Novel etiopathophysiological aspects of thyrotoxic periodic paralysis

REVIEWS Novel etiopathophysiological aspects of thyrotoxic periodic paralysis Rui M. B. Maciel, Susan C. Lindsey and Magnus R. Dias da Silva Abstract | Thyrotoxicosis can lead to thyrotoxic periodic paralysis (TPP), an endocrine channelopathy, and is the most common cause of acquired periodic paralysis. Typically, paralytic attacks cease when hyperthyroidism is abolished, and recur if hyperthyroidism returns. TPP is often underdiagnosed, as it has diverse periodicity, duration and intensity. The age at which patients develop TPP closely follows the age at which thyrotoxicosis occurs. All ethnicities can be affected, but TPP is most prevalent in people of Asian and, secondly, Latin American descent. TPP is characterized by hypokalemia, suppressed TSH levels and increased levels of thyroid hormones. Nonselective β adrenergic blockers, such as propranolol, are an efficient adjuvant to antithyroid drugs to prevent paralysis; however, an early and definitive treatment should always be pursued. Evidence indicates that TPP results from the combination of genetic susceptibility, thyrotoxicosis and environmental factors (such as a highcarbohydrate diet). We believe that excess T3 modifies the insulin sensitivity of skeletal muscle and pancreatic β cells and thus alters potassium homeostasis, but only leads to a depolarization-induced acute loss of muscle excitability in patients with inherited ion channel mutations. An integrated etiopathophysiological model is proposed based on molecular findings and knowledge gained from long-term follow-up of patients with TPP. Maciel, R. M. B. et al. Nat. Rev. Endocrinol. advance online publication 10 May 2011; doi:10.1038/nrendo.2011.58

Introduction Thyrotoxic periodic paralysis (TPP) is an acute condition that manifests as a sudden and reversible loss of muscle strength of the limbs and thyrotoxicosis associated with hypokalemia.1 TPP is the most frequent form of acquired acute flaccid paralysis in adults and is more prevalent in Asian populations2 than in other populations, although it can occur in any ethnicity.3–6 In this Review, we address important aspects of TPP, including its etiology, differen‑ tial diagnosis, clinical manifestations and management in 33 patients over a 10‑year follow-up period. We describe our experiences in treating >100 paralytic attacks. We also summarize research conducted in our laboratory in which we searched for susceptibility genes, which added to our understanding of the multifactorial patho­ physiological mechanism that underlies TPP. This article provides an overview of TPP as an endocrine channelo‑ pathy and emphasizes that it should be systematically included in the differential diagnosis of acute muscle weaknesses in young males independent of their ethnic background. Finally, we propose early, safe and effective treatment for patients with TPP.

TPP as an endocrine channelopathy TPP attacks occur with variable frequency, duration and intensity; therefore, a diagnosis of TPP is often only established long after the initial attack.3,7 The first report of an association between paralysis of the limbs and thyrotoxicosis in a 19-year-old man was Competing interests The authors declare no competing interests.

published in 1902 by Rosenfeld.8 The patient had been having typical symptoms of Graves disease for 1 year and presented with sudden and recurrent paralysis with absent patellar reflex. In 1931, Dunlap and Kepler described four patients who were treated at the Mayo Clinic, USA, for attacks of flaccid paralysis that were clinically identical to those of familial periodic para­lysis with associated exo­phthalmic goiter.9 The attacks did not occur once the exo­phthalmic goiter was controlled by clinical or surgical measures. Subsequently, the majority of case descriptions and patient series came from Asia.2 However, in 1968, Pereira et al.10 published the first case of TPP diagnosed in Brazil, a report that was accompanied by a study of the involvement of the glucose to insulin ratio in trigger­ing attacks. In 1983, Sterian and Maciel11 reported the first case of TPP in our clinic in a white patient, which increased our interest in documenting upcoming cases. A decade later, Ober 12 reported on seven American patients from dif­ferent ethnic backgrounds who had TPP. Although TPP is clinically similar (with the exception of thyrotoxicosis) to familial hypokalemic periodic paralysis,13 a disease that is caused by mutations in sodium and calcium channels, TPP was not reported as a potential channelopathy until 2002,14 a speculation that was support­ed by findings published in 2010.15

Retracing the epidemiology of TPP TPP is reported more often in Asian people than in those of other ethnicities.2,16–21 Western health-care pro­viders, therefore, have a reduced awareness of TPP, which can lead to its underdiagnosis in Western countries.22,23 How­ever,

NATURE REVIEWS | ENDOCRINOLOGY

Laboratory of Molecular and Translational Endocrinology, Department of Medicine (R. M. B. Maciel, S. C. Lindsey), Department of Biochemistry (M. R. Dias da Silva), Escola Paulista de Medicina, Universidade Federal de São Paulo, Rua Pedro de Toledo 669, 11º Andar, São Paulo 04039‑032, SP, Brazil. Correspondence to: M. R. Dias da Silva [email protected]

ADVANCE ONLINE PUBLICATION  |  1 © 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS Key points ■■ Thyrotoxic periodic paralysis (TPP) is characterized by attacks of acute muscle weakness, hypokalemia and thyrotoxic symptoms; it results from the combination of genetic susceptibility, thyrotoxicosis and environmental factors ■■ People of Asian descent are most often affected, but all ethnicities can present with TPP ■■ Hypokalemia is typical during paralysis, but is not always detected; administration of potassium during the attacks should be offered cautiously, preferably orally, to prevent rebound hyperkalemia ■■ Most patients are first diagnosed with paralysis and later with thyrotoxicosis; delay in the diagnosis is frequent and, therefore, all cases of acute weakness in young males should raise suspicion ■■ Early and definitive treatment of thyrotoxicosis is necessary to prevent new attacks of muscle weakness ■■ Mutations in Kir2.6, a potassium ion channel, which occur in up to 33% of patients, reinforces the hypothesis of genetic heterogeneity in patients with TPP

there have been reports of cases of TPP in many ethnic groups, including Native American (from North, Central and South America), Arabic, black, white, mixed black and white, and Hispanic people.4,20,24–37 McFadzean and Yeung studied 1,366 Chinese patients with thyrotoxicosis and found that 1.8% had signs of paralysis.38 At the Mayo Clinic, USA, the observed incidence of TPP was approxi‑ mately one-tenth of the reported rate for the Chinese population, ranging from 0.1% to 0.2%.5 Of all cases diagnosed in the USA, the estimated ethnic distribution was 45% white, 24% Asian, 15.5% Hispanic, 7% Native American, 7% African American and 1% other.12 In Japan, a decrease in the incidence of TPP was reported in a study that compared data from 1957 and 1991; the incidence dropped from 8.2% to 4.3% in men and from 0.40% to 0.04% in women.39 This change could be connected with a reduced salt and carbohydrate intake and an increased potassium intake in Japan during this period.39 In contrast to the finding that TPP is rare among African Americans residing in the USA,5,12 we identified TPP in 13 African American and mixed black and white people (39.4%) in our cohort (M. R. Dias da Silva, unpublished work). This finding is important as it demonstrates that TPP should not be excluded from the differential diagnosis of paraly‑ sis solely because of the ethnicity of the patient.22,40 On the basis of the greater prevalence of TPP in Asian and Latin American people,1,41 it is reasonable to hypothesize that there is a genetic predisposition to the development of TPP based on the common ancestry of these ethnici‑ ties, which might be connected by a migrating founder population that travelled from Asia to America through the Bering Strait.22 TPP occurs more frequently in men than women. In a review of 1,366 Chinese patients with thyrotoxicosis, 13% of males and 0.17% of females reported paralytic attacks.38 Similarly, among 45 Chinese patients evaluated for TPP from 1984 to 1993, all but one were male.17 Our study confirms the male preponderance of this condition (M. R. Dias da Silva, unpublished work), with a male to female ratio of ~30:1; in fact, only one female in our cohort was diagnosed with TPP (Box 1).

