Authors: Harold Henrison C. Chiu, Edrome F. Hernandez, Franz Michael M. Magnaye, Jereel Aron R. Sahagun, Jim Paulo D. Sarsagat, Jun-Sing Wang
Categories: Review, Hyperthyroidism, Radioactive iodine, Thyroid storm
The occurrence of post-radioactive iodine thyroid storm among patients with hyperthyroidism is relatively rare and only a few cases have been reported. We conducted a literature review of cases reported from 1951 to 2023 and determined the risk factors and clinical characteristics of patients who developed thyroid storm. A total of 19 cases were documented and reviewed. The mean age was 51.2 ± 20.1 years (range 7.5 to 75). Approximately two-thirds were females. Major etiologies were diffuse toxic goiter (57.9%) and nodular disease (36.8%). Mean dose was 11.3 ± 7.7 mCi (range 3.3 to 35), with 52.6% receiving less than 10 mCi. Mean interval time from administration to development of thyroid storm was 6.6 ± 5.5 days (range 0.5 to 20). The most common preexisting conditions were weight loss, heart failure, atrial fibrillation, hypertension and coronary artery disease. Thyroxine levels were not routinely measured prior to and during storm. Among those with available data, only 26.3% had hormone levels prior to and during storm. Thyroxine levels during storm (range 19.8 to 65 µg/dL) were higher than levels prior to storm (range 9.6 to 45 µg/dL). Pretreatment regimens varied consisting of no intervention, beta blockers, steroids, reserpine, phenobarbital and anti-thyroid drugs. Treatment regimens are more uniform and consistent with American Thyroid Association recommendations. The mortality rate remains high at ~ 26.3%. Statistical analyses did not show any significant differences. Even though the frequency of this condition is quite rare, it is an important and potentially prognostic condition underscoring the value of this review. The inclusion of this severe adverse effect should be part of patient discussion with emphasis on the need to seek early consultation when severe symptoms appear.
Keywords: Hyperthyroidism, Radioactive iodine, Thyroid storm
Thyroid storm is a preventable, severe and life-threatening progression of thyrotoxicosis, characterized by thermoregulatory, central nervous, gastrointestinal, hepatic and cardiovascular system dysfunction. The mortality rate is high; reaching from 20 up to 30% [1]. Radioactive iodine (RAI) therapy and withdrawal of antithyroid drugs (ATD) are two well-described precipitants of thyroid storm [2]. Although the pathophysiology of developing thyroid storm following RAI therapy has not yet been fully elucidated. It is postulated that the destruction of follicular cells within the thyroid gland results in the rapid release of stored thyroid hormone into the circulation, which can serve as a precipitating event for thyroid storm in a very small percentage of patients [2].
It has proven difficult to clearly define the risk factors for development of thyroid storm after radioiodine administration because of small number of reported cases [2]. Most reported cases of thyroid storm in literature are seen amongst the pediatric population which usually occurs 2 to 8 days following iodine-131 (I-131) treatment. In these cases, thyroid storm was attributed to the RAI therapy as no other cause was identified [3]. Given the benefits and perceived low risks of RAI when compared to surgery or long-term anti-thyroid medication, the trend towards RAI therapy is likely to continue.
Nevertheless, just with any therapeutic intervention, RAI is not without significant risks [4]. As seen from a systematic review conducted back in 1983 by McDermott et al. [5], among 2975 patients treated with RAI, the estimated frequency of thyroid storm is only at (10/2975) 0.3% while that of severe exacerbations is at (26/2975) 0.9%. Although thyroid storm after thyroid ablation is rare, the significant morbidity and potential mortality of patients warrant further studies to determine if patients with comorbidities, uncontrolled hyperthyroidism or those with markedly elevated thyroid hormone concentrations upon diagnosis should receive prolonged pretreatment with anti-thyroid drugs. While such an approach may reduce the efficacy of I-131, it can also reduce and possibly eliminate the risk of post-radioactive iodine thyroid storm, but its use remains controversial [4].
However, there is still paucity of data when it comes to the clinical profile, risk factors and effects of pretreatment with anti-thyroid medications among adult populations who develop thyroid storm resulting from RAI therapy in uncontrolled hyperthyroidism. Hence, this review aims to explore the clinical profile of patients who experienced and developed thyroid storm following I-131 for hyperthyroidism, identify the possible predisposing risk factors, compare treatment regimens and evaluate mortality and prognosis.
