The EANM guideline on radioiodine therapy of benign thyroid disease

Patient candidates for RAI therapy must be evaluated by a nuclear medicine specialist for discussing the indications, logistics, and expectations of RAI treatment. This consultation will also include a review of current medications and specific dietary preparation instructions in anticipation for RAI administration, as well as checking for current and/or prior use of iodine-containing products (Table 1] with resultant decreased radioiodine uptake (RAIU) in the thyroid gland which can reduce RAI therapy efficacy [1, 6, 10‐12]. Notably, RAIU testing and RAI therapy should be performed under the same condition (i.e., preparation protocol, medications).

Pregnancy is an absolute contraindication for RAI treatment. In all fertile females, it can be excluded by a careful history collection and demonstrated by obtaining hCG serum testing just before RAI administration (if not feasible within 72 h maximum). If there is a suspicion of early pregnancy (not detectable by beta-HCG) according to anamnestic data, RAI therapy should be postponed. Patients’ preparation protocols may differ depending on different thyroid diseases and treatment goals. Details are reported in the specific sections.

Any medical examination and treatment involving radiation exposure to patients, treating medical personnel, or members of the public should be carried out in accordance with the safety standards on radiation protection laid down in Chapter VII of the Council Directive 2013/59/Euratom [13].

With regard to the protection of patients treated with radiation, Article 56 of the directive is relevant: “For all medical exposure of patients for radiotherapeutic purposes, exposures of target volumes shall be individually planned and their delivery appropriately verified taking into account that doses to non-target volumes and tissues shall be as low as reasonably achievable and consistent with the intended radiotherapeutic purpose of the exposure.” [13].

Although the member states of the European Union were obliged to transpose the directive into national law, the requirement for individually planned treatments has not generally been adopted for nuclear medicine therapies. Irrespective of the present recommendations for action, medical practitioners must comply with the national formulations of the applicable laws, particularly when measurements of patient-specific biokinetics and retention time functions and verification of the administered radiation absorbed doses can be performed.

Depending on national regulations for treatments with unsealed radioactive agents, it may be necessary to have the treatment performed by specially trained personnel in a dedicated therapy ward with appropriately shielded rooms and suitable radiation protection equipment. Usually, such a ward has forced ventilation with adequate air exchange and is connected to a sewage system for the collection of contaminated wastewater. Treatment should be provided by a team of physicians, nurses, technologists, and medical physics experts with a high level of competence and a clear definition of tasks and responsibilities, considering legal requirements. Written instructions should be available for the handling of radioactive materials, waste, and treated patients. Adequate monitoring of personnel for exposure to direct radiation and/or contamination is required in line with national and regulatory requirements.

Dosimetry in conjunction with RAI therapy for benign thyroid disease can be performed either pre-therapeutically for determining the necessary Na[¹³¹I]I administered activity which will deliver the radiation absorbed dose to the target tissue that is expected to be effective in the particular type of disease, or post-therapeutically for evaluating the energy dose absorbed in tissue. The energy dose D from self-irradiation of a radioiodine accumulating target mass M, which can be a substructure of the thyroid gland such as a hyper-functioning adenoma or the whole thyroid gland such as in Graves’ disease or goiter, is given by the product of the mean energy Ē deposited in M per decay of ¹³¹I and the number of decays represented by the time integral of the activity A(t) in M. A(t) is often written as product Aa· RIU(t), where Aa is the administered activity and RIU(t) the radioiodine uptake, i.e., the fraction of the administered activity in M at a given time t after the administration. The activity necessary to achieve a specified radiation absorbed dose D in the target mass M is [14]:

The constant value recommended in [14] for the mean energy deposited in the target tissue per decay of ¹³¹I,which was calculated for thyroid with M = 20 g, will produce results with adequate accuracy for most of the patients. Mean energy deposition is about 5% higher due to increased photon absorption in goiters with M = 90 g.

Complete dosimetry requires measurements of the mass of the target tissue to be treated and the time function RIU(t) of iodine uptake in this mass. The activity required to achieve a specified absorbed dose is linearly dependent on the target mass. Any error in the mass estimate induces a correspondent error in the activity or absorbed dose calculations. For reliable dosimetry, morphological imaging should be used to measure the target mass. Two-dimensional ultrasound is inferior to three-dimensional ultrasound [15‐17]; however, it is easy to perform, cost-effective, and reliable in most patients [17‐20], and it is recommended in routine diagnostics. Computed tomography (CT) [21], magnetic resonance imaging (MRI) [20], and positron emission tomography (PET) [22] have been described as suitable, but as they are overly expensive and demanding, might however be considered in rare cases of large goiters with substernal extension. Planar and tomographic scintigraphies have limited spatial resolution and are sufficiently accurate only for target masses greater than approximately 20 g [20].