A few studies have described TPP in pregnant women;42–44 one case occurred following a p ­ rostaglandin-induced abortion during the second trimester.44 In addition, one case of a pregnant woman with TPP and impaired glucose tolerance has been reported,43 and although it is possible that the altered glucose metabolism was related to thyro‑ toxicosis, it is important to note the possible association between insulin resistance and TPP. In contrast to familial hypokalemic periodic paralysis, nearly all cases of TPP present as sporadic disorders.45 How­ever, TPP has been described in patients with a family history of paralysis, albeit rarely.7,38,46 The onset of TPP symptoms begins in young adulthood, unlike attacks in the familial form, which begin earlier, usually before 16 years of age. In addition, the age of onset of TPP usually coincides with the peak of the incidence of thyrotoxicosis, which is between the third and fifth decades of life. In our cohort, the age of diag­nosis varied from 19 to 51 years (M. R. Dias da Silva, unpublished work), similar to the ages of onset described in the litera‑ ture.13,47 One case of a 14‑year‑old Chinese boy with TPP has been reported.48

Any thyrotoxicosis triggers paralysis Patients with TPP develop attacks only while they are thyro­toxic, regardless of the variable etiology of the thyro­ toxicosis. These etiologies include Graves disease, toxic adenoma,49 Jod-Basedow,50 toxic multinodular goiter, amiodarone-induced thyrotoxicosis,51 TSH-producing pituitary tumor,52,53 lymphocytic thyroiditis1 and factitious thyrotoxicosis.1,54–56 In fact, any cause of thyro­toxicosis, including overzealous levothyroxine replacement therapy, can trigger paralysis in susceptible patients. Nevertheless, Graves disease is the most common cause of thyro­ toxicosis,1,12,13 which we also observed in our cohort (31/33 patients; M. R. Dias da Silva, unpublished work). In our series, exogenous thyrotoxicosis was observed in one patient who had been taking drugs for weight loss that contained thyroid hormone; the remaining patient had thyrotoxicosis attributable to acute thyroiditis.1

Clinical presentation of TPP As TPP is the most frequent form of acquired hypo­ kalemic paralysis that presents in emergency rooms world­wide, we would like to reinforce the importance of identifying TPP early as a classic triad of flaccid para­lysis, signs of thyro­toxicosis and hypokalemia during the para‑ lytic crisis. In 2001, Lin et al.57 reviewed the medical charts of 97 patients with severe hypokalemia and profound weakness who were initially diagnosed with hypokalemic period­ic paralysis in an emergency room. The initial diag‑ nosis was confirmed in 73 of these patients, of whom 39 were diagnosed as having TPP.57 In patients with TPP, paralysis attacks occur with vari‑ able frequency and duration, usually during the night. Acute muscle weakness is typically symmetrical but occasionally presents as asymmetrical,38 focal or general­ ized; the weakness usually affects the proximal muscles of the lower limbs or any recently exercised muscles. Some patients experience tetraparesis or tetraplegia; of our

2  |  ADVANCE ONLINE PUBLICATION

www.nature.com/nrendo © 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS patients, 20 presented predominantly with tetra­paresis or tetraplegia. Interestingly, some of the patients reported a prodromal symptom—a sensation that arose in the affected muscles that was described as aching, stiffness, pain or cramping.1,38 As frequently reported and observed in our cohort, the attacks start while the patient is in bed during the night or early in the morning following a day of vigorous exercise and/or consumption of a large amount of food. Patients who are aware of their prodromal symp‑ toms from their own previous experiences usually report that they can prevent these crises by increasing physical activity. Usually, TPP attacks do not affect bulbar or respi‑ ratory muscle function, and no sensory symptoms have been reported. Most patients present with attacks that last several hours. In our study group, most patients reported attacks that lasted, on average, 2–6 h.1 One of the most remarkable features arising from a physical examination during an attack is a depressed deeptendon reflex in a patient with symptoms of thyro­toxicosis, in whom we would expect an exacerbated reflex response. In 16 of 19 patients tested, the reflexes were absent or decreased.1 By contrast, an increased tendon reflex has been reported by others.51 Cardiac arrhythmias are infrequent, but there have been reports of ventricular tachycardia and ventricular fibrilla‑ tion in patients with TPP.58–61 Respiratory involvement is also uncommon; however, acute respiratory hypercapnic failure has been reported, including in one of our patients, who needed mechanical ventilation until potassium levels were restored.1,61 Triggering factors of TPP include carbohydrate-rich meals, rest after intense physical activity, emotional stress, the presence of fever or infection, trauma, smoking and menstruation. These factors are more often reported by patients with familial hypokalemic periodic paralysis than by those with TPP, perhaps because fami­lial hypokalemic periodic paralysis has a chronic course, which enables more information to be obtained from the patients. Many of our patients reported precipitating factors, such as meals with a high carbohydrate load and heavy meals during Christmas and Easter festivities. Two patients pre‑ sented symptoms after consuming pizza and soft drinks.1 In addition, a young man, who was later found to have Graves disease, had episodes of tetra­plegia that occurred exclusively during weekends, when he consumed several liters of carbohydrate-rich soft drinks while working at his computer.62 Interestingly, two patients reported sudden paraplegia during a rest period after strenuous exercise playing football; the attacks occurred when they were sitting down and removing their shoes. Furthermore, there have been other reports of TPP attacks that were induced by steroid treatment, which caused hyper­ insulinemia and hyperglycemia,60,63,64 especially after taking high doses of methylprednisolone.60,65 Patients usually do not recognize their thyrotoxic symp‑ toms and are first diagnosed with periodic paralysis,66 which occurred in at least 72% of our patients.3 However, upon questioning and careful physical examination, all but one were found to have symptoms and/or signs of thyrotoxicosis or Graves disease, including weight loss,

Box 1 | Clinical and laboratory results in patients with TPP* Ethnicity (n) ■■ White: 13 ■■ Mixed black and white: 11 ■■ Asian: 5 ■■ Amerindian: 2 ■■ African American: 2 Age at diagnosis (years) ■■ Mean: 28.15 ■■ Range: 19–51 Delay until diagnosis ■■ First attack (n): 7 ■■ Average (months): 13.88 ■■ Range (months): 0.25–144 Pattern of weakness (n) ■■ Paraplegia: 7 ■■ Tetraplegia: 11 ■■ Tetraparesis: 9 ■■ Paraparesis: 6 Deep tendon reflexes (n) ■■ Absent: 8 ■■ Hypoactive: 8 ■■ Hyperactive: 1 ■■ Normal: 2 ■■ Not available: 14 Duration of attacks (h) ■■ Mean: 6.37 ■■ Range: 2–24 Time of attacks (n) ■■ Day: 4 ■■ Night: 29 Precipitating factors (n) ■■ Rest after intense physical activity: 8 ■■ After meals: 6 Potassium levels (normal = 3.5–5.0 mmol/l) ■■ Mean (mmol/l): 2.28 ■■ Range (mmol/l): 1.2–3.4 Etiology of thyrotoxicosis (n) ■■ Graves disease: 31 ■■ Thyroid hormone abuse: 1 ■■ Thyroiditis: 1 Signs and symptoms of thyrotoxicosis (n)‡ ■■ None: 1 ■■ Mild (1–3): 11 ■■ Overt (4–6): 21 First diagnosed as (n) ■■ Paralysis: 24 ■■ Thyrotoxicosis: 7 ■■ Unknown: 2 TPP follow-up (years) ■■ Mean: 10.78 ■■ Range: 0.3–29 *Data from 33 patients from the authors’ cohort. ‡Signs and symptoms of thyrotoxicosis: goiter, weight loss, tachycardia or palpitations, tremors, ophthalmopathy, sweating. Abbreviation: TPP, thyrotoxic periodic paralysis.

NATURE REVIEWS | ENDOCRINOLOGY

ADVANCE ONLINE PUBLICATION  |  3 © 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS Table 1 | Differential diagnosis of acquired muscle paralysis in young adults Factor

Type of muscle paralysis TPP

Familial hypokalemic periodic paralysis

Guillain–Barré syndrome

Thyrotoxicosis

Yes, often oligosymptomatic

No

No

Age of onset (years)

20–45 (95%)

Before 16 (80%)

All, with incidence increased in ≥50

Male to female ratio

30:1

3:1

1.5:1.0

Ethnicity most frequently affected

Asian

White

Any

Family history of paralysis

No

Yes

No

Family history of thyroid disease

Frequent

No

No

Precipitating factors

High carbohydrate diet and/or salt ingestion, rest after exercise and stress

High carbohydrate diet and/or salt ingestion, rest after exercise and dehydration

Precedent infection in 70% of patients

Dysautonomia

No

No

In 70% of patients

Deep tendon reflexes

Usually absent or depressed

Usually absent or depressed

Absent or depressed

Severe respiratory muscle weakness

Very rare

Rare

10–30% of patients

Facial weakness

No

No

≥50%

Duration of the attacks

30 min–6 h

≥24 h

Progressive over days to 4 weeks

Potassium levels during the attack (mmol/l)

1.5–3.0

2.8–3.5

Normal

Cerebral spinal fluid

Normal

Normal

Albuminocytologic dissociation

Nerve conduction analysis

Not specific, not necessary

Not specific

Useful and helpful for diagnosis

Main treatment

For thyrotoxicosis

Acetazolamide, dichlorphenamide, spironolactone, potassium chloride

Immunotherapy

Clinical course

Remission after thyrotoxicosis is corrected

Chronic myopathy

Recovery; residual functional deficit in up to 20%; death in some patients

Genetic inheritance

Mutations in skeletal muscle potassium Kir2.6 channel in up to 33% of patients

Mutation in skeletal muscle calcium and sodium channels in 80% (Cav1.1) and 15% (Nav1.4)

