PDF
Citation
Print
INTRODUCTION
Graves’ disease (GD) is an autoimmune disorder that targets thyroid cells and represents the most prevalent form of hyperthyroidism. The annual incidence of GD is estimated at 20 to 30 cases per 100,000 individuals, with rates of 3.0% in females and 0.5% in males [1,2]. The primary goal of GD treatment is to normalize thyroid hormone levels. If unresolved, hyperthyroidism increases the risk of mortality from cardiovascular disease, thyroid storm, and other serious conditions [2]. Most patients with GD are treated with radioiodine therapy (RIT) or undergo total thyroidectomy (TT). However, TT carries the risk of serious complications, including recurrent laryngeal nerve palsy, hypoparathyroidism, wound infection, and hemorrhage [3]. Sundaresh et al. [4] reported that RIT is the most common treatment modality due to its superior safety and efficacy. Donovan et al. [5] conducted a cost-utility analysis comparing RIT, antithyroid drugs (ATDs), and TT as first-line treatments for GD, concluding that RIT is the most cost-effective option. A recent survey on GD management among clinicians in Asia and the Pacific region revealed a preference for RIT (19%) over surgery (2%) [6]. RIT offers the convenience of outpatient administration through self-administered capsules, eliminating the need for hospitalization. Presently, two main strategies are employed for RIT dosing: fixed doses or calculated doses [1,2]. The calculated-dose method provides a personalized approach, delivering a precise radiation dose to the thyroid and potentially achieving a euthyroid state. However, this approach is relatively expensive, requiring additional tests, potential travel and overnight stays, and more time away from work. Consequently, calculated doses may not be practical for patients in remote locations or those with limited financial resources. For these individuals, the fixed-dose method offers an acceptable alternative.
To date, no consensus has been reached on the optimal radioiodine (RAI) activity for treating GD using the fixed-dose method. The 2016 American Thyroid Association guidelines recommend a single RAI dose of 10 to 15 millicuries (mCi) [2]. Research on the use of fixed-dose RIT for GD over the past decade has shown varying success rates, ranging from 49.0% to 94.8%, with higher success rates associated with larger RAI doses (Supplemental Table S1) [7-13]. However, techniques for determining dosage vary among studies. Some have based RAI doses on thyroid size [9,13], while many have administered the same dosage to all patients [7,8,10-12]. Other factors in addition to thyroid size have been reported to influence treatment outcomes. Mariani et al. [14] proposed RAI doses of 5, 10, or 15 mCi based on disease severity, thyroid size, and radioiodine uptake (RAIU) values. However, their report did not provide guidance on how to integrate these factors to select the appropriate dose [14]. Consequently, applying the fixed-dose method based on the current literature is challenging.
To determine the optimal RAI activity for fixed-dose treatment in real-world practice, we developed two protocols. Protocol 1, which was based on thyroid size, was implemented at our facility from 2014 to 2017. Protocol 2, utilized from 2018 to 2021, incorporated a dose calculation formula that considers clinical factors associated with treatment failure. We hypothesized that the latter protocol would yield a higher success rate and a lower frequency of hypothyroidism compared to protocol 1. Our objective was to compare the success rates of RIT using these protocols to estimate the optimal RAI activity for treating patients with GD.
METHODS
Ethical considerations
The study was conducted in accordance with the tenets of the 2013 revision of the Declaration of Helsinki. Approval was obtained from the Institutional Review Board of Queen Savang Vadhana Memorial Hospital (No. 010/2565) and the Thai Clinical Trial Registry (TCTR20220527002). Due to the retrospective nature of the study, the need for informed consent was waived by the Institutional Review Board.
Study design
This retrospective quasi-experimental trial was conducted at Queen Savang Vadhana Memorial, Burapha University, and Queen Sirikit Hospitals.
