Findings from a study of patients who received radioactive iodine (RAI) treatment for hyperthyroidism show an association between the dose of treatment and long-term risk of death from solid cancers, including breast cancer.
The study, led by researchers at the National Cancer Institute (NCI), part of the National Institutes of Health, was published July 1, 2019 in JAMA Internal Medicine.
“We identified a clear dose-response relationship between this widely used treatment and long-term risk of death from solid cancer, including breast cancer, in the largest cohort study to date of patients treated for hyperthyroidism,” said Cari Kitahara, Ph.D., of NCI’s Division of Cancer Epidemiology and Genetics, lead author of the study.
“We estimated that for every 1,000 patients treated currently using a standard range of doses, about 20 to 30 additional solid cancer deaths would occur as a result of the radiation exposure.”
RAI, which has been used widely in the United States for the treatment of hyperthyroidism since the 1940s, is one of three commonly used treatments for hyperthyroidism.
The other two are anti-thyroid drugs, which have been rising in popularity, and surgical treatment, which is used least often.
The new findings are from a long-term follow-up study of a large cohort of people with hyperthyroidism (mainly Graves’ disease) who were treated with radiation between 1946 and 1964, the Cooperative Thyrotoxicosis Therapy Follow-up Study.
In the new analysis – which included nearly 19,000 people from the original cohort, all of whom had received RAI and none of whom had had cancer at study entry – the researchers used a novel, comprehensive method of estimating radiation doses to each organ or tissue.
Most of the radiation is absorbed by the thyroid gland, but other organs like the breast and stomach are also exposed during treatment.
The researchers observed positive dose-response relationships between the dose absorbed by an organ and mortality from cancer at that site.
The relationship was statistically significant for female breast cancer, for which every 100 milligray (mGy) of dose led to a 12% increased relative risk of breast cancer mortality, and for all other solid tumors considered together, for which relative risk of mortality was increased by 5% per every 100 mGy.
Based on these findings, the researchers estimated that for every 1,000 patients aged 40 years with hyperthyroidism who were treated with the radiation doses typical of current treatment, a lifetime excess of 19 to 32 radiation-attributable solid cancer deaths would be expected.
According to the researchers, in the United States, about 1.2% of the population has hyperthyroidism, and women are much more likely to develop the condition than men.
Therefore, the findings for breast cancer mortality are particularly relevant for the large population of women treated for hyperthyroidism, Dr. Kitahara said.
“We found the increased risks of death from solid cancer overall and from breast cancer more specifically to be modest, but RAI is still a widely used treatment for hyperthyroidism,” she said.
“It’s important for patients and their physicians to discuss the risks and benefits of each available treatment option. The results of our study may contribute to these discussions.”
The researchers wrote that additional research is needed to more comprehensively assess the risk-benefit ratio of radiation versus other available treatment options for hyperthyroidism.
Furthermore, because the types of anti-thyroid drugs administered to patients in the original cohort were different from those prescribed more recently, the researchers wrote that more studies are needed to evaluate long-term health effects of current anti-thyroid drugs, including in comparison to RAI treatment.
Hyperthyroidism is among the most prevalent of medical conditions.
In iodine-replete communities it is found in 0.5–2% of women and is 10 times more common in women than in men (1). It can result in significant short-term and long-term morbidity and, historically, before the advent of effective therapy, significant mortality from cardiovascular disease (2).
Recent population data on mortality in treated hyperthyroidism are inconsistent. Franklyn et al. (3) described a population-based study of mortality in a cohort of 7209 subjects with hyperthyroidism treated with radioiodine in Birmingham, United Kingdom, between 1950 and 1989.
The standardized mortality ratio (SMR) for radioiodine-treated patients (105,028 person-years follow-up) was 1.1 [95% confidence interval (CI), 1.1–1.2, P < 0.001].
Most of the excess mortality occurred during the first year after treatment, with the major association being the hyperthyroidism itself. Other significant factors contributing to the excess mortality were total cardiovascular disease (both cerebrovascular disease and ischemic heart disease) and fractures.
A subsequent population-based study from the Birmingham group (4) investigated 2,668 individuals aged over 40 yr treated with radioiodine for hyperthyroidism between 1984 and 2002 and included 15,968 person-years at risk.
These subjects were treated with relatively low doses of radioiodine, but one dose was effective in 84%, and 45% later had overt hypothyroidism.
Again there was a slight excess in mortality noted in the radioiodine-treated subjects (SMR, 1.14; 95% CI, 1.04–1.24), primarily due to death from cardiovascular disorders often associated with hyperthyroidism such as heart failure and arrhythmias.