In particular, planar scintigraphy for mass estimation should be reserved for patients in whom the target tissue cannot be delineated or completely visualized in ultrasound, which is considerably more accurate for volume measurement. The measurement must then be performed after administration of Na[^99mTc]TcO₄ using a camera system with high spatial resolution and, after suitable contour finding, the volume must be estimated from the projection area of the accumulating gland [20]. Before planar scintigraphy is applied for volume determination, phantom studies and comparisons on patients with both scintigram and ultrasound measurement must be performed to verify that the contour finding and the equation for calculating volume from area provide sufficiently correct results for the gamma camera and acquisition parameters actually used.

From planar images with high spatial resolution, usually obtained by ultrasound, the mass M of a substructure such as a thyroid lobe or a nodule can be estimated from the lengths of the longest diameter A and 2 orthogonal axes B and C in the transverse section with the largest area as M = A,·B,·C/2, with A, B, and C, in centimeters and M in grams [14]. For large structures approximately shaped like an ellipsoid, the error of the mass estimate is typically less than 20%. Higher errors are possible for very small nodules, if the target mass is not clearly circumscribed in imaging, in patients with toxic goiters with several nodules or disseminated disease, or if the morphological imaging does not match the scintigraphically documented activity distribution. Morphological images used to deduce a mass estimate should always be compared to a Na[^99mTc]TcO₄ or Na[¹²³I]I scintigram in order to verify that the measured volume corresponds to the accumulating target tissue.

The time integral of the uptake function determines the number of decays and thus the absorbed dose in the target mass per unit of activity administered. At least 3 uptake measurements are required to calculate the uptake function. Recommended time points for measurements are 4–6 h after the administration during the initial increase of uptake, after 1–2 days shortly after the time of maximum uptake, and after 5–8 days when the decrease of uptake allows for a reliable estimate of the effective half-life of the activity in the target tissue [14]. Making use of theoretical compartment model considerations [23, 24], the number of uptake assessments can be reduced to two measurements obtained after 1–2 days and after 5–8 days, or even to only one measurement obtained after 5–8 days with minor loss of accuracy [14, 23‐26]. To ensure reliable uptake measurements and consistent dose calculations, it is recommended to follow the standard operational procedures on pre-therapeutic dosimetry provided in [14].

If instead of a late measurement of RIU(t), only the maximum uptake is estimated from an early assessment after time te, the measured value is not representative of the time integral in the denominator of Eq. (1]. The activity to be administered in such an approach should be calculated by [14]:with a reasonable estimate T_est of the effective half-life of the activity in the gland. As recommended in [14], a fixed value T_est = 5.5 days might be used for all patients or disease-specific half-lives if they are known to be representative of the respective patient population. With T_est = 5.5 days, the absorbed dose is a factor of 1.24 higher than expected if the actual effective half-life is 7 days and factors of 1.31 and 1.65 lower for 4 and 3 days actual effective half-life, respectively. Accuracy increases with the increasing time of the measurement after the administration of Na[¹³¹I]I. The potential for error is lower with a measurement after 2 days than with a measurement after one day or earlier.

Activity dosage linear to the mass M (g) without any pretherapeutic assessment of iodine kinetics in the target tissue assumes that the time integral over RIU(t) matches a predefined value. Possible overdosing is limited by the theoretically possible maximum value of the integral, which results from almost complete retention in M and decay with the physical half-life of ¹³¹I. Significant underdosing and thus ineffective treatment is possible in the case of unexpectedly low uptake and/or half-life.

The highest deviations from an adequate target dose D are possible when fixed activities are administered without considering any influence of thyroid tissue mass, radiation absorbed dose, and iodine kinetics on treatment outcome. Increased rates of residual or recurrent disease are expected with large masses and rapid iodine kinetics. Substantial overdosage is possible with small masses, e.g., a fixed activity of 555 MBq is more than twice as high as typically required to effectively treat autoimmune hyperthyroidism in a patient with 20-g thyroid gland and will result in a high radiation exposure that is medically unjustified and not “as low as reasonably achievable (ALARA)” [13]. Administration of fixed activities without any adjustment for the mass to be treated and the expected kinetics is not recommended.

RAI treatment delivery depends on national legislation regarding the therapeutic use of Na[¹³¹I]I. In many countries, RAI therapy is performed in an outpatient setting, reserving an inpatient facility for patients affected by important comorbidities and/or in whom severe complications can be expected, and/or according to Na[¹³¹I]I activity administrated and local legislation. Finally, an inpatient facility should be preferred in cases of bladder or bowel incontinence since it allows for safe disposal of excreta.

RAI therapy must be performed by a trained nuclear medicine physician in the nuclear medicine unit, with the availability of nursing and medical physics support (mainly if a dosimetric approach is planned).

Adequate preparation before performing RAI therapy is very important for reducing the risk of early side effects, such as exacerbation of thyrotoxicosis (i.e., transient rise in FT3 and FT4 levels) due to radiation-induced thyroiditis, occurring in about 10% of all treated patients or, more rarely, the so-called “thyroid storm” (more common in GD patients with poorly controlled hyperthyroidism) [9, 10, 45, 46]. Thus, patients should be ideally treated in a euthyroid state (overt hyperthyroidism should be avoided mainly in patients with active GO). Accordingly, ATD should be stopped 2–5 days before RAI therapy. However, longer withdrawal periods (2–8 weeks) are advised when propylthiouracil (PTU) is used due to its more pronounced radioprotective effect on thyroid tissue than methimazole’s [1, 6, 47‐51]. However, high cure rates were reported after short PTU withdrawal before RAI therapy in patients treated with personalized dosimetric approaches [52]. Using a dosimetry strategy with an intended absorbed dose of 250 Gy (and achieved dose > 200 Gy) excellent ablative results for GD were obtained with only a 2-day PTU withdrawal period before RAIU testing and subsequent RAI therapy delivery [52].

