2025 Volume 72 Issue 1 Pages 23-36
Almost a century has passed since Plummer reported the efficacy of short-term preoperative inorganic iodine therapy for Graves’ disease in the 1920s. Since there were concerns about the escape phenomenon and exacerbation with inorganic iodine, antithyroid drugs became the mainstay of pharmacotherapy for Graves’ disease following their development in the 1940s. With regard to long-term inorganic iodine monotherapy, Trousseau reported a case in the 1860s, and several subsequent reports suggested its efficacy. Around 1930, Thompson et al. published a number of papers and concluded that long-term inorganic iodine monotherapy was useful if limited to mild cases under careful follow-up. From Japan, in 1970, Nagataki et al. reported that, of 12 patients treated with inorganic iodine, three remained eumetabolic for more than two years. Since 2014, some reports have also been published from Japan. A summary of these recent reports is given below. The starting dose of potassium iodide is around 50 mg/day, and candidate responders have mild disease, with FT4 <2.76 ng/dL (35.5 pmol/L), a small goiter, and are female and elderly. Response rates are relatively high, at 60–80%, and the remission rate is about 40%. In cases of insufficient response, changing therapy should be considered. Inorganic iodine can be used as a possible alternative if the patient experiences adverse events with antithyroid drugs and/or prefers conservative treatments, with an understanding of their efficacy and limitations. These recent reports have been published from Japan, where iodine is sufficient, and the dose of inorganic iodine is empirical and requires further study.
Almost a century has passed since Plummer reported the efficacy of inorganic iodine as a short-term preoperative therapy for Graves’ disease in 1923 [1]. With the development of antithyroid drugs (ATDs) in the 1940s [2], ATDs became the mainstay of pharmacotherapy for Graves’ disease, since inorganic iodine was thought to have a short-term effect, and there were concerns about the escape phenomenon and exacerbation. On the other hand, Trousseau administrated long-term inorganic iodine to a patient with Graves’ disease in 1863 [3]. Subsequently, reports of long-term inorganic iodine monotherapy for Graves’ disease were published from the late 1800s to the early 1900s. In 1930, Thompson et al. reported detailed cases of long-term inorganic iodine monotherapy, showing its usefulness in mild cases [4]. However, treatment efficacy was based on basal metabolism, pulse rate, and body weight, which are very different and not comparable to recent methods of assessing efficacy.
The main treatment for Graves’ disease in Asia and Europe is ATDs, whereas in the USA, radioiodine therapy (RAIT) used to be administered to most adult patients. However, recent reports indicate that the frequency of cases opting for RAIT has decreased, and more than half of the patients in the USA are receiving ATDs due to concerns about possible worsening of thyroid eye disease after RAIT [5, 6], etc. Pharmacotherapy with ATDs is now the main treatment for Graves’ disease worldwide. However, ATDs have the disadvantage of a high incidence of adverse events, some of which are rare but serious [7, 8]. It may be worth being reminded about long-term inorganic iodine monotherapy as a possible alternative to ATDs in the pharmacotherapy of Graves’ disease. In recent years, there has been a series of reports from Japan on long-term inorganic iodine monotherapy for Graves’ disease [9-13]. This paper aims to provide a narrative review of these reports, after summarizing the information on iodine up to recent years, to clarify its effectiveness and limitations.
Iodine is required for the synthesis of thyroid hormones (THs) and is an essential element for humans. Iodine is the rate-limiting substrate for TH synthesis, and TH production is dependent on the nutritional availability of iodine. Iodide (I-) is actively transported by the sodium-iodide symporter (NIS/SLC5A5) on the vascular side of thyroid follicular epithelial cells. On the luminal side, pendrin (SLC26A4) mediates the efflux of iodine into the lumen of the follicle. On the luminal side, iodine is oxidized by thyroid peroxidase (TPO) under H2O2 generated by dual oxidase 2 (DUOX2), requiring DUOXA2 for maturation and transport from the endoplasmic reticulum to the plasma membrane, and tyrosine residues are iodinated. This process produces monoiodotyrosine (MIT) with one iodine bond and diiodotyrosine (DIT) with two iodine bonds, followed by coupling reactions to produce triiodothyronine (T3) with three iodine bonds or thyroxine (T4) with four iodine bonds. The thyroglobulin-TH complex is reabsorbed into the follicular epithelial cells by endocytosis, and the thyroglobulin is hydrolyzed in lysosomes to release T3 and T4. The THs thus produced are secreted into the circulation by hormone transporters such as monocarboxylate transporter 8 (MCT8/SLC16A2) [14]. MIT and DIT are deiodinated by iodothyrosine dehalogenase (IYD/DHAL1) at the apical pole of the thyroid follicular epithelial cells for iodine recycling [15].