We performed a comprehensive literature search strategy from inception up to May 31, 2023 in the following PUBMED/MEDLINE, Ovid MEDLINE, Google Scholar, Embase, and Web of Science. The search strategy was limited the following types of case reports, case series, case control studies, retrospective and prospective cohort studies. There was no language restriction on the searches performed. To identify all the relevant studies, the following descriptors were used to build the search thyroid storm, radioactive iodine therapy, hyperthyroidism, and risk factors among others; terms were combined with the Boolean operators AND and OR. We supplemented our electronic search with manual searches and by cross-referencing included papers, relevant sections of clinical practice guidelines as well as relevant systematic and narrative reviews.
All investigators independently screened citations from the electronic search, reviewed full-text papers for inclusion, critically appraised the quality of the included studies, and abstracted the data and results. All relevant clinical characteristics and risk factors associated with the development of thyroid storm following radioactive iodine ablation in patients with uncontrolled hyperthyroidism were tabulated and compared.
The risk factors and predisposing conditions that can predict clinical presentation and outcomes are based on case reports and case series [2, 3, 5–18]. We documented 19 published cases from our literature search spanning a 70-year period from 1951 up to 2023.
The mean age of these patients is 51.2 ± 20.1 years with an age range of 7.5 to 75 years. The percentage of patients above 40 years old is 84.2% (16/19). Approximately two-thirds, 63.2% (12/19) were females. The major etiologies and indications for treatment were diffuse toxic goiter at 57.9% (11/19) followed by nodular disease at 36.8% (7/19). The average age among those with diffuse goiter is 40.9 ± 20.1 while the average age for nodular goiters is 66.0 ± 8.4 years. The mean RAI dose is 11.3 ± 7.7 mCi and this ranged from 3.3 to 35 mCi, with 52.6% receiving less than 10 mCi RAI dose. The mean interval time from RAI to development of thyroid storm is 6.6 ± 5.5 days with an interval range of 0.5 to 20 days. Among these patients, the most common preexisting conditions occurring prior to RAI therapy are weight loss, heart failure, atrial fibrillation, hypertension and coronary artery disease. There were no detailed descriptions on the presence of eye symptoms, grading of Graves’ ophthalmopathy, and measurement of clinical activity score.
Thyroxine levels were not routinely measured prior to and during storm. Among those with available data, only 26.3% had hormone levels prior to and during storm. Thyroxine levels during storm (range 19.8 to 65 µg/dL) were higher than levels prior to storm (range 9.6 to 45 µg/dL). For the rest of patients already with known hyperthyroidism but without available thyroxine levels, the diagnosis of thyroid storm was made primarily using clinical criteria (Burch-Wartofsky criteria) and validated scoring systems [19]. No data was available for thyroid receptor antibody level.
On one hand, pretreatment regimens varied ranging from no intervention 36.8% (7/19), beta blockers 15.8% (3/19), steroids 5.3% (1/19), reserpine 5.3% (1/19), phenobarbital 10.5% (2/19) and ATDs 42.1% (8/19). On the other hand, treatment regimens for thyroid storm across cases are more consistent and mostly based on published recommendations made by the American Thyroid Association utilizing a combination of propylthiouracil, steroids, beta blockers and potassium iodide [19]. Based on the review of the 19 documented cases, the mortality rate is roughly 26.3% (5/19). All mortality cases were reported before 1975. A summary is shown below (Table 1).
Further statistical analyses did not show any significant differences in outcomes in terms of age, sex, size of thyroid gland, radiation dose, presence of comorbidities and pretreatment (Table 2).