Performing RAI therapy in euthyroid status may be more important in patients with overt hyperthyroidism and/or in elderly patients and/or in patients with comorbidities (e.g., cardiovascular, cerebrovascular, systemic, or pulmonary hypertension, renal failure) [9, 10, 45, 46, 53‐55] for avoiding increased heart rate (due to the so-called T3-toxicosis) and possible, even if rare, supraventricular tachycardia including atrial fibrillation or flutter [9].

If possible, ATD medication should not be used in the days/weeks after RAI therapy administration since it can reduce therapy efficacy [10]. However, in elderly patients with pre-existing heart disease, even if asymptomatic, post-therapy use of ATD(s) could be considered starting 3–7 days after RAI therapy delivery [9].

Patient preparation before administration of RAI therapy includes discontinuation of iodine-containing products (e.g., toothpaste, disinfectant, hair dye) since they can reduce RAIU and, as per consequence, the success rate of RAI therapy.

RAI therapy must be postponed for at least 6 months in patients who received amiodarone treatment, and for several weeks to months for patients who received radiographic contrast agent, depending on the type of substance used, e.g., 6–8 weeks for water-soluble agents, vs. 3–6 months for lipophilic agents. In such patients, RAIU is obligatory before RAI therapy [11, 12] while an iodine urine measurement could be useful as well [6]. However, it is not available in everyday clinical practice of many nuclear medicine departments.

The ß-adrenergic blockade should be used both before and after RAI therapy in patients with severe and symptomatic hyperthyroidism, especially in elderly patients and those with cardiovascular disease, for avoiding side effects such as tachycardia or cardiac arrhythmia [6, 56, 57].

In NTG/TMNG patients, RAI therapy should be postponed if serum TSH is high-normal or elevated for avoiding the effect of RAI on normal thyroid tissue since it increases the risk of developing hypothyroidism [9, 58].

In GD patients with active and mild GO and without risk factors like high TRAb level (based on institutional assay and thresholds) and/or smoking, RAI therapy without steroid therapy should be considered as an acceptable approach mainly in patients with comorbidities such as diabetes and/or osteoporosis. Steroid therapy is recommended in such patients if risk factors are present [9]. In addition, in GD patients with active moderate to severe GO in whom RAI therapy is chosen since other treatments are not feasible, steroid treatment is recommended for minimizing the risk of GO worsening, mainly if concomitant risk factors (e.g., smoking) are present [1, 9, 59]. In such patients, intravenous administration of steroids is recommended using a starting dose of 0.5 g once weekly for 6 weeks, followed by 0.25 g once weekly for 6 weeks [58, 59].

For patients referred for RAI therapy who are at high risk of development/worsening of GD orbitopathy, prophylactic steroids oral administration is recommended using a daily dose of 0.2 mg prednisone (prednisolone)/kg body weight for 6 weeks after RAI therapy [59] or alternatively 0.3–0.5 mg/kg/bodyweight as an initial dose, subsequently gradually tapered and discontinued [58].

Steroid therapy was debated for a long time [60] but is not recommended for patients with present but inactive GO, and non-smoking GD patients without apparent GO by current EUGOGO recommendations [58]. Steroid therapy should not be started before RAI therapy being able to reduce RAIU and consequently the efficacy of the treatment [61]. Accordingly, steroid therapy should be started 1–3 days after RAI therapy and continued for several weeks [1, 4‐6, 9].

Lithium therapy is not frequently used in hyperthyroid patients. However, being able to lengthen ¹³¹I half-life in the gland (without reducing iodine uptake), it may be used in GD patients with low (< 20%) RAIU starting immediately after RAI therapy and continued until 7 days post-therapy. Presently, its clinical significance on RAI therapy effectiveness is not completely clear even if some literature data showed that adjuvant short-term (7 days) use of lithium (900 mg/day) increases RAI therapy efficacy without an increase in side effects (e.g., orbitopathy worsening in GD patients) [62, 63]. Since side effects are currently reported in about 10% of patients, the use of lithium should be evaluated case by case and considered only if RAI therapy is not feasible without it, and alternative treatment options are not available or not feasible [6].

In general, no special diet is strictly required before RAIU testing and subsequent RAI therapy administration. However, it seems sensible to avoid iodinated salt and reduce iodine-rich food (e.g., seaweeds, fish, milk, and eggs), respectively, for at least 7 days before RAIU [1]. In this light, an iodine urine measurement may be useful in selected cases (i.e., patients whose diet is mainly based on seaweed or is rich in supplements containing iodide) [9].