Deficiency or excess of iodine intake causes a variety of effects. Iodine is abundant in soil and water in coastal areas, and its global distribution is uneven. Iodine deficiency, which causes goiter, myxedema, and congenital hypothyroidism i.e., cretinism, has been a global problem. Severe and chronic iodine deficiency leads to diffuse enlargement of the thyroid gland, as well as nodular changes, and leads to a risk of developing toxic multinodular goiter, especially in elderly women. Before the association with iodine deficiency was established, many people around the world had these symptoms of unknown cause. In 1811, the French chemist Bernard Courtois accidentally discovered iodine during the manufacture of gunpowder. When heated, it produced a violet-colored vapor, hence named ‘iode’ from the Greek word ιώδης, meaning violet [16]. In 1820, Coindet discovered that iodine administration in Switzerland, a severely iodine-deficient country, reduced goiter [17]. In 1851, Chatin reported a high incidence of endemic goiter and cretinism in iodine-deficient areas [18]. This led to the widespread use of iodine as a treatment for goiter. Hunziker in 1915 and Marine and Kimball in 1918 proved the effectiveness of iodine prophylaxis in the prevention of endemic goiter and cretinism [19]. Subsequently, salt iodization spread worldwide and became national programs. According to the Iodine Global Network (IGN), a non-profit, non-governmental organization aiming for the sustainable elimination of iodine deficiency diseases worldwide, the number of countries with iodine-deficiency among schoolchildren decreased from 54 in 2003 to 21 by 2021 [20].
However, excessive iodine intake can also lead to goiter and hypothyroidism [21, 22]. A case of Japanese coastal goiter reported from a region with high ‘kombu (kelp)’ consumption showed that iodine restriction alone caused a reduction in goiter that improved more after TH replacement [23]. In addition to iodine excess due to dietary intake, gargles containing povidone-iodine and medicines such as amiodarone (containing 37.5 mg of iodine per 100 mg) may cause drug-induced hypothyroidism. In Japan, iodine excess has been of some concern due to the common diet of iodine-rich ‘kombu (kelp)’ and ‘kombu-dashi’ (Japanese soup stock made from kelp) [24]. In a recent study by Fuse et al., 32,025 schoolchildren aged 6–12 years without known thyroid disease from 46 geographical locations in prefectures across Japan were analyzed [25]. Although regional differences exist within Japan, with 12 of 46 regions showing excess, median urinary iodine levels from spot urine samples from the cohort indicated that iodine intake in Japan is adequate overall, consistent with the IGN’s current classification of Japan as ‘adequate.’
Recommended intakes and upper limits are given according to age and sex, and increases are preferred in pregnant and breastfeeding women. In the USA, the recommended daily intake is 150 μg for adults [26]. The maximum tolerated dose is estimated to be up to 1,100 μg/day in the USA [27] and 600 μg/day in Europe [27]. In Japan, the recommended daily intake is 130 μg/day for adults, with a maximum tolerated dose estimated at 3,000 μg/day [28].
As described above, iodine began to be used as a treatment for endemic goiter in the 19th century. Later on, it became known that treating goiter with iodine caused symptoms such as tachycardia, tremor, and weight loss, and people were warned to watch out for symptoms of hyperthyroidism when treating goiter with iodine. In 1910, Kocher coined the term ‘Jod-Basedow’ for the phenomenon that iodine treatment of goiters induced Graves’ disease [29]. In the 19th century Carl von Basedow [30], Robert James Graves [31], and others reported the current Graves’ disease and recognized it as a disease. It is assumed that goiters treated with iodine included exophthalmic goiter (Graves’ disease), as well as diffuse goiters and nodular goiters including autonomously functioning thyroid nodules, which are more common in iodine-deficient areas. It might be possible that some cases of ‘Jod-Basedow’ had latent functional nodules, and thyroid function might have worsened after the administration of iodine.
In 1923, Plummer of the Mayo Clinic in the USA administered iodide as Lugol’s solution (LS) preoperatively to patients with Graves’ disease. The signs and symptoms of hyperthyroidism disappeared within a few days, and the surgical mortality rate decreased from approximately 4–5% to less than 1%, showing a remarkable improvement in outcome [1, 32, 33]. Since these reports, the importance of administering high doses (100 to 400 mg) of inorganic iodine to patients with Graves’ disease in preparation for thyroidectomy in a short period was rapidly recognized. It was soon pointed out that the effect was usually only temporary, and that after a few weeks, the patient’s condition returned to the same level as before starting iodide, if it did not worsen. Thus, long-term therapy of Graves’ disease with iodine alone has come to be considered a treatment of limited benefit.