In this review, the mean age, the percentage of females, cases with diffuse toxic goiters, time interval and RAI dose are close in terms of value to that seen in the 1983 review of McDermott et al. [5] forty years ago. These findings are supported by available literature showing that hyperthyroidism is more common in women and diffuse toxic goiter is the most common etiology of hyperthyroidism [20]. The current recommendations for RAI dosing consist of a fixed dose regimen ranging from 10 to 15 mCi [19]. In our review, 52.6% of patients have received less than 10 mCi RAI dose, which is lower than what is currently recommended by the American Thyroid Association for the treatment of Graves’ disease, toxic adenoma, and toxic multinodular goiter, which are 10 to 15 mCi, 10 to 20 mCi, and 150 to 200 uCi/g, respectively [19]. Thus, the dose of RAI is unlikely to be correlated with increased risk of thyroid storm. However, while the RAI dose given to each patient is known, no data is currently available with regards to the measured absorbed dose to the thyroid gland. We surmise that if measured and made available, this parameter could be of interest and add more information as a possible predictor of the development of thyroid storm.
While RAI is the not first-line treatment of hyperthyroidism in the pediatric population, in this review, there were two reported cases of post-RAI thyroid storm occurring in pediatric patients [3, 17]. The first case published in 1978 was a 10-year-old female with Down syndrome and hyperthyroidism. She was initially started on propylthiouracil then shifted to methimazole at 30 mg daily where she developed generalized maculopapular rashes prompting definitive treatment with RAI [17]. The second case published in 2001 was a 7.5-year-old boy with difficult to control hyperthyroidism lasting for 3 years and necessitating a block and replace regimen with a methimazole dose of 50 mg daily in order to maintain a biochemical euthyroid state. Definitive treatment using RAI was subsequently offered after a thorough discussion with the patient’s family [3].
In terms of baseline hormone levels and severity of hyperthyroidism, we observed that only roughly a quarter of patients in this review had T4 or FT4 levels prior to and during presentation of thyroid storm. For the rest of patients with already known hyperthyroidism but without available thyroxine levels, the diagnosis of thyroid storm was made primarily using clinical criteria (Burch-Wartofsky criteria) and validated scoring systems [19]. In addition, the absence of thyroxine levels prior and during the storm will not change the management of thyroid storm. Among those with available information, hormone levels were noted to be higher during storm as compared to values obtained prior to RAI. However, on further statistical analyses, there were no significant differences between the thyroid hormone levels prior and during the storm. This is consistent and supported by the mechanism of action following RAI. Iodine is the precursor of thyroid hormone and its radioactive form is taken up by sodium iodide symporter of the thyroid follicular cell similar to how iodine from the diet is taken up [21]. The emitted beta particle from I-131 then destroys the follicular cells, gradually inducing thyroiditis with subsequent rapid release of stored preformed thyroid hormone into the circulation causing a rise in hormone levels [2]. The increased thyroid hormone release from degenerating follicles can then serve as a precipitating event for a thyroid storm. Studies have also shown that patients with thyroid storm and uncomplicated thyrotoxicosis had comparable total thyroxine (T4) and triiodothyronine (T3) levels, but that free T4 and dialyzable T4 levels were significantly higher in the patients with thyroid storm [3]. With the exception of the patient with Graves’ disease who developed thyroid storm and complete heart block following RAI treatment, majority of the cases presented with the classic signs and symptoms of thyroid storm based on the Burch and Wartofsky thermoregulatory, central nervous, gastrointestinal, hepatic and cardiovascular system dysfunction. This is similar to other etiologies as outlined in the 2016 American Thyroid Association (ATA) Guidelines for the Diagnosis and Management of Hyperthyroidism/Thyrotoxicosis [19]. In addition, it is also interesting to note that the measurement of hormone levels before and during thyroid storm was only consistently performed in cases reported after 1975 which could be attributed to improved assay performance, turnaround time and accessibility of the tests. Routine measurement of thyroid receptor antibody levels was not performed in all reported cases. Hence, establishing an association between severity of hyperthyroidism based on pre-RAI thyroid function tests and the development of storm is difficult due to the lack of available data in literature.