Pregnancy is an absolute contraindication for RAI treatment. In all fertile females, it can be excluded by obtaining hCG serum testing within 72 h before RAI administration. If there is a suspicion of early pregnancy (not detectable by beta-HCG) according to anamnestic data, RAI therapy should be postponed. Finally, pregnancy must be avoided for almost 4 months after RAI therapy [64]. Breastfeeding must have been discontinued for at least 6 weeks before RAI therapy administration for allowing time for mammary tissue involution with the goal of avoiding high radioiodine uptake and, consequently inadvertent radiation to hypertrophied mammary gland tissue [6, 9, 11, 65].

The nuclear medicine physician must provide verbal and written information to patients prior to therapy administration, addressing the following:

The patient’s written informed consent must be obtained before RAI therapy administration.

Na[¹³¹I]I is preferentially administered orally as a capsule since it allows an easier and safer treatment than using the liquid form or intravenous administration, thus reducing the risk of contamination. Intravenous administration can be used in patients with parenteral nutrition, vomiting, or other causes of the inability to process iodine through the gastrointestinal tract. The liquid form can be used in patients with severe swallowing difficulties; on the other hand, oral administration is not feasible in some countries due to legislative compliance, and therefore, intravenous administration may be required even for patients with swallowing difficulties. In addition, liquid Na[¹³¹I]I is generally supplied for intravenous injection, and therefore, administering liquid orally would be classed as an unlicensed way of administering Na[¹³¹I]I (off-label) with different absorbed doses to the mouth. If oral administration is chosen, as happen in most cases, patients have to be fasting overnight before RAI therapy and 1–2 h after treatment. Following RAI therapy administration, patients should be encouraged to drink a sufficient quantity of fluid for 24–48 h for reducing the radiation absorbed dose to the bladder (critical organ) and minimize body radiation exposure by enhancing urinary RAI excretion.

The administered Na[¹³¹I]I therapeutic activity depends both on the aim (e.g., ablative vs. functional therapy in GD patients) and the strategy (fixed-activity method vs. dosimetric approach) of RAI treatment.

Accordingly, the patients’ restrictions regarding work and with respect to family members and the general population will be different. In general, patients have to be advised to increase the distance and reduce the contact time when interacting with others (mainly children, pregnant, and breastfeeding women) for reducing radiation exposure that in Europe should not exceed 1 mSv cumulative dose per year [6, 13, 64, 66–68].

The administration of fixed RAI activity is the easiest and most commonly used method for performing RAI therapy in GD patients for the definitive treatment of hyperthyroidism.

The fixed-activity method was originally based on the administration of a fixed activity to all patients. Moreover it may be refined by estimating the thyroid gland size by palpation, morphological, or functional imaging (i.e., TUS or thyroid scintigraphy, respectively) (i.e., fixed administered activity per mass unit) [1, 6]. In addition, Na[¹³¹I]I thyroid gland maximum uptake and turnover should be empirically taken into account when therapeutic RAI activity is planned [88].

Most often, fixed activities between 370 and 555 MBq are employed to treat GD patients [1, 9, 11, 75]. However, higher activities may be required when the thyroid volume is higher than 40 ml. In this setting, a dosimetric approach should be considered [89].

In daily clinical routine, the main advantage of the fixed-activity method is the simplicity of pre-treatment logistics. However, a potential disadvantage of the fixed-activity method is the potential risk of over-treatment, associated with unnecessarily high radiation exposure to the patient, or under-treatment with a higher cumulative incidence of persistent/recurrent hyperthyroidism [1, 78, 90, 91]. Notably, however, neither randomized clinical trials nor retrospective comparison studies demonstrated the inferiority of the fixed-activity approach compared to dosimetric approach.

Results of studies reporting the relationship between the thyroid radiation absorbed dose and treatment outcome are conflicting [92]. However, a multicentric prospective randomized trial [93] and a systematic review and meta-analysis including the most relevant retrospective studies [7] on RAI therapy of Graves’ disease with complete dosimetry showed that there is a significant dependence of treatment success rates on the actual thyroid radiation absorbed dose. According to the combined analysis [7], the delivery of 150 Gy, 200 Gy, and 300 Gy to the thyroid gland is expected to eliminate hyperthyroidism in 74%, 81%, and 88% of cases, and restore euthyroidism in 38%, 35%, and 29% of cases, respectively. Besides the thyroidal radiation absorbed dose, a larger size of the thyroid gland (often associated with fast iodine kinetics) appears to be an independent risk factor for residual disease [76, 93, 94].

A target dose of at least 200 Gy is recommended for the definitive treatment of patients with refractory Graves’ disease. Up to 300 Gy is appropriate especially for large glands and very pronounced hyperthyroidism if a more reliable therapeutic effect is desired in the short term. If restoration of a euthyroid state is reasonably possible in milder disease with a low risk of recurrence and is desired by the patient, approximately 150 Gy target dose should be aimed for. The success of the therapy is not guaranteed due to uncertainties in determining the therapeutic RAI activity and individual variables that have not yet been sufficiently investigated. Patients should be informed that treatments with lower target radiation absorbed dose are more likely to require retreatment and that the likelihood of hypothyroidism requiring lifelong thyroid hormone replacement increases with an escalation of the target radiation absorbed dose. Even if euthyroidism is successfully restored after treatment, regular long-term monitoring of hormonal status remains necessary because the rate of hypothyroidism increases continuously over years following RAI therapy.

