Introduction
Amiodarone is a potent antiarrhythmic which has been used in Europe since the 1960s to treat various cardiac arrhythmias, ranging from paroxysmal atrial fibrillation to life-threatening ventricular tachyarrhythmias. Amiodarone use has also been associated with a variety of adverse effects, not least on thyroid function, and can lead to hypothyroidism or thyrotoxicosis in some patients.Reference Martino, Bartalena, Bogazzi and Braverman1 For patients requiring prolonged therapy, amiodarone-induced thyrotoxicosis is potentially life-threatening in those with significant cardiac comorbidities, and it can pose a challenge to medical management.
In the UK, amiodarone-induced thyrotoxicosis develops in 2 to 3 per cent of patients who have received therapy for an average of 21 months (although there is great variability in treatment duration prior to development of the condition); there is a much higher incidence in iodine-deficient areas.Reference Figge and Figge2 Amiodarone-induced thyrotoxicosis has been subdivided into three entities: type I, type II and a mixed type. Type I amiodarone-induced thyrotoxicosis is due to increased synthesis and secretion of thyroid hormone, in the presence of pre-existing thyroid pathology. Type II amiodarone-induced thyrotoxicosis occurs when a previously normal thyroid gland undergoes changes due to chronic amiodarone therapy, leading to a destructive thyroiditis with subsequent leakage of stored thyroid hormones (but no increase in production). There is no clear consensus on optimal medical management, and total or subtotal thyroidectomy (with its perceived risks) is often reserved as a final option in the treatment algorithm. Of concern is the reported incidence in the world literature of papillary thyroid cancer in patients with type II amiodarone-induced thyrotoxicosis, although whether this is coincidence or association is difficult to ascertain.Reference Cattaneo3, Reference Saad, Falciglia, Steward and Nikiforov4
General anaesthesia for non-cardiac surgery is not undertaken lightly in patients with Eisenmenger's syndrome as there is a significant risk of mortality. It has been suggested that these very patients with congenital cardiac disease, needing prolonged amiodarone therapy, may be at increased risk of developing amiodarone-induced thyrotoxicosis.Reference Thorne, Barnes, Cullinan and Somerville5 Furthermore, in view of the increasing use of amiodarone by physicians and the subsequent rise in the incidence of amiodarone-induced thyrotoxicosis, surgeons are likely to encounter this condition more frequently.
Case report
A 39-year-old man with a known history of Eisenmenger's syndrome (i.e. ventral septal defect, atrial septal defect and tricuspid atresia) presented with palpitations, dyspnoea and diarrhoea. He had not required corrective cardiac surgery and had been receiving amiodarone therapy (for 30 months) for paroxysmal atrial fibrillation, in addition to warfarin due to the risk of thromboembolism. There was no other comorbidity and no family history of thyroid or congenital heart disease.
On examination, the patient was anxious, dyspnoeic and peripherally cyanosed, with clubbing (in both hands) in addition to an obvious goitre. Clinical examination revealed severe thyrotoxicosis with a fine tremor and atrial fibrillation (ventricular response 160 beats/minute), but there were no signs of left ventricular failure.
Investigation revealed an erythrocytosis (haemoglobin concentration 20.0 g/dl) and biochemical confirmation of severe thyrotoxicosis (with a free thyroxine (T4) concentration of greater than 150 pmol/l and a free tri-iodothyronine (T3) concentration of 39.0 pmol/l). A technetium scan revealed poor thyroid uptake.
Despite withdrawal of amiodarone therapy and treatment with carbimazole (80 mg/day) followed by adjunctive steroids, there was continuing clinical and biochemical thyrotoxicosis, which was cause for increasing concern considering our patient's Eisenmenger's syndrome. Increasing doses of dexamethasone (6 mg/day) and initiation of propylthiouracil (increased to 800 mg/day) led to transient improvement of symptoms but relapse within days. Lithium therapy (a titrated increase to 1.2 g/day) brought no resolution of symptoms, and the patient's free T4 and free T3 concentrations reached 150 and 28.2 pmol/l, respectively. After a total of six weeks of attempted medical management, the patient was finally referred for thyroidectomy.
The patient was counselled on the significant risks of the procedure in this particular case, and full multidisciplinary consultation was initiated between the surgeon, endocrinologist, cardiologist and anaesthetists (both in theatre and intensive care), in order to minimise risk to the patient. Potassium percholate, maximal beta-blockade (with propanolol) and heparin infusion therapy was used to optimise the patient's condition prior to surgery, and intensive care was planned post-operatively.
After heavy premedication, an arterial line was inserted to allow rapid monitoring. Using careful titration, sodium thiopentone was used for induction of anaesthesia with paralysis by vecuronium. Intravenous opioid (remifentanil) was used to maintain anaesthesia, with inhalation of enflurane to prevent awareness and crystalloids for fluid replacement.