None

Abbreviation: TPP, thyrotoxic periodic paralysis. Modified with permission from Silva, M. R. et al. Arq. Bras. Endocrinol. Metabol. 48, 196–215 (2004).

palpitations, tremors, goiter and exophthalmos. Thus, a careful history and physical examination should be carried out to obtain a correct diagnosis. In a study by Ko et al.,17 28.9% of patients with TPP had a history of thyrotoxicosis before their first presentation with periodic paralysis, and 60% had clinical evidence of thyrotoxicosis. However, the diagnosis of periodic paraly­ sis is occasionally not established at the initial hospital visit, partly because the patients sometimes arrive when the weakness has reversed, and many are thought to have conversion disorders or anxiety and are given tran­ quilizers. For instance, seven of our patients were given drugs such as benzodiazepines.1 Apart from the thyrotoxic symptoms, the clinical fea‑ tures of TPP are similar to those observed in patients with familial hypokalemic periodic paralysis. However, some aspects might help to distinguish between the two forms of paralysis (Table 1). The clinical course following diag‑ nosis and treatment differs between the two conditions, as patients with TPP show no additional attacks after the thyro­toxic state has been resolved, while those with

familial hypokalemic periodic paralysis often progress to chronic myopathy, and require continuous pharmaco­ logical treatment and potassium supplementation. Other differential diagnoses of acute muscle weakness include Guillain–Barré syndrome, myasthenia gravis, acute myelo­ pathy (such as transverse myelitis), acute intermittent p­orphyria and botulism.

Laboratory hallmarks of TPP The most remarkable alterations in laboratory tests seen in patients with TPP are hypokalemia associated with suppressed levels of TSH and increased levels of thyroid hormones (total and free T4 and T3). Hypokalemia does not indicate a depletion of the total potassium pool but an increased influx of the ion to the intracellular compartment.67–69 Potassium levels are usually lower in patients with TPP than in those with familial hypokalemic periodic paralysis and are com‑ parable to the levels in patients with tumoral excess of mineralo­corticoid.1 A finding of normal or even increased levels of potassium is not uncommon if a blood sample

4  |  ADVANCE ONLINE PUBLICATION

www.nature.com/nrendo © 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS is drawn after the attack has passed or if the sample is handled inadequately. When collected after the attack, a change in the levels of potassium is usually caused by a sponta­neous recovery from the potassium shift, which is typically observed in patients with periodic muscle dis‑ orders and frequently leads to a rebound phe­nomenon;13 this alteration might contribute to a diag­nostic delay. Impor­tantly, potassium normalization invariably precede­s muscle contraction recovery.70,71 Some individuals also present with low serum levels of phosphorus during the attacks. However, these patients do not need replacement therapy because the levels increase after the attack, regardless of whether phosphorus replace‑ ment is used.72,73 In patients with TPP, calcium levels are usually within the normal range; when they are altered slightly, there is no correlation with the duration or inten‑ sity of the crisis. Patients with TPP might present with a higher calcium to creatinine ratio and a lower phosphate to creatinine ratio in spot urine samples than patients with nonthyrotoxic hypokalemic paralysis.74 As the clinical findings are variable among patients with thyrotoxicosis, measurements of TSH and free T4 are man‑ datory. All patients in our study had suppressed levels of TSH and high levels of thyroid hormones and most of them were positive for antithyroid antibodies attributed to Graves disease,1 as has been widely observed in the literature.75–77 Other tests, such as thyroid ultrasonography, thyroid scintigraphy and radioiodine uptake, might be necessary to better evaluate hyperthyroidism with regard to its etiol‑ ogy and treatment plan. Results from an electromyography in patients with TPP are atypical during attacks and normal after remission. In fact, when eight Chinese patients with TPP were tested most showed myopathic patterns during the paralytic attacks.78 The myopathic changes noted were a decrease in the duration of muscle action potentials, an increase in polyphasic potentials, a satisfactory interference pattern with reduced amplitude and a reduced amplitude of the evoked muscle action potential upon nerve stimulation. Peripheral nerve function was normal in these patients, which demonstrates that the weakness in patients with TPP is largely myopathic.78 Six of our patients underwent electromyographic tests; two had no specific myopathic changes, and four had normal results.1 Histological muscle findings in patients with TPP are also not specific and can be seen in any form of periodic paralysis. Five patients in our study had muscle biopsy samples taken; three showed vacuoles and tubular aggre‑ gates, and two were normal.79 An optical microscopic analysis of muscle biopsy samples taken during para­ lysis in 17 patients with TPP showed no abnormalities in 23.5% of patients, sarcolemmal nuclear proliferation in 35.5%, atrophy of muscle fibers in 29.4%, central nuclei in 23.5%, fatty infiltration in 17.6%, vacuolation in 11.8%, and sarco­plasmic masses in 11.8%. Of these patients, 10 also had their muscle specimens examined by electron microscopy. The main changes observed were vacuola‑ tion (90%), mitochondrial abnormalities (100%), accu‑ mulation of glycogen granules (100%), disruption of the myofibers (50%), and changes in the tubular T‑system

(40%). The mitochondrial abnormalities are not specific for TPP or other primary mitochondrial disorders, but secondary to thyrotoxicosis that leads to increased oxi‑ dative metabolism and muscular oxidative damage. An important observation was that there was no correlation between these findings and the severity of muscle weak‑ ness or hypokalemia.80 Another study that was based on muscle biopsy samples from three patients demonstrated that in patients with TPP the action of the thyroid hor‑ mones not only causes metabolic alterations but also leads to changes in the structure of the membranes of the sarcolem­ma and tubular T‑system, which could affect the processes of coupling excitation–contraction with subsequent paresis or paralysis.81 As these tests do not aid in the diagnosis or prognosis of TPP, we do not consider them in TPP diagnostic procedures. Similarly, an analysis of cerebral spinal fluid does not seem necessary. Four of our patients underwent cere‑ bral spinal fluid analysis at the emergency room and had normal results.1

Clinical management of TPP Thyrotoxicosis in patients with or without TPP should ini‑ tially be managed according to its cause, for example, with the use of thionamides (20–60 mg of methimazole once daily or 200–600 mg of propyl­thiouracil daily, divided into three doses) if the cause is hyperthyroidism or with the discontinuation of thyroid hormone supplements in patients with exogenous thyrotoxicosis. During the paralysis attacks, the main reason for potassium replace‑ ment therapy is to enable the patient to recover from paralysis and to prevent cardiac arrhythmia. We recom‑ mend the administration of potassium, preferably orally or 25–50 mmol/l intravenously at a slow pace (over 2 h) during attacks and in the hospital setting, which should be stopped at the first sign of muscle strength recovery. As the replacement therapy adds to the physiologic potas‑ sium efflux, rebound hyperkalemia might occur. This phe‑ nomenon was noted in two of our patients and in other cases;13,82 rebound hyperkalemia occurred in as many as 42% of patients in one series of 24 episodes of paralysis.13 As the potassium pool is not depleted and hypokalemia is attributable to potassium shift into the cells, potassium administration between attacks will not prevent future attacks and thus is not advised. In our clinic, propranolol was given to all patients as an adjuvant to the antithyroid drug (most often methimazole) and to nearly half of the patients as an initial treatment until the diagnosis of thyrotoxicosis was confirmed. In fact, two patients who received propranolol mono­therapy did not experience paralysis attacks while they were waiting for diagnostic tests. Hence, a non­selective β adrenergic blocker, such as propranolol (80–160 mg per day), might be useful in these situations. By blocking adrenergic overstimula‑ tion of the Na+/K+ ATPase pump, propranolol reduces the intracellular shift of potassium. Propranolol has been sug‑ gested as a first-line therapy for patients with TPP72 during acute attacks because it does not induce rebound hyper‑ kalemia. Propranolol might also resolve muscle weakness more rapidly 72,82,83 than potassium replacement alone and

NATURE REVIEWS | ENDOCRINOLOGY

ADVANCE ONLINE PUBLICATION  |  5 © 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS Feed-forward model for TPP pathophysiology

Thyrotoxicosis of any cause

Genetic susceptibility Kir2.6 and other channel mutations TPP

Precipitating factors Rest after intense exercise High carbohydrate and/or salt ingestion

Figure 1 | Etiopathogenic hypothesis. A combination of genetics, thyrotoxicosis and environmental factors result in TPP. Abbreviation: TPP, thyrotoxic periodic paralysis. Modified with permission from Silva, M. R. et al. Arq. Bras. Endocrinol. Metabol. 48, 196–215 (2004).