Conversely, there had been reports of beneficial results with long-term inorganic iodine monotherapy for Graves’ disease. In 1863, Trousseau in France administrated inorganic iodine by chance and observed long-term improvement in the symptoms of exophthalmic goiter [3]. Subsequently, from the late 1800s to the early 1900s, long-term treatment with small amounts of inorganic iodine alone for months to years was reported to be effective and safe in some patients, to completely remit the disease, and to eliminate the need for surgery, as noted in a paper by Thompson et al. [4]. In 1930, Thompson et al. published a detailed report on the effects of long-term treatment with inorganic iodine alone in 24 cases of exophthalmic goiter at Massachusetts General Hospital in the USA [4], and the clinical features were summarized as follows. Twenty-four patients (14 mild and 10 severe or moderate) were treated continuously or intermittently with iodine alone for 1.5 months to 3 years, with 13 patients being treated for more than 1 year. Of the 14 mild cases, treatment was completely successful in 9 cases (64%), led to some improvement in 3 cases (22%), and was insufficient in 2 cases (14%). In 10 cases of severe or moderate disease, one (10%) had an adequate response, four (40%) had an inadequate response, and five (50%) had a worse response. These results suggested that long-term inorganic iodine monotherapy was useful in the treatment of mild cases (total response rate of 86%), and that moderate or severe exophthalmic goiters rarely (10%) show more than temporary improvement, namely, long-term inorganic iodine monotherapy was not contraindicated if limited to mild cases, and patients were carefully monitored. Thompson et al. also published a number of studies [34-38] attempting to determine the smallest effective dose of iodine and found that small amounts of iodine (1.5, 3, and 6 mg/day) had effects in hospitalized patients with exophthalmic goiters on controlling basal metabolism. Referring to a previous study showing sufficient efficacy of approximately 6 mg/day [34], in a long-term study, a dose of 6 mg/day was used as the customary dose with adjustment accordingly, and a maximum of 540 mg of iodine/day was administered [4]. In 1970, the Japanese pioneer and leading authority on iodine research, Shigenobu Nagataki (Fig. 1), and his colleagues reported that, when 12 patients with Graves’ disease were treated with 10 mg of iodide t.i.d., clinical improvement was observed in all patients. Although recurrence occurred in half of them after 4–16 weeks of treatment, three of these patients (25%) remained eumetabolic for more than two years, demonstrating the effectiveness of long-term treatment [39]. Although ATDs have been the mainstay of pharmacotherapy for Graves’ disease after the development of ATDs in the 1940s [2], inorganic iodine has been used for more than a century to date as a treatment for Graves’ disease.
Under high-iodine conditions, three different effects can occur: i) inhibition of the organification of iodine (the Wolff-Chaikoff effect); ii) an escape phenomenon; and iii) suppression of TH secretion (the so-called Plummer effect).
i) Wolff-Chaikoff effectIn the 1940s, Wolff and Chaikoff reported that administration of large amounts of iodide in rats resulted in an initial increase of intrathyroidal iodide, followed within a few hours by inhibition of the organification of iodine and TH synthesis [40], later named the Wolff-Chaikoff effect. The Wolff-Chaikoff effect requires a high intracellular iodide concentration [41], and high concentrations of iodine suppress the production of H2O2 [42] and TPO [43] and inhibit iodine organification. Although the mechanism of the Wolff-Chaikoff effect remains to be elucidated, an iodolipid was reported to be involved by acting on NOX activity [44], alteration of DUOX1/2 expression [45], and inhibition of TSH-stimulated genes [45].
ii) Escape phenomenonThe Wolff-Chaikoff inhibitory effect is transient, and after the intracellular iodide concentration decreases, iodine organification resumes, and TH synthesis is restored [46]. This escape phenomenon is the result of adaptation to a high intracellular iodide load. The main known mechanism is downregulation of NIS, resulting in decreased iodide uptake. Iodide uptake by thyrocytes is inversely proportional to the serum concentration. A high iodine concentration inhibits NIS transcription [47], making transcripts more susceptible to degradation [48] and reducing protein levels [49]. Upregulation of pendrin may contribute to the escape process by moving iodide from thyrocytes to the lumen and decreasing the intracellular concentration of iodide [50]. In 1970, Nagataki et al. suggested that chronic iodine administration to rats may decrease iodine recycling in the thyroid gland [51]. Later on, downregulation of DEHAL and type I iodothyronine deiodinase (DIO1) at the plasma membrane was reported [52], which might reduce recycling of iodine from MIT, DIT, and TH, alleviating the high intracellular iodine situation and contributing to escape from the Wolff-Chaikoff effect.
iii) Plummer effectIodine metabolism was studied in patients with Graves’ disease using radioiodine from around the 1950s to the 1970s. Studies by Goldsmith [53] and Wartofsky [54] suggested that the rapid decrease in circulating THs with inorganic iodine was due to a suppression of TH secretion. In 1970, Nagataki et al. reported that iodine administration to patients with Graves’ disease suppressed iodine uptake by the thyroid, but exceeded that of normal controls, and that the total amount of iodine uptake by the thyroid (absolute iodine uptake) increased, because circulating iodine levels increased markedly, suggesting that the decrease in TH was caused mainly by decreased TH secretion [51]. The mechanism for reduced TH secretion by high-dose iodine, the so-called Plummer effect, has been attributed to decreased proteolytic cleavage of thyroglobulin-TH complex to release T4 and T3 [55] and inhibition of TH secretion from thyrocytes. Iodine overload has been reported to decrease mRNA and protein expressions of MCT8 in rats and to reduce secretion of T4 and T3 into the circulation [56].