If the therapeutic RAI activity is estimated from an early uptake measurement according to Eq. (3], an estimate of the effective half-life of T_est = 5.5 days suggests that approximately 6, 8, and 12 MBq multiplied by the mass in gram and divided by the 24-h Na[¹³¹I]I uptake (RAIU) are to be administered for target doses of 150, 200, and 300 Gy, respectively.

Example: A patient with 35-g thyroid mass and 65% 24-h Na[¹³¹I]I uptake, RAIU(t = 24 h) = 0.65, who is to be treated with a thyroid absorbed dose of 200 Gy, should be given Aa = 8·35/0.65 MBq = 430 MBq. In the case of fast iodine kinetics, recognizable by an early and high iodine retention and high Technetium Thyroid Uptake (TcTU) in the Na[^99mTc]TcO₄ scintigram, a shorter half-life in the tissue and therefore underestimation of the absorbed dose is to be expected. Especially with large thyroid glands, which might be less sensitive to radiation, this can lead to high failure rates [94] justifying to aim for absorbed doses higher than 200 Gy.

If a pre-therapeutic assessment of the kinetics is omitted, activity dosage should be linear to the mass M assuming that about half of the administered activity decays in the gland, corresponding, e.g., to iodine kinetics with 79%, 67%, or 57% 24-h Na[¹³¹I]I uptake and 4.5, 5.5, or 6.5 days, respectively, the effective half-life of the activity in the gland. The therapeutic activity is calculated as 9, 12, and 18 MBq multiplied by the mass in grams for target doses of about 150, 200, and 300 Gy, respectively. Example: A thyroid gland with 20-g mass is to be treated with Aa = 9·20 MBq = 180 MBq for an absorbed dose of 150 Gy.

As for GD patients, the administration of predefined fixed RAI therapeutic activities is the easiest and most commonly used method for the delivery of RAI therapy in TNG and TMNG patients.

According to this method, the estimation of hyperfunctioning nodule(s) volume is obtained by palpation or preferably by TUS or TS [1, 6] and ¹³¹I nodule(s) maximum uptake and turnover (evaluated by RAIU trend) should be considered for RAI therapy planning.

However, the fixed-activity method has limited accuracy [1, 91], and consequently, the cumulative incidence of persistent/recurrent hyperthyroidism (mainly in TMNG patients) or, conversely, of hypothyroidism (mainly in TNG patients whose extranodular parenchyma is not suppressed) is not negligible [1, 78, 90, 118].

Most often, fixed activities ranging from 370 to 740 MBq are used to treat TNG or TMNG patients [1, 9, 11].

RAI therapy efficacy is limited when hyperfunctioning nodule(s) volume is > 26 ml. In such patients, a dosimetric approach is considered more useful for obtaining an effective treatment [89].

The fixed-dose method offers the same advantages and limitations already discussed for RAI therapy in GD patients.

In non-autoimmune toxic goiter, dosimetry should ideally only take into account the kinetics in the autonomously functioning tissue, whereby not all, but only part of the thyroid gland parenchyma is affected, typically in the form of one or more nodules. With uptake probes usually used to measure activity kinetics [14], dosimetry is reliably possible if the mass of the autonomic tissue can be measured and the healthy thyroid tissue is completely suppressed. Underdosing and the consequent risk of persistent or recurrent hyperthyroidism occurs when a non-negligible fraction of the measured activity is located in unsuppressed healthy tissue or autonomously functioning tissue outside of the nodules taken into account in the mass determination. The main risks for developing hypofunction are incomplete suppression of the remainder of the gland and a high targeted absorbed dose in a large target volume. In toxic goiter with a single or only a few reliably measurable nodules, RAI treatment based on complete dosimetry of the autonomously functioning volume eliminates hyperthyroidism in 85–100% of individuals treated with 300–400 Gy target dose with rates of hypothyroidism typically between 10 and 20% [112, 113].

When calculating the target dose for the treatment of toxic nodule, the measured iodine uptake must be corrected by the nodule/whole thyroid uptake ratio obtained from the scintigram. The target dose for the nodule is 300–400 Gy.

When the estimation of a target mass is not possible, absorbed doses of 100–150 Gy to the mass of the entire gland should be applied to calculate the therapeutic RAI activity [92, 110, 112].