effectively reduces the frequency of attacks.72,84–86 Although acetazolamide and dichlorphena­mide, which are carbonic anhydrase inhibitors, are used to treat patients with familial hypokalemic periodic paralysis, they are not indicated in patients with TPP. Similarly, the use of spironolactone is not usually recommended because of the periodic features of the muscle manifestations of TPP, in which hypokalemia is transient and the potassium pool is normal.38 Early and definitive treatment should be pursued with the aim of resolving thyrotoxicosis; radioactive iodine administration for thyroid ablation is usually the first choice. Once euthyroidism has been established, the patients no longer present with paralytic attacks and show good clinical outcomes if treatment compliance is achieved. In the meantime, patients should be warned to avoid pos‑ sible precipitating factors, such as c­arbohydrate-rich meals and intense physical activity. This benign clini­cal course is characteristic of TPP, in contrast with the familial form, which requires continuous treatment and can lead to pe­rmanent muscle weakness.87 In our cohort of 33 patients, the ­follow-up time varied from 1 to 29 years.1,3,11,15 Seven patients were diagnosed with hypokalemic paralysis during their first attack, while the remaining 26 were given this diagnosis from 1 week to 12 years after the first attack (median of 5 months). Although they were sympto­matic for thyrotoxicosis, most patients were first diagnosed with periodic para­lysis, which reinforces the idea that the attending physician in the emergency room should suspect TPP even in the absence of a history of thyrotoxicosis. Five patients were lost to ­follow-up before they received definitive treatment for hyper­thyroidism. These data and the low level of compli‑ ance often observed in patients with Graves disease support our proposal of definitively treating all individuals with TPP as soon as possible. In fact, 22 of our patients received radioactive iodine therapy. None of the patients had further attacks after their thyrotoxicosis was resolved.

Increasing evidence indicates that TPP results from the combination of three factors: genetics, thyrotoxicosis and environment (Figure 1). The intersection of these factors makes this condition unique. This observation has led to the hypothesis that hormonal modulators (such as excessive levels of T3 and testosterone), c­arbohydrate-rich meals and rest following exercise could exert their effects by altering ion channel dynamics in the cell membranes of neuromuscular junctions; these factors converge to trigger paralysis in susceptible patients.1,3,14 In favor of this hypothesis, a patient who has had TPP in the past could experience new attacks if he or she is given excess amounts of thyroid hormones. In fact, when T3 was administered to a patient who had been diagnosed with TPP 7 years earlier, the same pattern of paralytic attacks was observed.88 Excess T3 has a major role in revealing genetic defects in patients with TPP.1,89 Several studies have shown that the activity of the 3Na+/2K+ ATPase pump is increased in patients with thyrotoxicosis and that it increases to a greater degree in patients who have TPP.69,90 In ad­dition, patients with TPP produce increased levels of in­sulin during attacks and have raised levels of basal insu­lin between attacks.91 Another study has confirmed this find­ing and asso­ciated the physio­pathologic role of hyper­insulinemia with increased activity of the 3Na+/2K+ ATPase pump in patients with TPP. This association was made because the patients showed an increased insulin response in a glucose tolerance test and higher platelet 3Na+/2K+ ATPase acti­vity compared with healthy control indivi­duals and thyro­toxic patients without paralysis.90 Curiously, in the four patients with TPP who were studied after they received anti­thyroid treatment, the results were similar to those of healthy controls.90 In a study published in 2009, Soonthornpun et al.92 compared insulin sensitivity between patients with a history of TPP and those with a history of thyrotoxicosis but no periodic paralysis, using a 75 g oral glucose toler‑ ance test and a euglycemic hyperinsulinemic clamp. They found that the patients with TPP had considerably lower insulin sensitivity than the patients who had thyro­toxicosis without periodic paralysis. They also noted that, overall, the patients with TPP had a higher BMI and greater waist circumference than the patients who were thyrotoxic without paralysis.92 We performed a glucose tolerance test in three patients, using a 75 g oral glucose overload. One patient presented with paraparesis with absent deep tendon reflexes within 30 min of the glucose overload.1 Levels of insulin, glucose and potassium were also mea‑ sured (Figure 2). This hyperinsulinemic response to a glucose stimulus is in accordance with published find‑ ings.90,91,93,94 However, it should be noted that we do not recommend this or other provocative tests. T3 also has an indirect effect through adrenergic stimu‑ lation, which results in an increase in 3Na+/2K+ ATPase activity. This effect is clinically evident in the sympto­ matic relief of patients with TPP after the administra­tion of propranolol, which prevents new paralytic attacks.72,83 As skeletal muscle cells have a high content of potas‑ sium channels, they efficiently control the exchange of

6  |  ADVANCE ONLINE PUBLICATION

www.nature.com/nrendo © 2011 Macmillan Publishers Limited. All rights reserved

a

4.0 – 3.5 – 3.0 – 2.5 – 2.0 – 0–

b

300 –

Level of insulin (UI/l)

250 – 200 – 150 – 100 – 50 – 0–

c

300 – 250 –

Level of glucose (mg/dl)

potassium between the cell and the extracellular compart­ ment, which makes the muscle tissue the guardian of the largest pool of potassium in the human body.95 The mechanism through which hypokalemia leads to muscle weakness relies on a change in the rest potential of the skeletal muscle-cell membrane. This change results in an excitation–contraction perturbation, a phenomenon known as depolarization-induced loss of muscle excit‑ ability, also considered to be the common final pathway in patients with familial hypokalemic periodic para­lysis.96–98 The resulting depolarized state inactivates the opening of the sodium channel, which inhibits the build-up of a new action potential.99 Depolarized intercostal muscle fibers of patients with familial hypokalemic periodic paralysis have increased action potential thres­holds and the fraction of excitable muscle fibers decreases with increasing fiber depolarization; moreover, insulin reduces the conductance of the inward rectifier K+ channel for outward-flowing current­s, potentiating the depolarization of these fibers.100 In a study published in 2010, Puwanant et al.101 elegantly demonstrated that inward rectifier K+ (IKir) and voltagegated Na+ currents were reduced, and the action potential threshold increased in intercostal muscle biopsy samples from one patient with TPP during thyro­toxicosis, whereas when the patient was euthyroid they had values similar to those of control individuals. Interestingly, this phe‑ nomenon was also observed in fast contracting muscle fibers (type II),101 on which T3 has its primary effect, and also increases the insulin-regulated glucose transporter (GLUT4) content in response to the high oxidative and glycolytic demand. In addition, the reduced outward component of IKir is suppressed further when insulin is present, which reinforces our hypo­thetical model of feedforward thyrotoxic pathophysiology. This model repre‑ sents a complex metabolic postprandial response to the carbohydrate overload. T3 can regulate ion channels in a transcriptional15 or post-transcriptional manner and, indirectly, through the activity of other kinases or phosphatases,102,103 which causes channels to respond to thyrotoxic metabolic stress. Watanabe et al.104 analyzed the effect of T3 on rat atrium myocytes and demonstrated that T3 increases the outward currents of ion channels through the transcriptional regula­ tion of voltage-gated K+ channels (Kv1.5) and inward recti­ fiers. T3 also decreases the inward currents, particularly by decreasing the mRNA expression level of the L‑type calcium channel (Cav1.2).104 Consequently, excess T3 could precipitate paralysis attacks in patients with a primary susceptible genetic defect in one of these channels. In addition, thyrotoxic animals present predominantly with impairment of the fast skeletal muscle fibers, a type of fiber in which inward rectifier K+ channels are essential for facilitating K+ ion re-entry from K+-loaded transverse tubules after each action potential and to prevent dumping of dangerous amounts of K+ into the circulation from pathologically depolarized muscles.105,106 Androgens are also capable of increasing 3Na+/2K+ ATPase activity,107,108 which lends credence to the theory that there is a combined hormonal contribution to the attacks. Two white patients with TPP have been further

Level of potassium (mmol/l)

REVIEWS

200 – 150 – 100 – 50 – 0– 0

30

60 Time (min)

90

120

Figure 2 | Representative oral glucose tolerance test of a patient with thyrotoxic periodic paralysis—glucose, insulin and potassium levels. The red arrow indicates the moment of the attack of muscle weakness.

diagnosed with unilateral adrenal adenoma and hyper­ androgenemia.107 We speculate that estradiol and testo­ sterone are involved in protecting against and facilitating TPP, respectively, which has also been suggested by find‑ ings from studies based on the remodeling of K v4.3 po­tassium channel gene expression.109 Hyperinsulinemia, together with glucose, hyperfunction­ ing 3Na+/2K+ ATPase pump and post­exercise counterregulatory mechanisms, ultimately exacerbates potas­sium uptake by muscle cells. An abnormal Kir channel is unable to compensate for this potassium shift, there­fore, the arrested potassium content results in a more positive intracellular charge. Taking these find­ings together, we propose a unifying patho­physiological me­chanism that integrates genetic susceptibility and triggering factors