Recently, Uchida et al. conducted a detailed study of the long-term effects of inorganic iodine in a mouse model of Graves’ disease [57]. In the mouse model of Graves’ disease (the GD-C group), T4 levels in serum and thyroid gland were higher than in the control group, whereas in the sodium iodide-loaded Graves’ disease model mice (GD-NaI group), serum T4 levels decreased to normal levels during the study period, despite higher T4 levels in the thyroid gland than in the GD-C and control groups. In addition, in the GD-NaI group, the reverse T3 level in the thyroid gland was also higher than in the GD-C and control groups. In the GD-NaI group, genes involved in thyroid hormone synthesis and iodine transport, such as SLC5A5 (NIS), SLC26A4 (pendrin), TPO, DUOX2, and DUOXA2, were downregulated, and, therefore, thyroid hormone synthesis was assumed to decrease. In addition, the solute carrier organic anion transporter family member 4A1 (SLCO4A1), also referred to as OATP4A1, which belongs to the SLCO gene family mediating Na+-independent transmembrane transport of various substrates such as THs, was downregulated, which may lead to decreased thyroid hormone secretion. Furthermore, DIO1 was decreased, and DIO3 was upregulated, suggesting altered metabolism of thyroid hormones. These data indicate that long-term treatment with inorganic iodine has the combined effect of reducing both synthesis and secretion of THs, as well as altering thyroid hormone metabolism, which leads to normalization of serum TH levels [57] (Fig. 2).
The gene expression of inorganic iodine administration in a mouse model of Graves’ disease with hyperthyroidism. Relative mRNA expression levels of Slc5a5 (NIS), Slc26a4 (PDS), Tpo, Douxa2, Slco4a1 (OAPT4A1), Dio1, and Dio3 determined with quantitative real-time PCR are shown as bar graphs with dots indicating individual data (n = 4). Data are presented as the means–SE. *p < 0.05 and **p < 0.01. mRNA, messenger RNA; NIS, sodium/iodide symporter; PCR, polymerase chain reaction; PDS, pendrin; Control, control mice; GD-C, mice model of Graves’ disease without treatment; GD-NaI, mice model of Graves’ disease with treatment of sodium iodine.
Inorganic iodine pharmaceuticals used for treatment include LS, saturated potassium iodide solution (SSKI), and potassium iodide (KI) tablets. LS generally contains 5–8 mg of iodide per drop; SSKI generally contains 50 mg of iodide per drop; and KI tablets are available in some countries. In Japan, tablets containing 50 mg KI are available. It should be noted that LS and SSKI are often dispensed within individual medical institutions, and the iodide content may vary depending on the dispensing facility.
Table 1 shows the applications of inorganic iodine for Graves’ disease. Inorganic iodine is used for preoperative preparation and for control before and after radioactive iodine (131I) therapy when ATDs cannot be used due to adverse events. In 1923, Plummer demonstrated the effectiveness of inorganic iodine as a preoperative treatment [1], and since then, administration of inorganic iodine just before surgery has been shown to reduce thyroid blood flow, the vascular bed, and intraoperative bleeding during thyroidectomy [58-60]. The 2016 American Thyroid Association guidelines recommend administering 5 to 7 drops of LS or 1 to 2 drops of SSKI (50 to 100 mg iodide) three times a day for 10 days before surgery (strong recommendation, low quality of evidence) [61]. However, in recent years, there have been reports that there is no difference in blood loss and that thyroid weight increases [62, 63], so there is still debate about its necessity. Iodide is used in combination with ATDs when rapid control of thyroid function is required due to thyroid crisis or comorbidities (atrial fibrillation, heart failure, etc.) and, in Japan, during initial treatment of Graves’ disease. Thiamazole (MMI) has been reported to have dose-dependent adverse effects that include skin disorders [64] and hematological disorders such as agranulocytosis and granulocytopenia [65], which occur more frequently with high doses of MMI. It has been reported that, in severe cases, adding KI to MMI to avoid high-dose use of MMI results in fewer adverse events than high-dose MMI, earlier improved thyroid function, and no difference in the remission rate [66, 67]. The Japan Thyroid Association recommends that, depending on thyroid function, a low dose of 15 mg per day of MMI should initially be administered in mild cases, and in severe cases, 15 mg per day of MMI in combination with 50 mg per day of KI to avoid high doses of MMI [68], which appears to be the practice [69]. Long-term inorganic iodine monotherapy may be used to control Graves’ disease, as described next. Due to the limited number of literature reports, at present, based on the quality of evidence, the administration of KI is not a global recommendation for the initial treatment of Graves’ disease [61].
| (i) | Pre-operative preparation |
| (ii) | Pre- and post- radioactive iodine (131I) therapy |
| When ATD cannot be used due to adverse events | |
| (iii) | Rapid control of thyroid function |
| Thyroid crisis | |
| Comorbidities (e.g., atrial fibrillation and heart failure) | |
| (iv) | Initial treatment (in combination with ATD)* |
| (v) | Long-term inorganic iodine monotherapy* |
*Proposed by the Japan Thyroid Association [68].