If complete dosimetry is omitted, activity dosing should be based on an early uptake measurement, as indicated by a meta-analysis [123]. Equation (3] suggests administering approximately 4, 6, 8, 12, or 16 MBq multiplied by the mass in gram and divided by the 24-h Na[¹³¹I]I uptake for target doses of 100, 150, 200, 300, or 400 Gy, respectively. The suggested value of 5.5 days for T_est is about 10% less than the mean effective half-lives typically observed in most toxic nodular goiters [110, 113] but limits the potential error to ± 25% for the vast majority of individuals [124]. Examples: A patient with a 10-g nodule with 30% 24-h Na[¹³¹I]I uptake should be administered Aa = 12·10/0.3 MBq = 400 MBq for a nodule absorbed dose of 300 Gy. Another patient with multinodular goiter in a gland with 30-g total mass and 40% 24-h Na[¹³¹I]I uptake needs Aa = 6·30/0.4 MBq = 450 MBq for a mean thyroid absorbed dose of 150 Gy.

The dosimetric approach following the European standard operational procedure provides the most accurate results regarding achieved and intended target doses. The Marinelli formula seems to enable minimal deviations of achieved doses from target doses in TMNG, although other individualized strategies taking into account at least the thyroid volume also result in low deviations of presumed achieved dose from the intended dose [14, 125].

The patient facility can be different according to national legislation on the therapeutic use of ¹³¹I. Please refer to the paragraph on RAI therapy for hyperthyroidism.

As per hyperthyroid patients, iodine-containing products such as administration of amiodarone or radiographic contrast agents should be discontinued well in advance of therapy, or better avoided altogether since they reduce RAIU and, consequently, RAI therapy efficacy can be decreased [1, 6, 9‐12]. Patients with (very)-large NTNG (≥ 100 ml) and severe trachea narrowing (especially if < 1 cm) should always be treated under steroid coverage (e.g., prednisolone, 25 mg daily for 14 days) [6, 126]. Moreover, consideration for inpatient RAI therapy, mainly for elderly patients with comorbidities (e.g., heart failure), will ensure a hospital setting for addressing possible post-treatment complications [6, 144, 158, 170].

Patient preparation, logistical information, and instructions for the delivery of RAI therapy for NTG patients are similar to hyperthyroid patients regarding:

As for hyperthyroid patients, Na[¹³¹I]I is preferentially administered orally in the form of a capsule but liquid form or intravenous administration can be used in patients with severe swallowing difficulties or patients with parental nutrition, respectively.

For rapid 131-I absorption, it is recommended that oral RAI therapy is administered on an empty stomach and this condition must be maintained in the following 1–2 h.

To reduce the radiation absorbed dose to critical organs and diminish radiation exposure to the entire body, patients should be encouraged to drink a large quantity of fluid for at least 24–48 h following RAI therapy.

The prescribed Na[¹³¹I]I therapeutic activity largely depends on the target thyroid volume, the RAIU values, and radioiodine distribution in the gland.

Accordingly, the instructions regarding patients’ restrictions on work, with family members, and with respect to the general public will vary depending on the amount of prescribed radioactivity.

Prior to RAI therapy administration all patients must receive detailed information for reducing both radiation exposure and contamination risk to others, particularly to children, pregnant, and breastfeeding women [6, 13, 64, 66, 67,68].

The easiest and most common method used for performing RAI therapy in NTNG patients is the administration of fixed RAI activities.

According to this method, the estimation of the target treatment volume is obtained by palpation and/or TUS and/or CT or MR (in cases with sub-sternal extension), functional mapping of the goiter is obtained by TS, and RAIU testing should be performed for RAI therapy planning [1, 6, 136, 160, 171‐178].

There is no consensus regarding the fixed RAI activity prescribed for NTG treatment and a wide range of RAI activities have been reported [92, 126, 136, 160, 171‐179]. Most often, fixed activities ranging from 600 to 1110 MBq are used to treat NTNG patients [126, 136, 159, 160, 171‐177].

Although the volumes of euthyroid goiters often are substantially increased, the biokinetics usually corresponds approximately to those of healthy thyroids with normal or only moderately increased radioiodine uptake and a long half-life within the gland [168]. After pre-therapeutic dosimetry, treatment should target a mean thyroid absorbed dose of about 100 Gy [134], in selected cases up to 150 Gy. An often-used activity of 3.7 MBq multiplied by the mass in gram and divided by the 24-h ¹³¹I uptake administers a mean thyroid absorbed dose of 120 Gy to a large goiter treated for size reduction if the effective half-life of the activity in the gland is 7 days.

The dosimetric approach following the European standard operational procedure provides the most accurate results regarding achieved and intended target doses [14, 125].

Although rhTSH use was initially restricted to thyroid cancer patients, it has been used since the early 2000s to increase thyroidal RAIU in patients with NTG [173, 180, 181]. Administration of rhTSH is associated with goiter reduction enhancement by 35–55% as compared with Na[¹³¹I]I therapy without rhTSH stimulation. The adjuvant role of rhTSH in RAI therapy for NTG has been studied mainly for patients with very large goiters and in those with a low baseline RAIU [179]. However, rhTSH is not yet approved for this indication, and administration in a patient with NTG represents an off-label use. Therefore, the current guidelines cannot endorse the general use of rhTSH for the treatment of NTG, unless rhTSH-aided RAI therapy is regulated according to local rules (e.g., by authorization of the local Ethical Committee) and cautiously chosen individually for each patient in view of the risks, benefits, experience of the treating physician, and patient’s personal preference.