NATURE REVIEWS | ENDOCRINOLOGY

ADVANCE ONLINE PUBLICATION  |  7 © 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS Thyrotoxicosis – Antithyroid drugs and β2 antagonists

Antithyroid drugs – and β2 antagonists Catecholamine-induced hyperinsulinemia

Postexercise reuptake of K+ into muscle cell

Hypersensitivity to glucose load

Rebound response after intense exercise

Pancreatic β cell (with carbohydrate load)

Skeletal muscle cell (postexercise)

3Na/2K ATPase activity

3Na+ 2K+

Nav1.4

Cav1.2/1.3 ATP

+ +

3Na+

Ca2+

Phospho pumps

2K+

Nav1.4

+ +

+ + +

ATP

A2 RC RyR SE

GLUT4

– ATP/ADP

ATP

ATP

Glucolysis

KATP Kir + + +

Glucolysis GLUT2

3Na+

Ca2+

Na+

PKA cAMP

SUR

Phospho pumps

2K+

+ + +

PKA cAMP Transient [K+] i

Insulin receptor

Ca2+

Na+

Cav1.1

ATP ∆Ψ

Cav1.1

Skeletal muscle cell (postexercise and/or with glucose load)

Ca2+ Sustained [K+]i ATP/ADP K+ – SUR SUR + + + + KATP KATP

3Na+

cAMP

K+ Kir KATP

2K+

ATP/ADP

GLUT4 β-oxidation

∆Ψ Kir Kir

+

K Normal [K+] i

PKA Transient [K+]i

SUR

K+

Raised [K+] i

Normal channel

Mutated channel

Normal muscle excitation–contraction coupling

Depolarization-induced acute loss of muscle excitability

ATP 2K+ 3Na+

Figure 3 | Thyrotoxic periodic paralysis pathophysiology hypothesis; major events that determine insulin sensitivity and the response of potassium intracellular shift in skeletal muscles are enhanced by excess T 3 and increased activity of the 3Na/2K ATPase pump. Exercise causes increased glucose transport into the muscle fibers independently of insulin activity, and rapidly reverses after cessation of exercise. A high-carbohydrate meal after exercise shortens the reversal of insulin sensitivity; this response is enhanced by excess levels of T3. Glucose transport mediated by increased GLUT4 is high in type II muscle fibers. T3 modulates these fibers and is also a potent inducer of GLUT4 content in skeletal muscle. The rebound response seen after the hyperkalemia induced by intense exercise might lead to transient hypokalemia, which could trigger paralytic attacks in genetically susceptible thyrotoxic patients. Administration of propranolol can prevent the postexercise potassium rebound. Control of potassium homeostasis is impaired in thyrotoxic patients with mutations in the skeletal muscle inward-rectifier potassium channels, resulting in membrane depolarization attributable to potassium arrest inside muscle cells that leads to sodium channel inactivation. Depolarization-induced loss of muscle excitability is considered the common pathway in all forms of hypokalemic periodic paralyses. Abbreviations: cAMP, cyclic AMP; GLUT4, glucose transporter type 4; Kir, inward rectifier K+ channel; KATP, ATP-sensitive K+ channel (Kir + SUR); PKA, protein kinase A; RyR, ryanodine receptor; SERCA2, sarcoplasmic/endoplasmic reticulum calcium ATPase 2; SUR, sulfonylurea receptor.

under thyrotoxic conditions, in which the effects of excess T3 on skeletal muscle and pancreatic β cells converge to facilitate a depolarization-induced acute loss of muscle excitability (Figure 3).

Genetic susceptibility to TPP For many years, we have been searching for candidate genes for TPP; however, the genetic screening process is difficult because of the largely sporadic presentation and low incidence of the condition, which increases the complexity of testing associations between mutations and even predisposing polymorphisms. Thus, we first analyzed

the genetic findings from studies in patients with familial hypokalemic periodic paralysis, an autosomal dominant neuromuscular disorder,13,45 which is largely caused by mutations in the skeletal muscle calcium channel gene (CACNA1S) and the skeletal muscle sodium channel gene (SCN4A).110 Our group and others have demon‑ strated that known mutations in CACNA1S and SCN4A are not present in patients with TPP.1,3,14,16,111,112 In addi‑ tion, a mutation that was pre­viously identified in KCNE3 (Arg83His),97 the same mutation that we found in one proband and in two (of three) of his descendents,14 was later found to be a polymorphic variant.113,114 A search for

8  |  ADVANCE ONLINE PUBLICATION

www.nature.com/nrendo © 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS genetic poly­morphisms in CACNA1S found no associ­ ation between its genetic variations and TPP in 98 Chinese patients.115 A study of patients with Andersen syndrome (periodic paralysis associated with cardiac arrhythmia and dysmorphic features), identified a mutation in KCNJ2 that encodes the inward rectifier potassium channel Kir2.1;116 however, we found no mutant form of this gene in our TPP group.1 As other Kir2.x paralogs have constitutive cis-­elements in their regulatory regions at the genomic level that respond to T3, we screened the coding sequence of KCNJ12 (Kir2.2), KCNJ4 (Kir2.3) and KCNJ14 (Kir2.4). While performing low-stringency PCR and direct sequencing of Kir2.2, we found a new paralog and named it KCNJ18 (Kir2.6).15 This novel channel gene had been repeatedly reported as Kir2.2 single nucleotide polymorphisms; however, its hidden linkage disequilibrium seemed unknown. KCNJ18 is located on 17p11.1–2 and encodes a functional inwardly rectifying potassium channel. This channel is expressed in skeletal muscle and is transcriptionally regulated by T3.15 Six mutations in Kir2.6 have been found to be associated with TPP: Arg205His, Thr354Met, Lys366Arg, Arg399X, Gln407X and Ile144fsX7.15 When analyzing different ethnic groups, these mutations were found with different frequen‑ cies and were present in up to 33% of patients from Brazil, France and the USA.15 The missense mutations Arg205His, Thr354Met and Lys366Arg are located at residues that are conserved among human Kir2 family members and across other species. The plasma membrane phosphoinositide phosphatidylinositol 4,5-bisphosphate (PIP2) controls the activity of many ion channels through direct electrostatic interactions, which leads to the deregulation of mutated channel activity in tissues where it fully modulates cell repolarization.117 We also demonstrated that Arg205His, Thr354Met and Lys366Arg mutations are hypermorphic alleles and seem to increase PIP2 affinity, probably altering its performance under thyrotoxic conditions.15 However, the most common Kir2.6 mutations, 15 Arg399X and Gln407X, should underlie a loss-of-­function mechanism that would probably interfere with the assem‑ bly of potassium channels either as homo­tetramers or hetero­tetramers. Both mutations are located at the C‑terminus of the channel and lack the PDZ binding domain boundaries that can compromise assembly and trafficking of the protein to the cell membrane. 118 As demon­strated in Kir2.2 in skeletal muscle, mutations in this region might result in altered sub­cellular localiza­ tion.119 In addition, immunofluorescent localization 1.

2. 3.

Silva, M. R., Chiamolera, M. I., Kasamatsu, T. S., Cerutti, J. M. & Maciel, R. M. Thyrotoxic hypokalemic periodic paralysis, an endocrine emergency: clinical and genetic features in 25 patients [Portuguese]. Arq. Bras. Endocrinol. Metabol. 48, 196–215 (2004). Lin, S. H. Thyrotoxic periodic paralysis. Mayo Clin. Proc. 80, 99–105 (2005). Dias da Silva, M. R. et al. Mutations linked to familial hypokalaemic periodic paralysis in the calcium channel alpha1 subunit gene (Cav1.1) are not associated with thyrotoxic hypokalaemic periodic paralysis. Clin. Endocrinol. (Oxf.) 56, 367–375 (2002).

4.

5.

6.

7.

studies have demonstrated that Kir2.2 co-localizes with syntrophin, dystrophin and dystrobrevin at skeletal muscle neuromuscular junctions, dynamically anchor‑ ing and stabilizing channels in the plasma membrane. We believe that amorphic and antimorphic alleles could produce decreased outward potassium currents, leading to depolarization and transitioning of voltage-gated channels to their inactivated states, whereas hypermorphic alleles could cause hyperpolarization and difficulty in reaching the action potential threshold.