In recent years, several studies of long-term inorganic iodine monotherapy have been reported from Japan (Table 2).
| Author/Year | Okamura/2014 [9] | Uchida/2014 [10] | Honda/2017 [11] | Suzuki/2020 [12] | Okamura/2022 [13] |
|---|---|---|---|---|---|
| Type of study | Retrospective | Retrospective (propensity score matching) |
Retrospective | Prospective, observational |
Retrospective |
| No. of subjects | 44 (A:B group⁂ = 29: 15) |
20 (KI = 20 vs. MMI = 20) |
24 | 122 | 504 |
| Untreated/Previously treated | Previously treated GD experienced adverse events of ATD | Untreated, newly diagnosed GD | Previously treated GD experienced adverse events of ATD | Untreated, newly diagnosed GD | Untreated, newly diagnosed GD |
| Age | A:B group⁂ = 41(20–73): 42 (18–69)*** |
48.4 ± 17.6* | 45 (32–56)** | 48 (15–69)*** | <40: 40–65: >65 = 274: 202: 29 |
| Male: Female | 7:37 | 4:16 | 4:20 | 13:109 | 96:408 |
| FT3 (pg/mL) [reference interval] |
N/A | 6.57 ± 1.85* [2.4–4.5] |
5.2 (3.6–10.2)** [2.4–4.5] |
8.55 (3.7–19.8)*** [2.2–4.3] |
<7: 7–10: >10 = 88: 99: 317 [2.2–3.8] |
| FT4 (ng/dL) [reference interval] |
A:B group⁂ = 4.5 ± 3.2 : 5.4 ± 3.0* [0.7–1.7] |
2.49 ± 0.70* [1.00–1.70] |
1.9 (1.4–3.9)** [1.00–1.70] |
2.88 (1.32–4.9)*** [0.8–1.6] |
<4: 4–7: >7 = 212: 186: 106 [0.8–1.7] |
| TRAb [reference interval] |
A:B group⁂ = 31.8 (3.5–85.2)% : 43.2 (8.7–80.8)%*** [<15] |
4.4 (2.8–7.9) IU/L** [<1.0] |
9.5 (6.1–15.6) IU/L** [<2.0] |
4.65 (0.3–40.0)IU/L*** [<2.0] |
N/A |
| Thyroid size | A:B group ⁂ = 29 (8–104 ): 25 (15–80)g*** | 25.29 ± 7.75 g*♯ | 29.2 (22.4–36.5) g**♯ | 22.1 (9.7–85.5) mL† | <20: 20–40: >40 g = 164: 257: 83‡ |
| Initial daily KI dose (mg) [actual dose] | 13–100 [<50: n = 15, 50: n = 10, 100: n = 19] |
50 – 100 [21.4: n = 1, 50: n = 18, 100: n = 1 ] |
50 or 100 [50: n = 16, 100: n = 8] |
50 | 100 |
| Follow up period | Median 17.5 years (8.6–28.4) | 12 months | 6 months | Median 27.6 months (1.6–46.7) | 2.5–19.9 years |
| Response rate (%) | 65.9 | 85 | N/A | 58.2 | 83.7 |
| Remission rate (%) | 38.6 | N/A | N/A | N/A | about 40¶ |
| Factors expected to respond | Lower maximum dose of KI (<200 mg/day) |
Mild hyperthyroidism | Mild hyperthyroidism | Female Mild hyperthyroidism | Older generation Mild hyperthyroidism Small goiter |
| Conclusion | Effective in 2/3 of the subjects, and responded well in 41%. | Effective for 85% of the patients with mild GD. KI could be an alternative option to ATD. | Efficacy of KI was inversely correlated with the severity of hyperthyroidism. | KI is effective and potentially safer for 60% of female patients with mild GD. | The susceptible group were a high rate of 83.7%. MMI and RAIT was effective in cases of insufficient response. |
KI, potassium iodide; MMI, thiamazole; GD, Graves’ disease; ATD, anti-thyroid drug; RAIT, radioiodine therapy; TRAb, TSH receptor antibody; N/A, Not Available; *, mean ± standard deviation; **, median (IQR); ***, median(range); ⁂, A group included well controlled patients with KI alone and B group included not controlled patients with KI alone; ♯, Thyroid volume (cm3) was calculated by ultrasonography using the ellipsoid model: width × length × thickness × π/6 for each lobe + width × length × thickness for isthmus; †, Thyroid volume (cm3) was estimated by ultrasonography according to the following formula: 0.7365 × right lobe length × width × depth + 0.7412 × left lobe length × width × depth – 0.55 ; ‡, Thyroid volume was estimated by palpation and ultrasonography using the following formula: 0.7 × the maximum width × the maximum thickness × the maximum length for each; ¶, Calculated as approximate values from the figure.