Declining efficacy with increasing goiter size and low thyroidal RAIU are the major challenges in conventional RAI therapy for NTG, and these factors represent the main indications for rhTSH-stimulated RAI therapy [182]. The potential benefits of rhTSH-aided RAI therapy are based on the marked increase in thyroidal RAIU, resulting in increased radiation absorbed dose to the thyroid without increasing the administered RAI activity (superiority strategy) or reduction of the administered RAI activity while maintaining the same thyroidal radiation absorbed dose (dose-reduction or equality strategy) [183]. According to published literature, rhTSH should be administrated 24 h before RAI therapy (optimal time interval) rather than 2 h, 48 h, or 72 h for maximizing the thyroidal RAIU increase [92, 126, 168, 184]; however, no consensus exists about the dose of rhTSH for aiding RAI therapy. Although in most studies a dose of 0.1–0.3 mg has been given, a wide range of doses (i.e., 0.005–0.9 mg) have been used [92, 126, 174, 185]. However, the choice of rhTSH dose to use for aiding RAI therapy is strongly related to the purpose of the treatment. Indeed, a dose of 0.1 mg is recommended in the equality strategy, whereas a higher rhTSH dose is employed (i.e., 0.3 mg) coupled with higher prescribed therapeutic RAI activity in the superiority strategy [92], resulting in a higher goiter reduction as compared to the equality approach [92]. All in all, based on the present knowledge the optimal rhTSH dose for aiding RAI therapy is most likely in the range of 0.03–0.1 mg. Indeed, in this dose range, a significant improvement of the thyroidal RAIU is obtained while minimizing the risk of side effects [126].

The rationale for using rhTSH-aided RAI therapy in NTG is to obtain an enhanced goiter volume reduction (GVR) which is maintained for several years after therapy with the goal of reducing the need for additional therapy, such as a second RAI treatment or thyroid surgery. However, little is known about the long-term risk of goiter recurrence after RAI therapy, whether with or without rhTSH stimulation [186]. Reported adverse events include administered RAI activity-dependent increased risk of thyroid hyperfunction and acute thyroid swelling when RAI therapy is preceded by rhTSH stimulation. While other serious acute side effects seem of little concern when using low rhTSH doses, the more pronounced goiter reduction, seen with rhTSH-stimulated RAI therapy, is unfortunately achieved at the expense of an up to fivefold increase in the rate of permanent hypothyroidism [175, 182].

Pretreatment with a single, low rhTSH dose allows a reduction of the prescribed RAI activity in patients with nodular goiter and significantly enhances goiter volume reduction in some studies. Notably, however, available evidences do not support its routine clinical use, and the rhTSH is not EMA-approved for this indication. Its out-of-label use requires strict compliance with local laws and recommendations under the responsibility of nuclear medicine physicians.

Autonomous nodules or obstructive goiter are very rare in children; nearly all patients referred for RAI therapy have been diagnosed with GD refractory to medical therapy with ATD. Therefore, unless explicitly stated otherwise, the recommendations and considerations in this section will pertain to GD.

The reluctance for employing RAI therapy in pediatric GD increases with decreasing patient’s age. According to the recently published European Thyroid Association Guidelines [82] on the treatment of Graves’ disease in children and adolescents, Na[¹³¹I]I can theoretically be used in any patient with GD. However, a very young age (< 5 years; because of a greater long-term theoretical risk of malignancy) is seen as a near-absolute contraindication. Age 5–10 years is regarded as a relative contraindication. Interestingly, an update in this guideline is a recommendation that ATD should be maintained for multiple years in pediatric GD patients so that rather than being referred for surgery, children with a near-absolute or relative age-based contraindication for RAI therapy patients can often be kept on ATD until they reach an age adequate for undergoing RAI therapy.

Other than the standard preparations as described for adults (e.g., stopping ATD), no special medical preparations are necessary for pediatric patients. Of course, depending on the age of the patient and the possibilities allowed by local radiation protection regulations, it may be possible or even necessary to allow a parent to be co-hospitalized with the patient after appropriate preparation of the parent/caregiver.

Thyroid swelling is an early side effect that can be observed in patients with large toxic or, more frequently, non-toxic goiter. Patients could manifest thyroid pain and sensation of anterior neck discomfort due to the inflammation process caused by irradiation of thyroid tissue (i.e., radiation-induced thyroiditis). There are rare reported cases of elderly patients with very-large nodular goiters experiencing rapid thyroid enlargement due to edema which may cause tracheal compression and dyspnea. This possible side effect should always be considered, particularly when rhTSH-aided RAI therapy is planned in patients with large NTG or patients with known tracheomalacia [1, 160].

The aforementioned symptoms are generally mild or moderate. As per consequence, medical therapy is not usually required or may be considered to alleviate local pain symptoms (e.g., paracetamol). However, in patients with higher risk factors (e.g., larger goiter size), they can be prevented/treated using a non-steroidal anti-inflammatory agent for 24–48 h after RAI administration while corticosteroids should only be used in selected patients since it can reduce RAI therapy efficacy [1, 61].