Pre-treatment with thionamides accounted for less than half of the total reported cases, only 8 out of 19 (42.1%) of patients. It is important to note that while most cases of thyroid storm may occur after a precipitating event such as infection, pancreatitis, myocardial infarction, discontinuation of maintenance ATDs remains to be a common cause [2]. The relatively lower proportion of patients pre-treated with thionamides [2, 5] as well as the discontinuation of maintenance thionamides [10, 18] in these patients may be a possible risk factor for developing thyroid storm post-RAI. The current available evidence on pretreatment is based on the clinical practice recommendations of the American Thyroid Association (ATA) back in 2016 [19]. RAI treatment is known to cause a transient exacerbation of hyperthyroidism through the release of preformed hormones following thyrocyte destruction. Medical treatment and control of any comorbid conditions should be the priority goal before proceeding to RAI therapy. Patients with severe hyperthyroidism, the elderly, and individuals with cardiovascular complications such as atrial fibrillation, heart failure, or pulmonary hypertension and those with renal failure, infection, trauma, poorly controlled diabetes mellitus, and cerebrovascular or pulmonary disease are at a higher risk to develop exacerbation of hyperthyroidism [19]. Based on low-quality evidence (weak recommendation), beta-adrenergic blockade may be considered even if asymptomatic, among the elderly and those with comorbidities above, who may be at an increased risk for complications due to worsening of hyperthyroidism. In addition, based on moderate-quality evidence (weak recommendation), pretreatment with MMI prior to RAI therapy should be considered in patients who are at increased risk for complications due to worsening of hyperthyroidism; MMI should be stopped at least 2 to 3 days prior to RAI. Furthermore, resuming MMI 3 to 7 days after RAI administration should be considered based on low-quality evidence (weak recommendation). Several studies have looked into the subgroups that would benefit from pretreatment prior to RAI. A study by Burch et al. [22] in 2001 compared the occurrence of worsening thyrotoxicosis among those pre-treated with ATDs (30 mg of methimazole per day for at least two months prior to RAI administration adjusted to maintain FT4 within the normal range) versus those without ATDs. Once normal FT4 was achieved, methimazole was discontinued for six days prior to RAI ablation. Those in the non-pretreated group received RAI ablation within a week of initial evaluation. Five patients (11.9%), including three in the pretreatment arm and two in the no pretreatment arm experienced a late exacerbation of thyrotoxicosis after RAI therapy. After ablative therapy, the pretreated patients experienced a 52.4% increase (95%CI 26.4%, 78.5%) in FT4 and a 61.8% increase (95%CI23.5%, 100.0%) in FT3, with the upper limit of normal for FT4 exceeded in 23.8% of patients before receiving RAI and in 33.3% of patients one day after RAI. Conversely, the majority (90.5%) of non-pretreated patients registered a substantial decline in thyroid hormone levels after RAI treatment. Mean FT4 levels decreased from 85.8 ± 60.4 to 58.0 ± 76.5 pmol/L, representing a 32.4% decrease, whereas mean FT3 levels decreased from 16.1 ± 8.0 to 10.8 ± 11.1 pmol/L, representing a 32.9% decrease. Three patients developed worsening thyrotoxicosis in the pretreatment group whereas two developed in the non-pretreatment group, illustrating that pretreatment may not be necessary for all patients. In contrast, in a retrospective cohort study conducted by Vuayakumar et al. [23]. in 2006 which looked at pre-treatment with beta blockers involving 122 patients treated with RAI 10 to 20 mCi, no one developed thyroid storm despite the 25.0% of patients exhibiting high risk factors for thyrotoxic crisis (RAI uptake > 65.0%, very marked signs and symptoms of hyperthyroidism, and markedly elevated FT4 and/or FT3) [23]. They concluded that I-131 can be safely administered to patients with uncontrolled hyperthyroidism without fear of development of thyroid storm, provided that beta blockers are used to control the signs and symptoms [23]. Furthermore, Bonnema et al. [24]. conducted a clinical trial in 2001 to determine whether or not the resumption of methimazole seven days after I-131 influences the final outcome of ablative treatment. Three weeks following RAI therapy, the FT4 index slightly decreased by 5.7% (95%CI -15.5, 5.4%) in the ATD group, versus an increase of 35.9% (95%CI, 18.8, 55.5%, p = 0.001) in the no intervention group [24]. Thus, resuming methimazole after I-131 may help alleviate the early and transient thyrotoxic phase. Since ATDs are known to interfere with RAI uptake in the thyroid, they are routinely discontinued prior to administration of RAI [5]. Beta blockers are then used as an alternative in controlling thyrotoxic symptoms and as deterrent to the development of thyroid storm until RAI takes full effect [25]. Based on the review of several pre-treatment and post-treatment recommendations and strategies above [22–24], pre-treatment with beta blockers and ATDs, as well as post-treatment with ATDs are suitable options for prevention among high-risk groups of patients namely, severe thyrotoxicosis, the elderly, and those with comorbidities.