In 2014, Okamura et al. reported that the effects of inorganic iodine monotherapy may be sustained over the long term [9]. This was a retrospective study of 44 patients with Graves’ disease who switched to iodine due to adverse events with ATDs, and their long-term course was followed over a period of 8 to 28 years. Twenty-nine patients (66%) were well controlled with KI (as shown in Table 2, these cases were classified as group A and the rest as group B) (Table 2), of which 58.6% (38.6% overall) achieved remission after an average of 7.4 years (2–23 years) of treatment. Of interest, in these remission cases, TSH receptor antibody (TRAb) levels (reference interval <15%) were significantly reduced and completely normalized from a median of 43.2% (range 6.9–76.4%) to 2.3% (0.1–7.7%). The maximum dose of KI was 800 mg, which is quite high. Initial parameters including free T4 concentration and goiter size were not indicators of responsiveness to KI or long-term prognosis. When a maximum KI dose of 200 mg/day or more is required, the remission rate is 35%, and when it is less than 200 mg/day, the remission rate is 70.8%.
Uchida et al. reported a retrospective study in which cases of KI and MMI monotherapy for newly diagnosed untreated Graves’ disease were identified, and 20 cases of each were compared using propensity score matching [10]. The starting KI dose was 53.6 ± 11.7 mg/day, with most patients treated with 50 mg/day. KI treatment alone was successful in 17 patients (excluding 2 escape cases and 1 refractory case), and MMI treatment was successful in 20 cases, with normal thyroid function after 6 months and 1 year. In cases with KI alone, TRAb levels (reference interval <1.0 IU/L) decreased similarly to MMI [KI: median 5.7 (interquartile range 2.7–8.1) IU/L before treatment vs. 2.4 (1.6–3.8) IU/L after 1 year of treatment, p < 0.05, MMI: 4.4 (2.8–7.9) IU/L before treatment vs. 2.3 (1.5–3.0) IU/L after 1 year of treatment, p < 0.05]. The characteristic background of KI-ineffective cases was not clear. The pre-treatment FT4 level was 2.49 ± 0.70 ng/dL, corresponding to relatively mild hyperthyroidism, which may have resulted in the high response rate.
Honda et al. conducted a retrospective study of 24 patients who had adverse events with ATDs and were switched to KI, dividing the patients into groups based on the presence or absence of aggravation within 180 days after switching [11]. The aggravation group had higher TH levels at switchover than the non-aggravation group (FT3, 9.3 [interquartile range, 5.2–11.6] vs. 3.7 [interquartile range 3.3–4.8] pg/mL, p = 0.02 and FT4, 3.6 [interquartile range 1.8–4.5] vs. 1.4 [interquartile range 1.2–1.9] ng/dL, p = 0.02). In the aggravation group, the KI effect was inversely correlated with FT3 and FT4 values at switchover. There was an increase in TRAb levels during the treatment period in 9 of the 24 patients (37.5%). Seven of 13 patients in the aggravation group had an increase from before the start of KI, but only two of 11 patients in the non-aggravation group had a minor increase, suggesting that an obvious increase may be a factor that should be considered for definitive therapy.
Suzuki et al. conducted a prospective, observational study of 122 patients with newly diagnosed, untreated Graves’ disease, with mild to moderate FT4 levels defined as less than 5.0 ng/dL [12]. Administration of KI was basically started at 50 mg/day, and the KI dosage was adjusted as shown in the flowchart in Fig. 3. If the FT4 value did not decrease to within the reference range even after increasing the dose of KI to 100 mg, it was considered a non-response. Of the 122 patients, 71 (58.2%) patients responded successfully, and 36 (29.5%) were able to stop KI, and it was possible to induce remission. However, 51 were non-responders. The median duration for non-responders to be judged non-responsive was 5.9 months. The factors associated with non-response to inorganic iodine treatment were FT4 (odds ratio (OR) 2.19, 95% confidence interval (CI) 1.28–3.75; p = 0.0007) and male sex (OR 3.58, 95%CI 1.04–12.3; p = 0.04). The cutoff value for FT4 calculated using the receiver-operating characteristic curve was 2.76 ng/dL (=35.52 pmol/L) (Fig. 4). The median TRAb levels (reference interval <2.0 IU/L) of non-responders (n = 51) were 9.1 IU/L (range 0.3–40.0) IU/L when judged non-responders after treatment initiation, significantly higher than the 4.4 (0.3–30.3) IU/L before treatment (p < 0.0001). An increasing trend in TRAb levels may also indicate the need for changing treatment. A total of 91 cases in which thyroid volume was re-evaluated one year later showed that there were cases of enlargement, cases of non-enlargement, and cases of shrinkage. The median thyroid volume change was +2.6 mL in total subjects, +1.94 mL in KI responders, and +7.32 mL in KI non-responders.