Participants
We enrolled 824 Thai patients with GD, aged 18 years or older, who received RIT at the Radiology and Nuclear Medicine Unit of Burapha University Hospital between January 2014 and December 2021. Patients were diagnosed with GD based on the following criteria: suppressed thyroid-stimulating hormone (TSH) levels, elevated serum-free thyroxine (fT4) or triiodothyronine levels, and the presence of thyroid-associated ophthalmopathy and/or diffuse goiter [4]. All participants received fixed-dose RIT. From the initial cohort of 824 patients, we excluded 166 for various reasons: loss to follow-up or a follow-up period shorter than 6 months (57.8%); the need for high RAI activity to ablate the patient’s thyroid gland, as determined by the clinician (25.3%); a thyroid larger than 105 g (9.0%); or a history of RIT or thyroid surgery (7.8%). We excluded patients who required high doses of RAI to ablate the thyroid gland because these cases did not require the application of a formula to calculate RAI dosage. We also excluded patients with thyroids larger than 105 g because the formula would yield an RAI dose of 30 mCi or greater, which is not suitable for outpatient administration. Additionally, patients who had previously undergone RIT or thyroidectomy were excluded, as the determination of RAI activity would be influenced by prior treatments. The remaining patients were considered eligible for inclusion. Patients treated between 2014 and 2017 received RAI doses based on thyroid size (protocol 1), while those treated between 2018 and 2021 received RAI based on our modified dose calculation approach (protocol 2).
The sample size for this trial was calculated using the following equation.
Here, n represents the required sample size, while p1−p2 denotes the difference in the success rates of RIT between two groups: one treated with 10 to 15 mCi RAI (p1=0.75) [7] and the other treated with protocol 2 (estimated p2=0.85). p* indicates the average of both groups (p*=0.80). Using an alpha value of 5% (Zα/2=1.96) and a power of test of 80% (Zβ=0.84), we calculated that each group required a sample size of 250 cases. To account for potential loss to follow-up, we increased the sample size by an additional 30%, resulting in a requirement of 325 cases per group. Consequently, the total sample size for this study was 650 cases.
Procedure
In 2018, we developed an RIT data collection method following the introduction of the new fixed-dose protocol at our institution. The procedures for both fixed-dose protocols were meticulously planned and standardized. All female patients of childbearing age were required to undergo pregnancy testing within 24 hours before receiving RAI therapy and were counseled to use contraception for at least 6 months after treatment. Furthermore, radiation protection guidance was provided to patients and their families. Patients with inactive moderate ophthalmopathy were treated with RIT in conjunction with corticosteroid therapy [4]. Those taking ATDs were instructed to discontinue these medications for 3 days before and after RIT. Thyroid volume (TV) was assessed using two ultrasound devices: the Aplio 200 and the Aplio 300, both equipped with a 7 to 14 MHz linear transducer and manufactured by Toshiba (Tokyo, Japan). We measured each thyroid lobe in three dimensions and calculated the total TV using the standard ellipsoid volume formula, where TV=0.52×(width×depth×length) [15].
Between 2014 and 2017, we estimated the optimal RAI activity based on thyroid size (protocol 1) as follows: 10 mCi for thyroids ≤15.0 g (normal size), 15 mCi for thyroids between 15.1 and 45.0 g (enlarged to three times normal size), 20 mCi for thyroids between 45.1 and 75.0 g (enlarged to five times normal size), 25 mCi for thyroids between 75.1 and 105.0 g (enlarged to seven times normal size), and 30 mCi for thyroids ≥105.0 g. After 2017, we developed a new protocol (termed protocol 2) that utilized the following dose calculation formula:
RAI activity (µCi)=thyroid size (g)×200 µCi/g×(1/24-hour RAIU [%]) [4].
We estimated the mean 24-hour RAIU to be 70.0%, based on the analysis of data from 304 patients who received calculated-dose RIT between 2014 and 2017. To determine the appropriate RAI activity, we considered key clinical risk factors such as male sex and smoking [16,17], and the presence of first-degree family members with autoimmune thyroid diseases [17]. These factors were corroborated by the results of our pilot study, which included 136 patients with GD treated under a fixed-dose protocol from 2014 to 2016. The study identified male sex (relative risk [RR], 1.2; 95% confidence interval [CI], 0.8 to 2.0; P=0.371) and smoking (RR, 1.7; 95% CI, 0.9 to 3.1; P=0.984) as factors associated with RIT failure within 6 months. When calculating the dose, we increased the RAI activity by 10% for each clinical risk factor. Patients adhered to a low-iodine diet for 7 days before RIT. They then ingested a sodium iodide capsule on an empty stomach. To assess each patient’s thyroid functional status, we conducted thyroid function tests and measured thyroid size at 6 months and again at 12 to 13 months after RIT. The primary outcome was treatment success or failure. Treatment success was indicated by the presence of euthyroidism (normal TSH and fT4 levels) without ATDs or the development of hypothyroidism (high TSH and low fT4 levels) requiring levothyroxine therapy within 6 months following RIT. Treatment failure was indicated by continued hyperthyroidism (low TSH and high fT4 levels) and/or the continued necessity for ATDs post-RIT [4]. The secondary outcome was also treatment success or failure, defined by the same criteria but evaluated at 12 to 13 months after RIT. Patients who experienced hyperthyroidism after at least 6 months of remission were classified as having relapsed.