The prevalence of radiation thyroiditis is very low with about 1% of all treated patients. It may occur during the first weeks after RAI therapy and can produce a transient rise of serum thyroid hormone levels due to follicular cell damage.

As per consequence, a mild to a severe worsening of thyrotoxicosis may occur in about 10% of patients, mainly among those with poorly controlled hyperthyroidism before RAI therapy [54]. Conversely, the so-called thyroid storm is a rare but life-threatening condition that requires immediate diagnosis, comprehensive, and advanced medical treatment using anti-thyroid drugs (ATDs), inorganic iodide, beta-adrenergic receptor antagonists, corticosteroids, and antipyretic drugs [191] and, if all else fails, emergency thyroidectomy.

A careful selection of the optimal time point for performing RAI therapy and correct patients’ premedication are mandatory for avoiding/reducing the risk of thyrotoxicosis worsening and, especially, thyroid storm condition [1].

Although it is a more frequent complication in case of RAI therapy for differentiated thyroid cancer, radioiodine-induced sialadenitis can also occur in hyperthyroid patients, especially if treated with a high RAI activity (e.g., ≥ 1000 MBq), both as an acute and a late side effect [1]. Iodine is concentrated in the salivary gland by the NIS and then secreted into saliva. During this process, salivary glands are exposed to dose-dependent radiation damage [192]. Clinical symptoms include salivary gland swelling (frequently transient), local pain, xerostomia, or dysgeusia (taste dysfunction), which may occur in as many as 20–30% of patients [192, 193]. The use of lemon juice and/or salivation-inducing snacks (like lemon/orange candy), starting 24 h following RAI therapy is a valid radioprotective procedure to diminish RAI damage to salivary parenchyma [1, 194‐196].

The so-called Marine-Lenhart syndrome [197] is a rare side effect of RAI treatment observed in TNG patients during the first days after RAI therapy, with a reported prevalence ranging from 2.7 to 4.1% [198, 199]. However, its incidence is significantly higher in patients with elevated serum levels of thyroid peroxidase antibody (TPOAb) before RAI therapy, as well as in patients who became TPOAb positive after RAI therapy [198, 200‐203]. These observations support the hypothesis that immunogenic hyperthyroidism in such patients represents an exacerbation of a pre-existing and occult immunogenic thyroid disorder [1].

RAI administration, as well as surgery, could produce a transient increase of TRAb due to the release of thyroid antigens from damaged follicular cells. As per consequence, autoimmune hyperthyroidism can be observed in such patients, but no correlation has been found between serum TRAb rise and RAI activity administered [1, 204].

The most serious consequence of immunogenic side effects is the activation or new onset of Graves’ orbitopathy (GO) [92].

Persistent hypothyroidism can be counted among side effects of RAI therapy for TNG or NTNG patients [1, 92]. However, in these patients, thyroid surgery will produce hypothyroidism too [92]. Conversely, hypothyroidism represents the optimal result (i.e., main goal) when RAI therapy is performed in GD patients according to the newest ablative dose concept [1, 6, 9, 52, 76, 77]. However, some patients experience transient hypothyroidism that may manifest 2–5 months after RAI therapy and spontaneously recover within a few months without the development of symptoms or signs of hypothyroidism [1, 205]. Differentiating between transient and permanent hypothyroidism in the early months following RAI therapy remains a clinical conundrum that requires adequate clinical and laboratory follow-up to establish the optimal treatment plan, especially for GD patients with orbitopathy [1, 206, 207].

Several studies demonstrated a significant association between RAI therapy and the development or worsening of GO [1, 208]. In a systematic review that included nine studies comprising a total of 1773 patients, Li et al. reported that RAI therapy is a significant risk for new-onset or worsened orbitopathy as compared to ATD treatment (p < 0.00001), while prophylactic administration of steroids has been demonstrated to be an effective method for preventing this event (p = 0.002), especially in patients with moderate orbitopathy [208]. High TRAb levels before RAI administration, smoking, and uncorrected post-treatment hypothyroidism are associated with a more aggressive orbitopathy and a worse prognosis [70]. For this latter reason, thyroid function should be always evaluated in the first weeks after RAI therapy to avoid the risk of development or worsening of GO due to uncorrected hypothyroidism. However, early (i.e., after 2 weeks following RAI therapy) levothyroxine substitutive treatment has been already proposed for preventing hypothyroidism [207]. In GD patients at high risk for the development/worsening of orbitopathy following RAI therapy, the administration of prophylactic oral steroids is recommended using a daily dose of 0.2 mg prednisone/kg body weight for 6 weeks after RAI therapy [59]. Conversely, intravenous administration of steroids is recommended in GD patients with moderate to severe active GO using a starting dose of 0.5 g once weekly for 6 weeks, followed by 0.25 g once weekly for 6 weeks [59].

There is no convincing evidence for an increased risk of secondary cancers and/or increased mortality in patients treated by RAI therapy [209‐211].