal during the first two months of pregnancy (38). MMI was discontinued 5-6 months after the start of the pregnancy and the patients were divided into two groups. Group A was given thyroxine (100 mg/day), and group B was given no drugs from 5 months after the start of the pregnancy to 1 year after delivery. The results showed that thyroxine administration during pregnancy and after delivery can effectively reduce TRAb levels and prevent postpartum recurrence of hyperthyroidism. It is now generally accepted that ATD combined with thyroxine is not recommended during pregnancy, because the side effects of high-dose ATD are more serious in pregnant women (45). Because there are only few studies on this topic, it is unclear whether adding thyroxine during pregnancy can obtain the same clinical benefits as the above studies.
Parents generally consider non-pharmacological treatments for children with Graves' disease because of the poor compliance or severe side effects caused by ATD. Due to concerns about postintervention complications, non-pharmacological treatments are often not accepted by children's parents when the patient is very young (46). Vigone et al. retrospectively studied the effects of a medium dose of ATD with levothyroxine (L-T4) versus monotherapy with ATD in pediatric patients with unstable Graves' disease (39). The results showed that ATD with L-T4 did not increase the remission rate. However, this combination therapy may be useful when it is difficult to control the thyroid function using ATD therapy alone, and it may be a treatment option to defer certain treatments to the appropriate age.
Although such studies are few and insufficient as clinical evidence, they can also provide some basis for clinical treatment and further research.

Thyroxine Addition in the Non-Pharmacological Treatment of Graves' Disease
Four of the clinical studies screened were on the nonpharmacological treatment of Graves' disease, two of which were RCT (40,43). Non-pharmacological therapies for Graves' disease mainly include radioactive iodine 131 (RAI) and surgical treatment.
Taïeb et al. randomly divided 94 patients with Graves' disease into group A (50 µg L-T4 daily starting 15 days post-radioiodine and the dose of L-T4 was adapted to the thyroid function test results at 1, 3, and 6 months) or group B (follow-up every 4 weeks and treated with L-T4 when needed), and the follow-up period was 6 months (43). The results indicated that L-T4 administered early after RAI may have potential benefits to  patients' quality of life, but the optimal strategy for L-T4 dosage and timing remained to be determined. Tallstedt et al. retrospectively analyzed the clinical data of 492 patients who received RAI treatment, divided in 2 groups (41). One group of patients was administered thyroxine prophylactically 2 weeks after the RAI treatment, while the other group was administered thyroxine only when hypothyroidism occurred. After 18 months of follow-up, it was shown that early thyroxine administration after RAI treatment can reduce the occurrence of Graves' ophthalmopathy. The remaining two studies were related to the role of thyroxine in thyroid surgery. Kaur et al. randomly assigned 34 patients with Graves' disease to three treatment groups, first using ATD to normalize thyroid function, and then supplementing with thyroxine (40). The first group of patients stopped ATD and thyroxine treatment 10 days before partial thyroidectomy and received only Lugol's iodine treatment. In the second group, only ATD and thyroxine were used for treatment until surgery. In the third group, ATD and thyroxine continued until the day of surgery, and patients also received Lugol's iodine treatment for 10 days. The main role of iodine is to inhibit the release of thyroxine from the thyroid gland, which is often used in preparations for thyroidectomy. The results of the study showed that in patients who had used ATD and thyroxine to normalize the thyroid function, preoperative iodine therapy had no effect on the blood vessels of the gland, blood loss from surgery, or size of the thyroid follicles. Zhu et al. divided 471 patients with hyperthyroidism into an experimental group and a control group (42). The control group was given ATD and iodine preparations before surgery, while the experimental group was further divided into four subgroups and treated with "sequential thyroid defunctionalization followed by thyroxine supplementation." The four subgroups received different doses of ATD and thyroxine at different time periods. The results showed that "sequential thyroid defunctionalization followed by thyroxine supplementation" could effectively reduce the bleeding volume and the incidence of postoperative complications of partial thyroidectomy.
From these results, some clinical benefits could be seen, but these studies were small in size and scattered in purpose. If these conclusions are to be confirmed, high-quality clinical research is needed.

DISCUSSION
The relative advantages and disadvantages of various treatments for Graves' disease are still controversial. Each method is effective, but none is perfect. The choice depends on the treating physician, and increasingly on the wishes of the patient and local medical resources. Therefore, it is necessary to study the clinical efficacy of various treatments, including remission/relapse rate, quality of life and so on, which can provide the basis for the choice of treatment.