Data analysis
Baseline assessment and follow-up data were obtained from our hospital’s medical database. Clinical parameters were reported as mean±standard deviation (SD), median (interquartile range [IQR]), or median (range) for quantitative variables and as frequency (%) for qualitative variables. The Mann–Whitney U test was employed to compare continuous baseline variables between the two independent dosage method groups. The chi-square test was utilized to assess proportional differences in RIT success rates between the groups. Confounding factors, including sex (female/male), thyroid size category (≤45.0 g vs. >45.0 g, and ≤75.0 g vs. >75.0 g), smoking status (yes/no), and age, were evaluated using logistic regression to determine the ratio of success rates for protocols 1 and 2. RRs were also reported with 95% CIs. A P value of less than 0.05 was considered to indicate statistical significance.
RESULTS
The mean±SD patient age was 39.2±11.9 years, and most patients were female (64.8%). The median thyroid size was 26.8 g (IQR, 15.0 to 42.8). Table 1 presents the success rates of RIT for the two protocol groups. The group receiving protocol 2 displayed a slightly lower RIT success rate compared to those undergoing protocol 1 (63.6% vs. 67.2%), but the protocol 2 group also displayed significantly fewer cases of hypothyroidism (39.0% vs. 48.9%, P=0.0117). Based on the effect measures of protocol 2 and protocol 1, the risk of treatment failure did not differ significantly between groups (RR, 1.1089; 95% CI, 0.8937 to 1.3758; P=0.3477). Additionally, the median dose of RAI for protocol 2 was significantly lower than for protocol 1 (10.7 mCi vs. 15.0 mCi, P=0.0079). The prevalence of thyrotoxic periodic paralysis was higher in the protocol 2 group (13.6% vs. 7.8%, P=0.0217), as shown in Table 1. Moreover, no significant differences were noted in clinical factors or indications for RIT between the groups.
Table 2 presents the effects of age, sex, thyroid size, and smoking status on the success rates of the two protocols. A smaller thyroid gland (≤75.0 g) was associated with higher treatment success rates for both protocol 1 (RR, 2.230; 95% CI, 1.430 to 3.476; P=0.011) and protocol 2 (RR, 1.968; 95% CI, 1.407 to 2.753; P=0.003). Neither age nor smoking status significantly influenced the success rates. However, female sex was associated with a lower success rate for protocol 2 (RR, 0.815; 95% CI, 0.704 to 0.943; P=0.009).
After 12–13 months of RIT, 29.9% (197) of patients were lost to follow-up. Loss to follow-up was more common among recipients of protocol 2 compared to protocol 1 (33.6% vs. 24.6%), yielding a significantly lower RIT success rate for the protocol 2 group (64.1% vs. 74.2%). Additionally, the incidence of treatment failure was significantly greater for protocol 2 than for protocol 1 (34.0% vs. 23.8%, P<0.05). During the follow-up period, the RAI dose used in protocol 2 remained significantly lower than that for protocol 1 (10.9 mCi vs. 15.0 mCi, P=0.0263). Relapse rates after achieving at least 6 months of remission were exceedingly low in both groups, with no significant difference between them (1.9% vs. 2.0%, P=0.9695) (Table 3).