Treatment of post-RAI storm remains similar as with the other etiologies of thyroid storm [19, 26, 27]. Management is directed toward its four metabolic manifestations. The first of these is to interrupt synthesis and release of thyroid hormone through the effective use of high doses of iodine and ATDs. The second is blocking the continuing action of the thyroid hormone on the peripheral tissues and the peripheral conversion of T4 to T3. Propylthiouracil, intravenous corticosteroids, and beta blockers are excellent drugs to address this problem. The third component in management is the failure of the peripheral tissues to meet the demands of the severe hypermetabolic state. Specific therapy includes correction of dehydration, hyponatremia, reduction of fever by cooling blankets, and other replacements as needed. Lastly, in order to reverse and prevent thyroid storm post-RAI ablation, it is also important to address other possible factors such as the presence of intercurrent illness [2]. Known precipitating factors which can contribute to thyroid storm such as infection, myocardial infarction and heart failure, should therefore be adequately treated prior to RAI ablation for hyperthyroidism [2]. It is worth noting that those treated with PTU-based regimens for thyroid storm had a relatively high survival rate, 6 out of 7 (85.7%) patients.
In our review, 5 out of the 19 patients (26.3%) died. This rate remains high and is within the reported range of thyroid storm mortality ranging from 20 to 30% [1]. It is worth noting that all mortalities occurred before 1975 and this improved survival in modern times could be attributed to the advancement of medical science, drug therapeutics and patient care.
Our study has several limitations. First, because thyroid storm is infrequent in clinical practice, this makes it difficult to study a large number of cases, and great caution must be exercised before drawing conclusions from case reports. The results described in this paper are not general and largely biased towards reported cases. Second, the sample population is limited, highly heterogeneous and without a common denominator despite the long-time interval from the very first reported case. This precludes the performance of detailed quantitative analysis and the outcomes cannot be described in terms of odds ratios or relative risk. Third, it is difficult to establish an association between the severity of hyperthyroidism based on baseline thyroid hormone levels and the development of thyroid storm due to lack of available data in most reported cases of post-RAI storm. Lastly, it remains difficult to conduct large randomized studies because of ethical considerations and radiation exposure. Since there is not much change in terms new available data over the past 4 decades, clinical decision making in this area would remain to be from anecdotal evidence, evidence synthesis and clinical practice guidelines.
In summary, based on the limited available evidence, post-RAI storm is rare yet possible. Hence, inclusion of this severe adverse effect as part of patient education should be included during discussion with patients on the post-RAI side effects with emphasis on the need to seek early consultation when severe symptoms appear. It is prudent to identify and closely monitor patients because of the high mortality rate. Although it is difficult to draw conclusions from limited number of cases, several observations and findings deserve emphasis. Although not statistically significant, post-RAI storm has been shown to be more common in female patients, above 40 years old, and those with preexisting conditions such as weight loss, heart failure, atrial fibrillation, coronary artery disease and hypertension. However, there is no clear trend when it comes to the dose of RAI and the duration of symptom onset after RAI. Compared to baseline, hormone levels were noted to be higher during storm but the degree of association between hyperthyroidism severity using baseline hormone levels and the development of thyroid storm is difficult to establish due to lack of available data. While pretreatment prior to RAI remains an area of research, clinical practice guidelines recommend beta blockers and ATDs as well as post-RAI administration of ATDs may benefit certain at-risk individuals such as severe thyrotoxicosis, the elderly, and those with untreated and uncontrolled comorbidities. The treatment strategy of post-RAI storm is the same as that of thyroid storm and carries a high survival rate. Even though the frequency of this condition is quite rare, it is an important and potentially prognostic condition underscoring the value of this review.
None.
HC, EH, FM, JAS, and JPS conceptualized the project, gathered the articles, collected the data and wrote the initial draft of the main text. HC and JSW revised the manuscript and wrote the final version of the manuscript. All authors have reviewed the manuscript and approved the final version.
The authors received no funding for this review.
All data underlying the results are available as part of the article and no additional source data are required.
The authors declare no competing interests.
All data underlying the results are available as part of the article and no additional source data are required.