However, the increased all-cause and cardiovascular disease mortality was observed only in subjects not taking T4 replacement and included those with subclinical hypothyroidism. The higher mortality in the group not requiring T4 presumably reflected the ongoing deleterious effects of the preceding hyperthyroidism.
The authors therefore argued that the goal of radioiodine therapy should be overt hypothyroidism and that early evidence of thyroid failure should be treated.
In contrast Flynn et al. (5) recently published a large population-based study from Dundee, United Kingdom, of 3888 subjects with hyperthyroidism identified and treated from 1994 to 2001.
This study found no increase in all-cause mortality or serious vascular events in the treated hyperthyroid subjects (21,190 person-years follow-up).
However, among those subjects with hyperthyroidism there was an increase in dysrhythmias (SMR, 3.22; 95% CI, 1.94–5.03).
The reasons for this difference are not clear because no treatment modality was identified. Radioiodine may be more likely to be given to older subjects, and presumably this study included younger patients who are less likely to have cardiac complications and who were more likely to be treated with antithyroid drugs. Furthermore, this study demonstrated an increased frequency of serious vascular events in treated hypothyroid subjects.
In the article in this issue of JCEM Metso et al. (6) have studied 2793 subjects (457 men and 2336 women) of median age 62 yr who received radioiodine treatment (mean dose of 305 MBq) for hyperthyroidism in Finland between 1965 and 2002.
The median follow-up was 9 yr, although there was a wide range from a minimum of 1 month to a maximum of 32 yr 11 months, and included 31,072 person-years at risk.
The major strengths of this study include an age- and gender-matched reference control population, the completeness of follow-up in both the radioiodine treatment group and controls, the ascertainment of the etiology of the hyperthyroidism (including verification by thyroid scintigraphy), and the detail of other treatments for hyperthyroidism at baseline. Fifty-seven percent of subjects had Graves’ disease, whereas the remainder had toxic nodular disease and 80% required only a single dose of radioiodine.
No mortality data were available for those treated with thyroidectomy or long-term antithyroid drug therapy, but this accounted for less than 10% of all hyperthyroid subjects treated.
The analyses also included adjustments for prevalent cardiovascular disease, previous cancer, diabetes mellitus, and age but not smoking history in the subject and the control populations.
An overall increased all-cause mortality of 453 vs. 406 per 10,000 person-years was found in the radioiodine-treated men and women compared with controls [rate ratio (RR), 1.12; 95% CI, 1.03–1.20].
The risk of death appeared to increase soon after radioiodine treatment and was maintained for 25 yr of follow-up.
Mortality from cerebrovascular disease but not ischemic heart disease accounted for the majority of the increased risk in both men and women and was increased significantly in only the subjects older than 60 yr at the time of treatment.
An increase in overall mortality was seen in nodular thyroid disease but not in Graves’ disease, presumably because these subjects were older. The radioiodine subjects had an increased risk of atrial fibrillation compared with controls (mortality, 29.3 vs. 17.5 per 10,000 person-years; RR, 1.68; 95% CI, 1.20–2.34).
The next most frequent increased cause of death was from malignant tumors, mainly upper gastrointestinal tumors, although this was significant only in men over 70 yr. Esophageal cancer caused death in seven subjects compared with two controls, and gastric cancer caused death in 24 subjects and 11 controls.
There was an excess mortality attributed to hyperthyroidism and respiratory diseases observed only in women. Confirming the findings of Franklyn et al. (4), there was a lower mortality in subjects who developed hypothyroidism during follow-up (RR, 0.53; 95% CI, 0.45–0.60).
Those subjects (20%) who received more than one dose of radioiodine had the same risk of dying as those whose hyperthyroidism was corrected by a single dose.
Prior thyroid surgery reduced mortality, presumably because the hyperthyroidism was rendered less severe and more easily treated subsequently by radioiodine, and no difference was observed with antithyroid drugs.
What are the lessons to be learned from this study?
This study confirms that hyperthyroidism is a serious condition with increased mortality.
As the authors state, it is not possible to say whether the increased mortality is due to the disease or the treatment, unless an untreated group is used as a reference.
However, it is far easier to identify and follow-up subjects treated with radioiodine than those treated with antithyroid drugs, and only a small proportion of patients are treated by surgery.
The available data suggest that it is the condition not the treatment modality that is the cause of the excess mortality. It confirms that cardiovascular mortality is increased and that the majority of this increase can be attributed to cerebrovascular disease and atrial fibrillation.
Atrial fibrillation occurs in 5–15% of patients with hyperthyroidism, and hyperthyroidism is the third most common cause of atrial fibrillation after ischemic heart disease and valvular heart disease (7, 8).
Those at greatest risk include those patients over 60 yr of age, men, and those with a coexistence of coronary heart disease, heart failure, and valvular heart disease.