DISCUSSION
The European Association of Nuclear Medicine (EANM) recently published new guidelines on the use of RIT for benign thyroid disease. They proposed two strategies for determining the RIT dosage in the treatment of GD. The first strategy is functional dosing, in which the primary outcome is achieving euthyroidism. The second is ablative dosing, which targets the induction of hypothyroidism as rapidly as possible [18]. In our clinical practice, we have opted for ablative dosing for patients with GD who have extremely large thyroid glands (exceeding 105 g), who exhibit hyperthyroidism, or who have recently experienced a severe complication. We believe that patients who do not meet these criteria should have the opportunity to attain euthyroidism following the use of RIT for thyroid size reduction. Research indicates that most patients with GD (approximately 80%) treated with ablative dosing, who subsequently require lifelong management of hypothyroidism, are dissatisfied with this outcome [19]. Furthermore, over one-quarter of these patients encounter issues with thyroid hormone replacement therapy related to overtreatment or undertreatment (21% and 9%, respectively) [20].
To date, the fixed-dose method remains the simplest and most convenient approach for accurately estimating RAI activity. However, this method may lead to an increased incidence of hypothyroidism after RIT due to excessively high RAI activity; alternatively, persistent or recurrent hyperthyroidism can arise due to insufficient RAI activity. Over the past decade, numerous fixed-dose methods have been developed to determine the optimal RAI activity. However, these studies have produced a wide range of results and success rates, varying from 49.0% to 94.8% (Supplemental Table S1) [7-13]. In the present study, we developed and evaluated two fixed-dose methods for functional dosing, aiming to achieve either euthyroidism or short-term hypothyroidism in patients with GD. The technique in protocol 2 facilitated more precise and individualized calculations of RAI activity than the one in protocol 1, as it was adapted from a dose calculation formula based on individual thyroid size. Although the success rate of RIT using protocol 2 was marginally lower than that of protocol 1 (63.6% vs. 67.2%) (Table 1), no significant difference was observed in the risk of treatment failure between groups (RR, 1.1089; 95% CI, 0.8937 to 1.3758; P=0.3477). Furthermore, we were able to reduce the RAI dose (10.7 mCi in protocol 2 vs. 15.0 mCi in protocol 1; P=0.0079) while maintaining a similar treatment success rate to that of protocol 1 (Table 2). Thus, less radiation was administered. Moreover, the incidence of hypothyroidism during short-term follow-up was significantly lower in the protocol 2 group, with a difference of 9.9% relative to those receiving protocol 1 (39.0% vs. 48.9%, P=0.0117).
At the 12- to 13-month follow-up, the incidence of hypothyroidism had increased in both protocol groups (44.8% in the protocol 2 group vs. 58.4% in the protocol 1 group; P=0.0037), yet it remained significantly lower in the protocol 2 group. Additionally, the rise in incidence was less pronounced for protocol 2 compared to protocol 1 (5.8% vs. 9.5%) (Table 1). Notably, the incidence of disease relapse after at least 6 months of GD remission was exceedingly low for both groups (1.9% vs. 2.0%, P=0.9695) (Table 3). We compared this finding with results from a study by Kiatkittikul et al. [13], which also employed a fixed-dose protocol based on thyroid size. Relative to that study, our protocol 2 yielded a higher rate of hypothyroidism (39.0% vs. 33.1%) due to the effects of higher RAI activity across all thyroid sizes. Consequently, the success rate in our second protocol was greater than that reported by Kiatkittikul et al. [13] (63.6% vs. 50.0%).
Although protocol 2 reduced the incidence of post-RIT hypothyroidism, the success rate of 63.6% fell short of our anticipated target. Our findings indicate that a larger thyroid gland has a detrimental impact on the success of RIT. Specifically, thyroid sizes greater than 45.0 and 75.0 g were associated with lower success rates of 50.5% and 31.6%, respectively (Table 2). These findings align with previous research, which has shown that larger thyroid glands—ranging from 25.6 to 60.0 g—represent a key factor in RIT failure when fixed-dose protocols are employed [7,9,10,13]. The EANM guidelines acknowledge that higher RAI activity may be necessary for patients with GD who have a thyroid larger than 26.0 g [18]. This recommendation is supported by a clinical trial conducted by Vija Racaru et al. [21], which achieved an excellent RIT success rate of 91.0% and supported the use of the calculated-dose method in patients with GD whose thyroid exceeds 26.0 g.