Conclusions TPP is often underdiagnosed. This disease should be included in the differential diagnosis of acute muscle weakness in young patients, especially males, regardless of their ethnic background. Awareness of TPP among healthcare providers and good clinical assessments are important because many patients are unaware of thyrotoxicosis but will reveal signs and/or symptoms if properly questioned and examined. Thyroid function tests are mandatory in patients with periodic paralysis. Normal potassium levels during paralysis attacks do not exclude the possibility of TPP, as this finding might reflect recovery or factors such as hemolysis. Oral, rather than intravenous, potassium should be administered to correct hypokalemia, shorten the duration of the paralytic attack, and avoid severe cardiac complications secondary to potassium rebound. Propranolol is beneficial as a first-line therapy to control these attacks while waiting for a confirmation of thyro‑ toxicosis and planning definitive treatment. Prophylactic administration of potassium between attacks is not recom­mended. The aim should be definitive treatment of thyrotoxicosis. Candidate genes should be examined for mutations to confirm genetic heterogeneity in TPP, as has been seen for other channelopathies.120 Review criteria The content of this Review is based on the collective knowledge of the authors and literature collected over the course of their careers. Therefore, a formal literature search was not required to find the majority of the papers cited. Scielo, MEDLINE and PubMed were searched with no language or date restrictions. The search terms used were “thyrotoxic periodic paralysis”, “thyrotoxicosis”, “hypokalemic periodic paralysis” and “channel gene mutation”. Most references are full-text articles. We also searched the reference lists of identified articles for further papers.

Hackshaw, K. V. & Coker, E. Hypokalemic periodic paralysis in a hyperthyroid black woman. J. Natl Med. Assoc. 80, 1343–1344 (1988). Kelley, D. E., Gharib, H., Kennedy, F. P., Duda, R. J. Jr & McManis, P. G. Thyrotoxic periodic paralysis. Report of 10 cases and review of electromyographic findings. Arch. Intern. Med. 149, 2597–2600 (1989). Hay, I. D. & McDougall, I. R. Letter: Thyrotoxic periodic paralysis in Britain. Br. Med. J. 4, 761 (1975). Engel, A. G. Thyroid function and periodic paralysis. Am. J. Med. 30, 327–333 (1961).

NATURE REVIEWS | ENDOCRINOLOGY

8.

Rosenfeld, M. Acute aufsteigende lahmung bei morbus Basedow [German]. Klin. Wochenschr. (Berl.) 39, 538–540 (1902). 9. Dunlap, H. F. & Kepler, E. J. A syndrome resembling familial periodic paralysis occurring in the course of exolphthalmic goiter. Endocrinology 15, 541–546 (1931). 10. Pereira, V. G., Wajchenberg, B. L. & Quintão, E. R. Periodic paralysis associated with hyperthyroidism. Study of the precipitating factors. Arq. Bras. Endocrinol. Metabol. 17, 55–67 (1968). 11. Sterian, L. & Maciel, R. M. B. Paralisia periódica tirotóxica, relato de um caso e comentários

ADVANCE ONLINE PUBLICATION  |  9 © 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS

12.

13.

14.

15.

16.

17. 18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

etiopatogênicos [Portuguese]. Arq. Bras. Endocrinol. Metabol. 27, 161–162 (1983). Ober, K. P. Thyrotoxic periodic paralysis in the United States. Report of 7 cases and review of the literature. Medicine (Baltimore) 71, 109–120 (1992). Manoukian, M. A., Foote, J. A. & Crapo, L. M. Clinical and metabolic features of thyrotoxic periodic paralysis in 24 episodes. Arch. Intern. Med. 159, 601–606 (1999). Dias da Silva, M. R., Cerutti, J. M., Arnaldi, L. A. & Maciel, R. M. A mutation in the KCNE3 potassium channel gene is associated with susceptibility to thyrotoxic hypokalemic periodic paralysis. J. Clin. Endocrinol. Metab. 87, 4881–4884 (2002). Ryan, D. P. et al. Mutations in potassium channel Kir2.6 cause susceptibility to thyrotoxic hypokalemic periodic paralysis. Cell 140, 88–98 (2010). Chen, L. et al. Clinical and molecular analysis of Chinese patients with thyrotoxic periodic paralysis. Eur. Neurol. 49, 227–230 (2003). Ko, G. T. et al. Thyrotoxic periodic paralysis in a Chinese population. QJM 89, 463–468 (1996). Zhang, M. Y. Clinical analysis of 50 cases of thyrotoxic periodic paralysis [Chinese]. Zhonghua Shen Jing Jing Shen Ke Za Zhi 20, 33–36 (1987). Tamai, H. et al. HLA and thyrotoxic periodic paralysis in Japanese patients. J. Clin. Endocrinol. Metab. 64, 1075–1078 (1987). Okihiro, M. M. & Beddow, R. M. Thyrotoxic periodic paralysis in Hawaii. Its predilection for the Japanese race. Neurology 15, 253–255 (1965). Shiang, J. C. et al. Therapeutic analysis in Chinese patients with thyrotoxic periodic paralysis over 6 years. Eur. J. Endocrinol. 161, 911–916 (2009). Maciel, R. M. B. Screening for thyrotoxic paralysis out of Asia: are we underestimating its frequency in the West? Presented at the 12th International Thyroid Congress, Kyoto, Japan. Endocr. J. 47 (Suppl. 2) WS-3 (2000). Magsino, C. H. Jr & Ryan, A. J. Jr. Thyrotoxic periodic paralysis. South. Med. J. 93, 996–1003 (2000). Bernard, J. D., Larson, M. A. & Norris, F. H. Jr. Thyrotoxic periodic paralysis in Californians of Mexican and Filipino ancestry. Calif. Med. 116, 70–74 (1972). Castro Vilanova, M. D., Berbel, A., Posada, I. J. & Bermejo, F. Periodic thyrotoxic paralysis in Hispanics [Spanish]. Neurologia 12, 428–429 (1997). Cesur, M. et al. Thyrotoxic hypokalaemic periodic paralysis in a Turkish population: three new case reports and analysis of the case series. Clin. Endocrinol. (Oxf.) 68, 143–152 (2008). Chatterjee, A. K., Dhakam, Z., Morrish, N. J. & Manford, M. R. Thyrotoxic periodic paralysis in a Caucasian male. Hosp. Med. 65, 182–183 (2004). Elston, M. S., Orr-Walker, B. J., Dissanayake, A. M. & Conaglen, J. V. Thyrotoxic, hypokalaemic periodic paralysis: Polynesians, an ethnic group at risk. Intern. Med. J. 37, 303–307 (2007). Gordon, D. L., Agrawal, L., Swade, T. F. & Lawrence, A. M. Thyrotoxic hypokalemic periodic paralysis: six cases in non-Asian patients. Endocr. Pract. 4, 142–145 (1998). Kilpatrick, R. E., Seiler-Smith, S. & Levine, S. N. Thyrotoxic hypokalemic periodic paralysis: report of four cases in black American males. Thyroid 4, 441–445 (1994). Pichon, B. et al. Thyrotoxic periodic paralysis in Caucasian patients: a diagnostic challenge. Eur. J. Intern. Med. 16, 372–374 (2005).