A total thyroidectomy was performed, with meticulous haemostasis to minimise blood loss and blood sampling to monitor serum free T4 throughout the procedure. The patient did not suffer any intra- or post-operative complications. Post-operatively, there was rapid resolution of amiodarone-induced thyrotoxicosis, both clinically and biochemically.
Histological examination of the resected thyroid confirmed the presence of changes consistent with amiodarone-induced thyrotoxicosis. There was no evidence of malignancy.
Literature review
A literature review was undertaken to assess the reported morbidity and mortality associated with non-cardiac surgery for Eisenmenger's syndrome and amiodarone-induced thyrotoxicosis. In view of continuing controversies and difficulties in the diagnosis and medical management of our patient's amiodarone-induced thyrotoxicosis, the alternative option of surgery for amiodarone-induced thyrotoxicosis was evaluated.
A Medline literature search was performed of articles from 1975 to the present, using combinations of the following keywords: ‘amiodarone’, ‘thyrotoxicosis’, ‘surgery and Eisenmenger's syndrome’ and ‘surgery’. Articles referring to non-cardiac surgery for Eisenmenger's syndrome were then manually selected. Analysis was also performed on selected case series to determine significant morbidity and mortality associated with surgical management of amiodarone-induced thyrotoxicosis.
A number of papers had been published on non-cardiac surgery in both adult and paediatric patients with Eisenmenger's syndrome. We reviewed a total of 19 non-cardiac procedures described in the world literature, involving 16 patients.Reference Thorne, Barnes, Cullinan and Somerville5–Reference Temelcos, Kuhn and Stribley10 In total, eight patients (undergoing 11 procedures) were children and eight were adults. No patients were older than 40 years, and none had thyroid pathology. A large variety of anaesthetic techniques were employed, ranging from xenon gas to the use of different intravenous agents. A common theme throughout the various studies was careful patient selection, intensive monitoring and maintenance of systemic vascular resistance and blood pressure. Apart from two documented episodes of transient bradycardia,Reference Lyons, Motherway, Casey and Doherty6, Reference Kraayenbrink and Steven7 general anaesthesia was well tolerated; no intra-operative mortality was recorded. Two patients had post-operative complications related to the surgical procedure (i.e. glossopharyngeal nerve palsy and hoarseness of voice); no complication could be attributed to the general anaesthetic.Reference Kraayenbrink and Steven7, Reference Kopka, McMenemin, Serpell and Quasim8 No post-operative deaths were recorded, despite some patients undergoing major procedures.
A number of case series have been published detailing the surgical management of amiodarone-induced thyrotoxicosis. Apart from our own patient, analysis of 110 cases of surgery (total or subtotal thyroidectomy) in patients with amiodarone-induced thyrotoxicosis revealed low levels of significant morbidity (7.2 per cent), with two cases each of post-operative lower respiratory tract infection, cardiac dysrhythmia and symptomatic hypocalcaemia. Two further instances of thromboembolic stroke and sepsis were also recorded.Reference Houghton, Farley, Brennan, van Heerden, Thompson and Grant11, Reference Hamoir, Meurisse, Defechereux, Joris, Vivario and Hennen12 Until 2004, there was no recorded mortality due to surgery for amiodarone-induced thyrotoxicosis, and even with the three recorded cases in one series at one surgical unit, the overall reported mortality rate was 2.7 per cent.Reference Houghton, Farley, Brennan, van Heerden, Thompson and Grant11 It must be noted that none of these patients had significant congenital cardiac comorbidity prior to surgery.
The majority of patients treated (56 per cent (55/98)) had failed medical therapy prior to surgery. Interestingly, a significant proportion (48 per cent) had surgical intervention as first-line treatment of amiodarone-induced thyrotoxicosis.Reference Houghton, Farley, Brennan, van Heerden, Thompson and Grant11–Reference Franzese, Fan and Stack21 Of the 49 cases in which classification of amiodarone-induced thyrotoxicosis could be ascertained, 14 (29 per cent) were type I and 35 (71 per cent) type II.Reference Hamoir, Meurisse, Defechereux, Joris, Vivario and Hennen12–Reference Bogazzi, Miccoli, Berti, Cosci, Brogioni and Aghini-Lombardi16, Reference Painvin, Dupuis, Victor, Rohmer, Ronceray and Tadei19, Reference Franzese, Fan and Stack21 All patients were treated definitively, resulting in rapid resolution of amiodarone-induced thyrotoxicosis, and there was no reported recurrence in those recommenced on amiodarone therapy.