Interestingly, and unlike the results observed for protocol 1, female patients exhibited a lower success rate for protocol 2 relative to male participants (RR, 0.815; 95% CI, 0.704 to 0.943; P=0.009) (Table 2). The prior identification of male sex as a risk factor prompted a 10% increase in the RAI dose for male patients, which improved the success rate of RIT (72.4% for protocol 2 and 60.0% for protocol 1). However, the success rate for female patients was lower in protocol 2 compared to protocol 1, at 59.0% versus 70.8%. One key factor influencing the success or failure of RIT was the rapid turnover rate of the thyroid gland. Rapid turnover rate is characterized by an early-tolate uptake ratio of ≥1, or a 4-hour RAIU that exceeds the 24- hour RAIU [22]. Factors associated with rapid turnover include a thyroid larger than 56 g (odds ratio [OR], 3.7; 95% CI, 3.032 to 4.559), age under 38 years (OR, 2.3; 95% CI, 1.906 to 2.856), and female sex (OR, 2.2; 95% CI, 1.757 to 2.791) [23]. Our institution employs both methods to determine RAI activity, considering clinician requirements and patient convenience. Thus, we collected RAIU values for patients with GD who received the calculated RIT dose during phase I (n=304, 2014–2017) and phase II (n=124, 2018–2021). We observed more patients exhibiting a rapid thyroid turnover rate in phase II compared with patients in phase I (11.3% vs. 5.6%) (Supplemental Table S2). Notably, the rapid thyroid turnover rate was exclusively observed in female patients (n=14), with a median age of 29.5 years (IQR, 26.5 to 38.8). Regarding RAI kinetics within the thyroid gland, patients with GD and a rapid turnover of RAI are expected to have a significantly shorter effective half-life for RAI clearance. This leads to lower absorption of the radiation dose delivered to the thyroid gland, necessitating an increase in RAI activity by approximately 1.5 to 2.0 times [24].
Based on the above results, we recommend using protocol 2 for patients with GD who have a thyroid 45.0 g or smaller. For thyroids exceeding 45.0 g but no larger than 75.0 g, we prefer the calculated-dose method, especially for female patients under 30 years of age. Alternatively, protocol 2 is recommended for patients who prefer a fixed dose approach and are willing to undergo a second RIT should the disease not be cured. For those with a thyroid larger than 75.0 g, we advise treatment with an ablative dose.
GD is a multifactorial disorder influenced by both genetic and environmental factors. While genetics are thought to account for 70% to 80% of GD cases, modifiable environmental factors, such as smoking, also contribute to the outcomes of RIT [17]. Plazinska et al. [25] found that current smokers with GD had a significantly lower RIT success rate than their non-smoking counterparts. Furthermore, a recent study indicated an elevated risk of GD in patients who smoke and have a first-degree relative with the disease (hazard ratio, 4.68), noting a statistically significant interaction between these factors (relative excess risk due to interaction, 0.94; 95% CI, 0.74 to 1.19). Consequently, it is advisable for patients with GD and their first-degree relatives to quit smoking [26]. Our institution is in the eastern economic corridors of Thailand, a region characterized by its industrial and coastal nature. Consequently, patients in this area tend to consume diets high in iodine and experience a relatively high incidence of stress-related conditions. Several studies have revealed correlations between these factors and the risk of GD [17]. However, no research to date has been focused on the impacts of stress and high iodine intake on the outcomes of RIT. Future studies are warranted to explore how these patient-specific variables affect the success of RIT.
Our study had several limitations. First, TSH receptor antibody titer is known to correlate with RIT failure [12]; however, we were unable to analyze this variable due to the prohibitive cost of testing all patients with GD. Second, 197 of our participants (29.9%) were lost to follow-up at 12 to 13 months, with an unequal rate of loss between the two groups (24.6% for protocol 1 vs. 33.6% for protocol 2). This discrepancy may have introduced attrition bias, potentially affecting the precision of the long-term success rates reported in this study. Third, TV measurements were conducted by three radiologists experienced in ultrasound, who concurred on the ultrasound technique used. However, we did not assess interobserver variation, as the data were collected retrospectively and one of the radiologists transferred to another hospital in 2020. Finally, our study was a quasi-experimental retrospective trial, which may have been influenced by confounding factors due to the absence of randomization.
In conclusion, the success rate of the modified dose calculation protocol was not inferior to the thyroid size-specific RAI dose protocol. This protocol yields a favorable success rate, especially in patients with GD who have a thyroid no larger than 45.0 g.
XML Download