32. Piraino Neuenschwander, P., Pumarino Carte, H., Bidegain González, F., Zura Jiménez, M. L. & Ferreiro Merino, F. Thyrotoxic hypokalemic periodic paralysis: 18 cases with different forms of thyrotoxicosis [Spanish]. Rev. Clin. Esp. 195, 294–297 (1995). 33. Reisin, R. C. et al. Thyrotoxic periodic paralysis in caucasians. Report of 8 cases. Neurologia 15, 222–225 (2000). 34. Sinharay, R. Hypokalaemic thyrotoxic periodic paralysis in the UK. QJM 96, 611–613 (2003). 35. Wild, E. Thyrotoxic periodic paralysis in a Maori patient. N. Z. Med. J. 117, U1204 (2004). 36. Williams, R. Thyrotoxic periodic paralysis in Navajos--HLA typing. West. J. Med. 142, 260–261 (1985). 37. Tran, H. A. et al. Thyrotoxic, hypokalaemic periodic paralysis in Australasian men. Intern. Med. J. 33, 91–94 (2003). 38. McFadzean, A. J. & Yeung, R. Familial occurrence of thyrotoxic periodic paralysis. Br. Med. J. 1, 760 (1969). 39. Shizume, K. et al. Comparison of the incidence of association of periodic paralysis and hyperthyroidism in Japan in 1957 and 1991. Endocrinol. Jpn. 39, 315–318 (1992). 40. McKenzie, T. Hypokalemic periodic paralysis of thyrotoxic origin: a case complicated by cultural differences. Can. Med. Assoc. J. 128, 950–951 (1983). 41. Nellen, H. et al. Thyrotoxic periodic paralysis in Mexican mestizo patients: a clinical, biochemical and HLA-serological study. Arch. Med. Res. 30, 74–76 (1999). 42. Iheonunekwu, N. C., Ibrahim, T. M., Davies, D. & Pickering, K. Thyrotoxic hypokalaemic paralysis in a pregnant Afro-Caribbean woman. A case report and review of the literature. West Indian Med. J. 53, 47–49 (2004). 43. Donovan, L., Parkins, V. M. & Mahalingham, A. Thyrotoxic periodic paralysis in pregnancy with impaired glucose tolerance: a case report and discussion of management issues. Thyroid 17, 579–583 (2007). 44. Richey, S. D. & Wendel, G. D. Jr. Thyrotoxic hypokalemic periodic paralysis following secondtrimester prostaglandin-induced abortion. Obstet. Gynecol. 82 (Suppl.), 696–698 (1993). 45. Ptácek, L. J. Channel surfing. J. Clin. Endocrinol. Metab. 87, 4879–4880 (2002). 46. Leung, A. K. Familial ‘hashitoxic’ periodic paralysis. J. R. Soc. Med. 78, 638–640 (1985). 47. Hsieh, M. J. et al. Hypokalemic thyrotoxic periodic paralysis: clinical characteristics and predictors of recurrent paralytic attacks. Eur. J. Neurol. 15, 559–564 (2008). 48. Wong, G. W., Leung, T. F., Lo, A. F., Ahuja, A. T. & Cheng, P. S. Thyrotoxic periodic paralysis in a 14‑year‑old boy. Eur. J. Pediatr. 159, 934 (2000). 49. Tagami, T., Usui, T., Shimatsu, A. & Naruse, M. Toxic thyroid adenoma presenting as hypokalemic periodic paralysis. Endocr. J. 54, 797–803 (2007). 50. Ober, K. P. & Hennessy, J. F. Jodbasedow and thyrotoxic periodic paralysis. Arch. Intern. Med. 141, 1225–1227 (1981). 51. Laroia, S. T., Zaw, K. M., Ganti, A. K., Newman, W. & Akinwande, A. O. Amiodarone-induced thyrotoxicosis presenting as hypokalemic periodic paralysis. South. Med. J. 95, 1326–1328 (2002). 52. Kiso, Y. et al. A case of thyrotropin (TSH)secreting tumor complicated by periodic paralysis. Jpn. J. Med. 29, 399–404 (1990). 53. Hsu, F. S. et al. Thyrotropin-secreting pituitary adenoma presenting as hypokalemic periodic paralysis. Am. J. Med. Sci. 325, 48–50 (2003).

10  |  ADVANCE ONLINE PUBLICATION

54. Chen, Y. C., Fang, J. T., Chang, C. T. & Chou, H. H. Thyrotoxic periodic paralysis in a patient abusing thyroxine for weight reduction. Ren. Fail. 23, 139–142 (2001). 55. Chou, H. K., Tsao, Y. T. & Lin, S. H. An unusual cause of thyrotoxic periodic paralysis: triiodothyronine-containing weight reducing agents. Am. J. Med. Sci. 337, 71–73 (2009). 56. Hannon, M. J., Behan, L. A. & Agha, A. Thyrotoxic periodic paralysis due to excessive L‑thyroxine replacement in a Caucasian man. Ann. Clin. Biochem. 46, 423–425 (2009). 57. Lin, S. H., Davids, M. R. & Halperin, M. L. Hypokalaemia and paralysis. QJM 96, 161–169 (2003). 58. Fisher, J. Thyrotoxic periodic paralysis with ventricular fibrillation. Arch. Intern. Med. 142, 1362–1364 (1982). 59. Loh, K. C., Pinheiro, L. & Ng, K. S. Thyrotoxic periodic paralysis complicated by near-fatal ventricular arrhythmias. Singapore Med. J. 46, 88–89 (2005). 60. Miyashita, Y. et al. Ventricular fibrillation due to severe hypokalemia induced by steroid treatment in a patient with thyrotoxic periodic paralysis. Intern. Med. 45, 11–13 (2006). 61. Wu, C. Z., Wu, Y. K., Lin, J. D. & Kuo, S. W. Thyrotoxic periodic paralysis complicated by acute hypercapnic respiratory failure and ventricular tachycardia. Thyroid 18, 1321–1324 (2008). 62. Kokenge, F. & Heiner, M. Thyrotoxic periodic paralysis: a peculiar case with weekend-related quadriplegia. Endocrinologist 15, 297–299 (2005). 63. Liu, Z., Braverman, L. E. & Malabanan, A. Thyrotoxic periodic paralysis in a Hispanic man after the administration of prednisone. Endocr. Pract. 12, 427–431 (2006). 64. Kane, M. P. & Busch, R. S. Drug-induced thyrotoxic periodic paralysis. Ann. Pharmacother. 40, 778–781 (2006). 65. Wongraoprasert, S., Buranasupkajorn, P., Sridama, V. & Snabboon, T. Thyrotoxic periodic paralysis induced by pulse methylprednisolone. Intern. Med. 46, 1431–1433 (2007). 66. Kemperman, F. A., Hoff, H. C. & de Klerk, G. Hypokalemic periodic paralysis as the sole manifestation of hyperthyroidism [Dutch]. Ned. Tijdschr. Geneeskd. 139, 938–941 (1995). 67. Shizume, K. et al. Studies on electrolyte metabolism in idiopathic and thyrotoxic periodic paralysis. I. Arteriovenous differences of electrolytes during induced paralysis. Metabolism 15, 138–144 (1966). 68. Feely, J. Potassium shift in thyrotoxic periodic paralysis. Postgrad Med. J. 57, 238–239 (1981). 69. Chan, A., Shinde, R., Chow, C. C., Cockram, C. S. & Swaminathan, R. In vivo and in vitro sodium pump activity in subjects with thyrotoxic periodic paralysis. BMJ 303, 1096–1099 (1991). 70. Gaul, C., Leonhardt, G., Spens, A., Schneyer, U. & Zierz, S. Hypokalemic thyrotoxic periodic paralysis (HTPP). Rare differential diagnosis in case of acute tetraparesis in Europe [German]. Med. Klin. (Munich) 100, 583–586 (2005). 71. Neal, J. M. Thyrotoxic periodic paralysis in a Caucasian man: recognition and diagnosis. Indiana Med. 84, 696–699 (1991). 72. Lin, S. H. & Lin, Y. F. Propranolol rapidly reverses paralysis, hypokalemia, and hypophosphatemia in thyrotoxic periodic paralysis. Am. J. Kidney Dis. 37, 620–623 (2001). 73. Norris, K. C., Levine, B. & Ganesan, K. Thyrotoxic periodic paralysis associated with hypokalemia and hypophosphatemia. Am. J. Kidney Dis. 28, 270–273 (1996).

www.nature.com/nrendo © 2011 Macmillan Publishers Limited. All rights reserved

REVIEWS 74. Lin, S. H. et al. Early diagnosis of thyrotoxic periodic paralysis: spot urine calcium to phosphate ratio. Crit. Care Med. 34, 2984–2989 (2006). 75. Bernard, P., Rault, A. & Talarmin, F. Hypokalemic periodic paralysis at initiation of Graves’ disease treatment. Two observations [French]. Presse Med. 30, 221 (2001). 76. Navarro, V., Fournier, E., Girard, S., Delattre, J. Y. & Eymard, B. Periodic hypokalemic paralysis as the manifestation of Graves’ disease: clinical and electrophysiological study [French]. Rev. Neurol. (Paris) 156, 59–61 (2000). 77. Hawkins, B. R., Ma, J. T., Lam, K. S., Wang, C. C. & Yeung, R. T. Association of HLA antigens with thyrotoxic Graves’ disease and periodic paralysis in Hong Kong Chinese. Clin. Endocrinol. (Oxf.) 23, 245–252 (1985). 78. Puvanendran, K., Cheah, J. S. & Wong, P. K. Electromyographic (EMG) study in thyrotoxic periodic paralysis. Aust. N. Z. J. Med. 7, 507–510 (1977). 79. Tengan, C. H. et al. Periodic paralysis: anatomopathological study of skeletal muscles in 14 patients [Portuguese]. Arq. Neuropsiquiatr. 52, 32–40 (1994). 80. Cheah, J. S., Tock, E. P. & Kan, S. P. The light and electron microscopic changes in the skeletal muscles during paralysis in thyrotoxic periodic paralysis. Am. J. Med. Sci. 269, 365–374 (1975). 81. Sokolina, N. A., Il’ina, N. A. & Korolev, V. V. Change in the ultrastructure of the muscles in thyrotoxic periodic paralysis [Russian]. Arkh. Patol. 38, 27–33 (1976). 82. Tassone, H., Moulin, A. & Henderson, S. O. The pitfalls of potassium replacement in thyrotoxic periodic paralysis: a case report and review of the literature. J. Emerg. Med. 26, 157–161 (2004). 83. Shayne, P. & Hart, A. Thyrotoxic periodic paralysis terminated with intravenous propranolol. Ann. Emerg. Med. 24, 736–740 (1994). 84. Huang, T. Y. & Lin, S. H. Thyrotoxic hypokalemic periodic paralysis reversed by propranolol without rebound hyperkalemia. Ann. Emerg. Med. 37, 415–416 (2001). 85. Birkhahn, R. H., Gaeta, T. J. & Melniker, L. Thyrotoxic periodic paralysis and intravenous propranolol in the emergency setting. J. Emerg. Med. 18, 199–202 (2000). 86. Yeung, R. T. & Tse, T. F. Thyrotoxic periodic paralysis. Effect of propranolol. Am. J. Med. 57, 584–590 (1974). 87. Caciotti, A., Morrone, A., Domenici, R., Donati, M. A. & Zammarchi, E. Severe prognosis in a large family with hypokalemic periodic paralysis. Muscle Nerve 27, 165–169 (2003). 88. Okihiro, M. M. & Nordyke, R. A. Hypokalemic periodic paralysis. Experimental precipitation with sodium liothyronine. JAMA 198, 949–951 (1966). 89. Gallagher, E. J. Is thyrotoxic periodic paralysis a “channelopathy”? J. Emerg. Med. 17, 123–127 (1999). 90. Chan, A., Shinde, R., Chow, C. C., Cockram, C. S. & Swaminathan, R. Hyperinsulinaemia and Na+, K(+)-ATPase activity in thyrotoxic periodic paralysis. Clin. Endocrinol. (Oxf.) 41, 213–216 (1994). 91. Shishiba, Y., Shimizu, T., Saito, T. & Shizume, K. Elevated immunoreactive insulin concentration