Discussion
The pathogenesis of amiodarone-induced thyrotoxicosis is not fully understood. Being a fat-soluble drug that contains iodine and having a molecular structure similar to both T4 and T3, amiodarone therapy generates immense expansion of both systemic and thyroidal iodine pools. A typical 200 mg dose of amiodarone will cause the release of 100–200 times the recommended daily intake of iodine.Reference Martino, Bartalena, Bogazzi and Braverman1 Amiodarone also decreases peripheral deiodination of T4 to T3, leading to an increase in serum T4 and a decrease in T3; these changes, in addition to the cytotoxic effects of amiodarone on thyroid follicles, have been implicated in the pathogenesis of amiodarone-induced thyroid dysfunction.Reference Martino, Bartalena, Bogazzi and Braverman1, Reference Chiovato, Martino, Tonacchera, Santini, Lapi and Mammoli22 Paradoxically, the reduction of peripheral deiodination may result in worsening rebound thyrotoxicosis when amiodarone is withdrawn, due to an increase in T3. Amiodarone-induced thyrotoxicosis may also sometimes occur after cessation of therapy, due to the long half-lives of amiodarone and its metabolite (which can average over three months).Reference Martino, Bartalena, Bogazzi and Braverman1
Diagnosis is often delayed due to masking of the clinical features of thyrotoxicosis by the anti-adrenergic effects of amiodarone. However, even when a diagnosis is finally made, treatment is both contentious and challenging. Different authors advocate various medical management regimens, sometimes of dubious efficacy, and there is not even consensus on the importance of determining the type of amiodarone-induced thyrotoxicosis prior to medical therapy.Reference Newman, Price, Davies, Gray and Weetman23–Reference Loh25
Surgery for amiodarone-induced thyrotoxicosis is usually reserved as a last resort, even though it has been shown to be a rapid, effective and definitive management option, with no risk of recurrence in patients recommencing amiodarone therapy. There are patients who do respond to medical treatment, but in those cases refractory to treatment there is often a delay of weeks or months before surgery is considered.Reference Hamoir, Meurisse, Defechereux, Joris, Vivario and Hennen12, Reference Brennan, van Heerden and Carney15, Reference Franzese, Fan and Stack21 The potential consequences of delay are brought into sharp perspective by the fact that there is significant cardiovascular morbidity and mortality in those with thyrotoxicosis, particularly in patients with underlying cardiac disease.Reference Roffi, Cattaneo and Brandle26 It is interesting to note that, in the published reports, many patients underwent surgery as first-line treatment of amiodarone-induced thyrotoxicosis; furthermore, in those who did there was no increased morbidity or mortality.Reference Hamoir, Meurisse, Defechereux, Joris, Vivario and Hennen12, Reference Meurisse, Detroz, Messens, D'Silva, Zakaria and Joris17, Reference Meurisse, Hamoir, D'Silva, Joris and Hennen18 This may be due to there being no pre-operative delay during which cardiac status could worsen.
There is a perception that many patients are simply too ‘unfit’ for surgery to treat amiodarone-induced thyrotoxicosis, as many have uncontrolled thyrotoxicosis and cardiac failure. Our experience is in agreement with that of the published literature, that is, even in high risk cases (such as our patient with Eisenmenger's syndrome), surgery appears to be a relatively safe option when compared with no treatment or continuation of failing medical therapy.Reference Hamoir, Meurisse, Defechereux, Joris, Vivario and Hennen12, Reference Franzese, Fan and Stack21, Reference Sutherland, Robinson and Delbridge27 Based on these findings, we would urge that surgery be considered much earlier in the treatment algorithm; furthermore, we believe a debate needs to be undertaken on whether it should be first-line therapy. We would propose that surgery appears to be a safe, rapidly effective and definitive treatment option that should be used in the first instance, especially in patients who need continuing amiodarone therapy or those in whom prolonged thyrotoxicosis is particularly hazardous.
• The mechanism of amiodarone-induced thyrotoxicosis is not well understood
• Surgical treatment of this condition is often reserved as a last resort
• This report discusses the peri-operative management of a patient with coexisting Eisenmenger's syndrome
Despite the fact that the presence of Eisenmenger's syndrome is often considered sufficient grounds to declare a patient ‘unfit for anaesthesia’, the literature would indicate otherwise for non-cardiac surgery. As in the case of our patient, with substantial premedication, insertion of an arterial line for rapid blood pressure monitoring and subsequent reduced use of hypotensive anaesthetic agents, it is possible to maintain systemic vascular resistance and to reduce arterial hypoxaemia due to right to left shunting. Prophylactic measures must be (and in our patient's case were) taken against thromboembolism (patients are frequently polycythaemic), both peri-operatively (due to arterial puncture) and post-operatively (due to patients' increased risk).Reference Foster and Jones28
As long as the principles of adequate pre-operative multidisciplinary consultation are implemented, it is possible to minimise the risk of morbidity and mortality for surgical patients, whether they have amiodarone-induced thyrotoxicosis, Eisenmenger's syndrome or both.
Conclusion
Surgery should be considered the treatment of choice in patients with amiodarone-induced thyrotoxicosis, before full failure of medical therapy occurs, in order to avoid intra-operative anaesthetic complications associated with the cardiovascular complications of the advanced condition.