Initial dose was set at 50 mg/day for all subjects. KI dose was increased if FT4 values did not decrease to the upper limit of normal range. KI dose was decreased if TSH values elevated to be within normal range. Otherwise, KI dose was continued at the same dose.
Receiver-operating characteristic (ROC) curve analysis of the relationship between FT4 value and KI responsiveness indicated an optimal cut-off FT4 value of 2.76 ng/dL (area under the ROC curve (AUC) = 0.684).
Okamura et al. reported the efficacy of KI with a starting dose of 100 mg/day in 504 patients with untreated Graves’ disease in 2022 [13]. The initial response to KI within 6 months was evaluated, and the patients were subclassified based on whether TSH and FT4 normalized (Table 3) in the KI-sensitive group (group A) with improved TSH suppression (A1: decreased group, A2: normal function group including TSH), and the KI-sensitive group with sustained TSH suppression (group B) (B1:FT4 decreased due to TH level, B2: FT4 normal, B3: FT4 normal but high FT3 group), and this group was defined as the susceptible group. These groups A and B were classified as the susceptible group, and cases in which FT4 did not reach normal values were classified as the resistant group (Group C). In this study, the starting dose was high, at 100 mg/day of KI, and a unique method of using LT4 in combination with KI during periods of high TSH levels was adopted. The susceptible group (group A and group B) had a high rate of 83.7% (422 cases) (Table 3). In group A2, in which FT4 and TSH normalized initially, there was no initial exacerbation within 6 months (Fig. 5A), and the remission rate in group A was about 60% (Fig. 5B). If escape phenomena were observed after normalization of FT4, or if the effect of KI was insufficient, concomitant use of MMI was initiated. As a result, improvement in thyroid function was observed (Fig. 5A). The subsequent remission rate was about 35% in group B and 30% in group C (Fig. 5B). When RAIT was selected, there were no clinical problems with its efficacy. Higher KI sensitivity was seen in elderly patients, those with a small goiter, and those with mild disease.
| Group | A1 | A2 | B1 | B2 | B3 | C |
|---|---|---|---|---|---|---|
| Serum free T4 | low | normal | low | normal | normal (high fT3) | remained high |
| KI sensitivity | too sensitive | sensitive | sensitive | sensitive | sensitive | resistant |
| Serum TSH | high | normal | suppressed | suppressed | suppressed | suppressed |
| n (%/Total) | 92 (18.3%) | 78 (15.5%) | 27 (5.4%) | 142 (28.2%) | 83 (16.5%) | 82 (16.3 %) |
Reproduced with permission from Okamura et al. (2022) Endocr J 69: 983–997 [13].
KI sensitivity was evaluated depending on the changes in serum fT4 and TSH levels during first 180 days. Group A: KI sensitive with recovered TSH response, hypothyroid (A1) and euthyroid (A2). Group B: KI sensitive but suppressed TSH, low fT4 and inappropriately suppressed serum TSH (B1), normal fT4 and fT3 even temporarily (B2), normal fT4 and high fT3 (B3-T3 toxicosis). Group C: KI resistant and suppressed TSH.
(B) The prognosis of the patients with Graves’ hyperthyroidism initially treated with 100 mg potassium iodide (KI) and followed for 2–23 years (n = 429), depending on the early response during 180 days. If patients wished to continue a small maintenance dose of KI (50 mg/day) even when TRAb was negative with a small goiter, they were defined as the possible remission group. See the legends for Figs. 1 and 2. The number in parenthesis is the number of patients in each group. Remission (including possible remission) and spontaneous hypothyroidism were significantly more frequent in Group A (74.3% and 11.1%, respectively), than in Group B (46.3% and 2.8%) or Group C (53.6% and 1.5%) (p < 0.0001). The prognosis was not markedly different between Groups B and C (p = 0.5117). The time required for remission was 1,025 (676–1,532) days in Group A (n = 84) and 1,847 (1,276–3,147) days in Groups B & C (n = 99). The difference was significant (p < 0.0001).
Except for Honda’s report in which no information on side effects of KI was available [11], in the remaining four reports, no serious side effects of KI were found, with only one patient experiencing a skin rash 4 days after the administration of KI in the 2022 report by Okamura et al. [13]. In their report in 2013, Okamura et al. also described a transient increase in TRAb and hypothyroidism in several patients as ‘side effects’ [9]. However, these may not be considered side effects because transient changes in TRAb levels are often experienced in the clinical course of Graves’ disease, and hypothyroidism is a likely pharmacological effect of KI.