The association with increasing age presumably reflects the age-related reduction in the threshold for developing atrial fibrillation.
Chronic atrial fibrillation carries an annual 3–6% risk of thromboembolic complications, which is five to seven times greater than that of controls in sinus rhythm.
Some controversy still exists, however, as to whether the risk of stroke in patients with hyperthyroidism is lower than believed previously, and hyperthyroidism is rarely included as a subgroup in the studies of outcome of atrial fibrillation (9–11).
The cardiovascular manifestations of hyperthyroidism are best corrected by treating the hyperthyroidism, whether with radioactive iodine or an antithyroid drug.
This study supports the aggressive management of new-onset atrial fibrillation, particularly in patients over 60 yr.
Spontaneous reversion to sinus rhythm may occur in up to 50%, and it typically does so within a few months after restoration of normal thyroid function (12).
Reversion to sinus rhythm is, however, much less likely in the elderly, those with underlying heart disease, and those in whom atrial fibrillation antedated the hyperthyroidism.
The 2006 American College of Cardiology/American Heart Association/European Society of Cardiology guidelines on the management of atrial fibrillation recommended oral anticoagulation with warfarin or other vitamin K antagonist to a goal International Normalized Ratio (INR) of 2.0–3.0, in addition to control of the ventricular rate with a beta blocker (or administration of a nondihydropyridine calcium channel blocker when a beta blocker cannot be used) (13).
These guidelines were similar to those recommended for euthyroid patients who have risk factors for stroke.
Once the patient is euthyroid, the recommendations for antithrombotic prophylaxis if atrial fibrillation persists are similar to those in patients without hyperthyroidism.
Although accepting that no randomized controlled trial data exist, anticoagulation therapy must be recommended if atrial fibrillation does not resolve quickly or there are additional risk factors for thromboembolism, such as heart failure, hypertension, or diabetes. Whether aspirin might be effective in these patients is not known.
Radioiodine is a very effective treatment for hyperthyroidism due to Graves’ disease or toxic adenoma or multinodular goiter and is the most popular treatment for hyperthyroidism in the United States.
A survey over 15 yr ago of North American thyroid specialists in a hypothetical patient with Graves’ disease found that 69% chose radioiodine as the preferred treatment. In Europe and Japan it is less popular, being chosen as first-line therapy by only 22% and 11% of thyroid specialists, respectively (14).
In the United Kingdom it has been recently demonstrated that radioiodine is the most cost-effective treatment modality (15).
Metso et al. (6) have confirmed that effective treatment of hyperthyroidism is crucial and that those rendered hypothyroid and on T4 replacement appear to have the best outcome.
Although patients often express the desire to be euthyroid rather than hypothyroid after radioiodine therapy, it remains advisable to err on the side of overtreatment and advise patients to accept iatrogenic hypothyroidism as a better outcome than risking treatment failure and recurrence of hyperthyroidism with the continued risk of cardiac events and mortality. Hypothyroidism is virtually inevitable over time after radioiodine therapy (16).
Finally, there will be some concern raised regarding the long-term safety of radioiodine therapy and cancer risk.
Overall the data accumulated from over 60 yr of follow-up since the introduction of radioiodine therapy for hyperthyroidism in the 1940s are very reassuring (16, 17).
It is very encouraging that the data presented show no increases in leukemia or thyroid cancer, although they do suggest a small increase in the risk of upper gastrointestinal cancer in elderly men.
This observation has not been confirmed by other studies. However, the radiation exposure to the stomach from the relatively small doses of radiation used in benign thyroid disease is limited.
The known latency period for solid tumors after radiation exposure is more than 10 yr, which suggests that the small increase seen in this study was of doubtful significance. Other environmental factors not recorded such as smoking history, dietary history, or family history may be particularly relevant in these tumors.
In summary, hyperthyroidism is a serious condition with an increased cardiovascular mortality, particularly in the elderly. Radioiodine is recognized as a safe and effective treatment.
Hypothyroidism is an acceptable outcome and possibly even a desirable outcome in the elderly patient with nodular disease. Atrial fibrillation, which is often present, may not be reversible after correction of the hyperthyroid status and needs to be actively treated as per the accepted guidelines including anticoagulation.
Although further reassurance has been gained regarding any link between radioiodine and cancer risk, clinicians must be aware that this study has observed a small increased risk of upper gastrointestinal cancer in elderly men, but this appears to be an isolated finding and has not been confirmed in other long-term studies.
More information:JAMA Internal Medicine (2019). DOI: 10.1001/jamainternmed.2019.0981
Journal information: JAMA Internal Medicine
Provided by National Cancer Institute