ATD have been part of the standard treatment for Graves' disease for many years and may lead to long-term remission. Searching through the results, we found two modes of combined use of oral ATD and thyroxine (1): the B&R regimen, that is, simultaneous use with high doses of ATD (2); adding thyroid in the titration regimen. It has long been adopted by many researchers that the recurrence of hyperthyroidism caused by Graves' disease or the interruption of the ATD treatment is associated with elevated TRAb levels in most patients. Earlier research by Hashizume et al. (26). suggested that exogenous thyroxine supplementation after ATD reduction and discontinuation could inhibit TSH secretion and reduce the thyroid-stimulating antibody (TSAb) production, reducing the recurrence of Graves' disease. The same findings were achieved in subsequent studies in pregnant women (26). However, the results of most subsequent studies in various countries showed that thyroxine addition to maintain normal TSH serum levels or inhibit TSH serum levels did not reduce TSAb levels, and the recurrence rate did not decrease. A prospective multicenter observational cohort study of 344 patients showed that the prediction of biological euthyroism during treatment with antithyroid drugs is higher during titration compared to B&R regime. Whether one regimen is clinically more advantageous than the other remains unclear (47). With regard to the huge differences in the results of these studies, existing explanations point to possible genetic and iodine intake differences (48,49). TRAb is due to an uncertain mechanism that changes the body's immune homeostasis (50)(51)(52). There is evidence that ATD has an immunosuppressive effect and may help suppress the production of TRAb, but this is still controversial (53,54). A significant reduction in TRAb can be observed in patients receiving ATD; however, TRAb may continue to increase in many patients after the long-term treatment achieves remission of hyperthyroidism (51). Evidence of a dose-dependent immunosuppressive effect of MMI on the intensity of autoimmune processes in the thyroid gland has not been found (55). It is now widely accepted that the relief of Graves' disease relates to the restoration of normal thyroid function and is not a special drug effect. This conclusion underlines the importance of maintaining a normal thyroid in patients with Graves' disease (56). Since thyroxine combined with ATD is more complicated than monotherapy, and from most research results, this combination therapy does not make thyroid function more stable. Moreover, the B&R regimen has produced conflicting results in terms of the possibility of disease remission and side effects (57). Most of the included studies did not adequately report side effects. In studies that mentioned side effects (19,20,23,25), the side effects of B&R regimen were more than those of titration regimen. From our research results, thyroxine combined with ATD cannot obtain a more stable thyroid function, higher remission rates and lower recurrence rates than ATD alone. However, in some patients treated with ATD who have poor thyroid function control, the addition of thyroxine can be tried. In the studies we reviewed, we also observed some positive effects, such as the benefits on bone metabolism in elderly patients (37), the benefits on goiter (29), etc. However, these observations need to be confirmed in further clinical trials. These clinical trials have shown some benefits in adding thyroxine to the surgical treatment of Graves' disease. Recurrent laryngeal nerve injury, parathyroid injury, and postoperative bleeding are serious complications of thyroid surgery and are closely related to the adequacy of preoperative preparation. Preoperative preparation for thyroid surgery, including ATD use to restore thyroid function, usually requires the use of a saturated potassium iodide solution to reduce the blood flow to the thyroid gland to limit the perioperative blood loss (58). However, there are some disadvantages. First, taking iodine for a long time or abruptly stopping it will cause a sudden increase in free thyroid hormone and cause thyroid crisis. Second, although the size of the thyroid gland and the blood supply can be reduced, the glandular tissue become hard and brittle, and there is significant surgical bleeding. In addition, iodine preparations are not appropriate for patients allergic to iodine. Using the "preoperative hypothyroidism and then add thyroxine" preoperative preparation method, exogenous thyroxine can effectively reduce thyroid congestion and swelling, decrease thyroid blood supply, soften thyroid tissue, and make the thyroid gland similar to the normal thyroid gland, thus allowing for a more convenient surgery and fewer surgical complications (42,59).
A study of 2,430 patients with Graves' disease followed for 6 to 10 years showed that only 35.7% of all patients reached a normal thyroid hormone status without levothyroxine replacement (60). Hypothyroidism is common after Graves' disease undergoes RAI and surgery. The results of the included studies showed that an early use of thyroxine after RAI could not significantly reduce the incidence of severe hypothyroidism but could reduce the incidence of Graves' ophthalmopathy and improve the patients' quality of life. Studies have shown that after RAI treatment of Graves' disease, the incidence of ophthalmopathy is worse than after oral ATD (61). Elevated TSH levels are positively associated with the development of Graves' ophthalmopathy, and thyroxine addition may enhance TSH inhibition after RAI (62,63).
In addition, there are many meaningful case reports showing the use of thyroxine in Graves' disease. ATD overuse in pregnant women with Graves' disease can cause hypothyroidism in fetuses, and intraamniotic injections of levothyroxine have been proven successful in fetal treatment (64).
In summary, the available evidence does not support indiscriminate thyroxine supplementation in the treatment of Graves' disease, especially when used in combination with oral antithyroid drugs. The focus of future research is to further clinical studies to explore the indications for adding thyroxine to the treatment of Graves' disease.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.