To summarize, there are no clear standards regarding the initial dosage, but the starting daily dose for KI was around 50 mg. Factors related to expected response included mild disease (e.g., FT4 <2.76 ng/dL = 35.52 pmol/L), female sex, small goiter, and elderly patient. The response rate was relatively good at 60% to 80%, but the remission rate was about 40%. It is necessary to properly evaluate the effect during the course of treatment, at around six months after starting treatment. In cases of insufficient response, changing treatment (RAIT, surgery, and if possible, ATD) should be considered. An increasing trend in TRAb levels may also indicate the need for changing treatment. In rats on a low-iodine diet, the iodine content of the thyroid gland was reported to be 0.5 μg [70], and the dose required for the Wolff-Chaikoff effect was 5–20 μg [71], suggesting that excessive iodine intake, far exceeding the original thyroid gland content, is required to induce the effect. The iodine content of the human thyroid gland is estimated to be about 10 mg [72]. Okamura et al. stated that, based on the above evidence, the initial dose of KI was set high at 100 mg/day [13]. In contrast, Thompson et al., based on the idea that the minimum dose should be used, vigorously searched for the minimum effective dose and reported that a small amount of iodine of 6 mg or less per day was effective [35]. Although the number of iodine-deficient areas is decreasing, recent reports of long-term iodine monotherapy come from iodine-sufficient areas of Japan. It should be noted that the conclusions may only apply to areas where iodine intake is currently sufficient, and that, as diets are diversified, iodine sufficiency can vary among regions and individuals within a country, and that there can be racial and genetic differences in iodine sensitivity. The initial and maximum doses of iodine in long-term inorganic iodine monotherapy are empirical and require further study. In addition, it is expected that the mechanism by which the effect of inorganic iodine administration persists in some cases will be elucidated in the future.
Although the exact mechanism must be further investigated, of interest, TRAb values with KI monotherapy were reduced [9, 10]. Since iodine can induce autoimmune changes in patients, further data on the potential risks are needed [73]. Few reports have been published on the treatment of KI during pregnancy; Momotani et al. reported that, in 35 pregnant women with Graves’ disease treated with iodine monotherapy (6–40 mg/day) up to delivery, the thyroid function of their born infant was almost within the reference range [74]. Switching to inorganic iodine treatment in early pregnancy is an option with the aim of avoiding the teratogenic effects of MMI during the early pregnancy period [75]. Significant reductions in major anomalies (from 4.14% to 1.53%) and MMI embryopathy (from 1.6% to 0.8%) have been reported when MMI exposure is avoided by this method. However, we have to pay attention, because such a switch to inorganic iodine in early pregnancy often results in worsening of thyroid function [76]. The thyroid gland completes organogenesis and differentiation at approximately 12 weeks of gestation, when NIS expression and fetal TH production begin. Since it has been reported that when pregnant women take excessive amounts of iodine, the fetus is prone to hypothyroidism [77], mothers should be treated with the minimum amount of iodide for the minimum period necessary. NIS is expressed in various organs, particularly in the mammary glands during lactation, and the concentration of iodine in breast milk is 20–50 times higher than in the circulation, so caution is needed when breastfeeding mothers take iodine. Several cases of neonatal hypothyroidism secondary to maternal seaweed consumption have been reported [78]. Hamada et al. reported thyroid function in infants of breastfeeding mothers taking a median of 50 mg/day (range, 10–100 mg/day) of iodine; 25 of 26 infants had normal thyroid function, but one infant had high TSH (12.3 μIU/mL) in the first month after breastfeeding started [79]. A subsequent report of 100 cases found subclinical hypothyroidism in approximately 10% of the infants [80]. Elevated TSH levels normalized in all children during or after cessation of iodine exposure [79, 80]. Therefore, hypothyroidism may occur in infants of breastfeeding mothers taking iodine, and though it is acceptable if there is a compelling reason, it should be avoided if at all possible, and, if taken, the infant’s thyroid function should be tested. Reports of iodine therapy during pregnancy and during breastfeeding are scarce, and more evidence is needed. Finally, LS and other iodide preparations appear to have a low frequency of adverse reactions, but more evidence is also required.
The suppressive effect of inorganic iodine on thyroid function persists in some cases with Graves’ disease. Candidates for long-term inorganic iodine monotherapy are mild cases (FT4 <2.76 ng/dL = 35.52 pmol/L) with a small thyroid gland, elderly persons, and women. Inorganic iodine can be used as an alternative in patients who experience adverse events with conventional ATDs and/or prefer conservative treatment. Doses of inorganic iodine monotherapy are empirical and require further study.
This narrative review was written for the upcoming 100th anniversary of the Japan Endocrine Society. The author would like to thank the late Dr. Shigenobu Nagataki for his pioneering research and encouragement to us all. The author would also like to profoundly thank Dr. Nami Suzuki for her detailed research and help in preparing this paper.
Author disclosure statementThe author is a member of Endocrine Journal’s Editorial Board.