during spontaneous attacks in thyrotoxic periodic paralysis. Metabolism 21, 285–290 (1972). 92. Soonthornpun, S., Setasuban, W. & Thamprasit, A. Insulin resistance in subjects with a history of thyrotoxic periodic paralysis (TPP). Clin. Endocrinol. (Oxf.) 70, 794–797 (2009). 93. Hamada, N. et al. Transient glucose intolerance during attacks of thyrotoxic periodic paralysis. Horm. Metab. Res. 17, 301–305 (1985). 94. Holloszy, J. O. & Hansen, P. A. Regulation of glucose transport into skeletal muscle. Rev. Physiol. Biochem. Pharmacol. 128, 99–193 (1996). 95. Clausen, T. Hormonal and pharmacological modification of plasma potassium homeostasis. Fundam. Clin. Pharmacol. 24, 595–605 (2010). 96. Antes, L. M., Kujubu, D. A. & Fernandez, P. C. Hypokalemia and the pathology of ion transport molecules. Semin. Nephrol. 18, 31–45 (1998). 97. Abbott, G. W. et al. MiRP2 forms potassium channels in skeletal muscle with Kv3.4 and is associated with periodic paralysis. Cell 104, 217–231 (2001). 98. Tricarico, D., Servidei, S., Tonali, P., Jurkat-Rott, K. & Camerino, D. C. Impairment of skeletal muscle adenosine triphosphate-sensitive K+ channels in patients with hypokalemic periodic paralysis. J. Clin. Invest. 103, 675–682 (1999). 99. Cannon, S. C. An expanding view for the molecular basis of familial periodic paralysis. Neuromuscul. Disord. 12, 533–543 (2002). 100. Ruff, R. L. Insulin acts in hypokalemic periodic paralysis by reducing inward rectifier K+ current. Neurology 53, 1556–1563 (1999). 101. Puwanant, A. & Ruff, R. L. INa and IKir are reduced in Type 1 hypokalemic and thyrotoxic periodic paralysis. Muscle Nerve 42, 315–327 (2010). 102. Lin, H. et al. Thyroid hormones suppress ε-PKC signalling, down-regulate connexin‑43 and increase lethal arrhythmia susceptibility in non-diabetic and diabetic rat hearts. J. Physiol. Pharmacol. 59, 271–285 (2008). 103. Muscella, A. et al. PKC‑ε‑dependent cytosol‑to‑membrane translocation of pendrin in rat thyroid PC Cl3 cells. J. Cell Physiol. 217, 103–112 (2008). 104. Watanabe, H. et al. Thyroid hormone regulates mRNA expression and currents of ion channels in rat atrium. Biochem. Biophys. Res. Commun. 308, 439–444 (2003). 105. Casas, F. et al. Overexpression of the mitochondrial T3 receptor p43 induces a shift in skeletal muscle fiber types. PLoS ONE 3, e2501 (2008). 106. Miyabara, E. H. et al. Thyroid hormone receptor‑β‑selective agonist GC‑24 spares skeletal muscle type I to II fiber shift. Cell Tissue Res. 321, 233–241 (2005). 107. Biering, H., Bauditz, J., Pirlich, M., Lochs, H. & Gerl, H. Manifestation of thyrotoxic periodic paralysis in two patients with adrenal adenomas and hyperandrogenaemia. Horm. Res. 59, 301–304 (2003). 108. Azzarolo, A. M. et al. Androgen support of lacrimal gland function. Endocrine 6, 39–45 (1997). 109. Song, M. et al. Remodeling of Kv4.3 potassium channel gene expression under the control of sex

NATURE REVIEWS | ENDOCRINOLOGY

hormones. J. Biol. Chem. 276, 31883–31890 (2001). 110. Ryan, D. P. & Ptácek, L. J. in The Molecular and Genetic Basis of Neurologic and Psychiatric Disease (eds Rosenberg, R. N., DiMauro, S., Paulson, H. L., Ptácek, L. J. & Nestler, E. J.) 550–568 (Lippincott Williams & Wilkins, Philadelphia, 2007). 111. Ng, W. Y., Lui, K. F., Thai, A. C. & Cheah, J. S. Absence of ion channels CACN1AS and SCN4A mutations in thyrotoxic hypokalemic periodic paralysis. Thyroid 14, 187–190 (2004). 112. Wang, W. et al. Mutation screening in Chinese hypokalemic periodic paralysis patients. Mol. Genet. Metab. 87, 359–363 (2006). 113. Sternberg, D., Tabti, N., Fournier, E., Hainque, B. & Fontaine, B. Lack of association of the potassium channel-associated peptide MiRP2– R83H variant with periodic paralysis. Neurology 61, 857–859 (2003). 114. Lehmann-Horn, F. Disease-causing mutations or functional polymorphisms? Acta Myol. 23, 85–89 (2004). 115. Tang, N. L. et al. The α1S subunit of the L‑type calcium channel is not a predisposition gene for thyrotoxic periodic paralysis. Clin. Endocrinol. (Oxf.) 66, 229–234 (2007). 116. Plaster, N. M. et al. Mutations in Kir2.1 cause the developmental and episodic electrical phenotypes of Andersen’s syndrome. Cell 105, 511–519 (2001). 117. Logothetis, D. E., Petrou, V. I., Adney, S. K. & Mahajan, R. Channelopathies linked to plasma membrane phosphoinositides. Pflugers Arch. 460, 321–341 (2010). 118. Nehring, R. B. et al. Neuronal inwardly rectifying K(+) channels differentially couple to PDZ proteins of the PSD‑95/SAP90 family. J. Neurosci. 20, 156–162 (2000). 119. Leonoudakis, D. et al. Protein trafficking and anchoring complexes revealed by proteomic analysis of inward rectifier potassium channel (Kir2.x)-associated proteins. J. Biol. Chem. 279, 22331–22346 (2004). 120. Rolim, A. L. R. et al. Ion channelopathies in endocrinology: recent genetic findings and pathophysiological insights. Arq. Bras. Endocrinol. Metabol. 54, 673–681 (2010). Acknowledgments The authors thank the team of the Laboratory of Molecular and Translational Endocrinology, especially Reinaldo Furlanetto, Luiza Matsumura, Jairo Hidal, João Roberto Martins, Teresa Kasamatsu, Maria Izabel Chiamolera, Janete Cerutti, Ana Luiza Rolim, Cléber Camacho and Beatriz Freitas and all the collaborators at the Division of Endocrinology, Department of Medicine, Escola Paulista de Medicina, Federal University of Sao Paulo. We also thank the colleagues who referred patients for our genetic study, Drs Adelaide Albergária-Pereira, Gisah Carvalho, Maria Cristina Costa, Cássia Jurado, Cláudio Kater, Maria Conceição Mamone, Gláucia Mazeto, Gabriela Saraiva, Célia Tengan and Matthias Kugler. The authors’ research is supported by São Paulo State Research Foundation (FAPESP), Brazilian Research Council (CNPq) and Fleury Group. Author contributions All authors contributed equally to all aspects of this review.

ADVANCE ONLINE PUBLICATION  |  11 © 2011 Macmillan Publishers Limited. All rights reserved