Preoperative thyrotoxicosis is a potentially life-threatening condition that requires medical intervention before surgery. The positive ionotropic and chronotropic effects of thyroid hormone on the heart culminate in an increase in cardiac output and arrhythmias. Elective surgeries should be postponed in patients with overt hyperthyroidism because of the risk of precipitating thyroid storm (TS) and potential cardiac complications. Although antithyroid drugs (ATDs) are the preferred initial treatment, clinical response may take weeks. There are potential limitations of ATDs in managing complex cases of thyrotoxicosis, such as in cases of ATD intolerance, where patients are unable to tolerate ATDs due to adverse effects such as agranulocytosis or hepatotoxicity, or ATD resistance (1–3), where there is failure to achieve or maintain euthyroidism despite adequate dosing and compliance with ATD therapy. Studies have reported the use of combination pharmacotherapy or therapeutic plasma exchange (TPE) to rapidly achieve clinical and biochemical euthyroidism in patients who require rapid optimization of thyrotoxicosis.

Second line pharmacotherapy in the management of thyrotoxicosis include the following: 1. Corticosteroids inhibit the peripheral conversion of thyroxine (T4) to the more active triiodothyronine (T3), helping to reduce the levels of T3, which is primarily responsible for the hypermetabolic symptoms of thyrotoxicosis (4). 2. Lithium inhibits the synthesis and release of thyroid hormones from the thyroid gland (5). 3. Cholestyramine binds thyroid hormone in the intestine and facilitates fecal excretion of thyroid hormone, thereby reducing enterohepatic recirculation and serum thyroid hormone levels (6). 4. Beta-blockers primarily mitigate the peripheral effects of excess thyroid hormone on the cardiovascular system. High dose Propranolol has been shown to inhibit peripheral conversion of T4 to T3, which further contributes to the amelioration of clinical symptoms and signs of thyrotoxicosis (7). Existing guidelines suggest initiation of beta-blockers when heart rate exceeds > 90 to aim for a normal pulse rate (8).

TPE, a class 2C recommendation for the treatment of TS by American society for apheresis (9), involves the extracorporeal removal of plasma containing excessive thyroid hormones, which are then replaced with donor plasma or albumin. The effectiveness of TPE in reducing thyroid hormone levels has been demonstrated in various studies, showing significant decreases in thyroid hormone levels post-treatment (10, 11).

For the study, 12,876 records were retrieved from MEDLINE (3335), Google Scholar (1010), Embase (8327) and CENTRAL (Cochrane Central Register of Controlled Trials) (204). The study flow diagram is shown in Figure 1 . The complete list of articles obtained through the systematic search was screened to remove duplicates and exclude ineligible articles. Seventy-six studies were selected from title and abstract screening and retrieved for full-text screening. Sixty studies were deemed ineligible and 16 studies were included in the systematic review. Study quality was assessed using NOS; two studies obtained 7 points, twelve obtained 6 points, and the remaining 2 obtained 5 points ( Table 1 ).

The current recommended goal of therapy in patients with hyperthyroidism is to render patients as close to clinical and biochemical euthyroidism prior to surgery, and current guidelines propose the use of ATDs as first line therapy in doing so. While most patients tolerate and respond well to ATDs, the development of serious complications precludes its use in certain scenarios. Severe hepatotoxicity and neutropenia are widely regarded as contraindications to ATDs in view of the potential devastating consequences. Other cited reasons for ATD intolerance include adverse events of vasculitis, severe cutaneous reactions, allergic reactions, arthralgia and drug-drug interaction with warfarin. Concerns of safe ATD usage in patients with pre-existing hepatic dysfunction have also prompted the use of 2^nd line pharmacotherapy in some studies (20).

ATDs may not be sufficient for patients who require rapid control of hyperthyroidism prior to surgery, and this group would benefit from combination pharmacotherapy or TPE to adequately prepare them for surgery. In clinical practice, inadequate control of hyperthyroidism is often the result of suboptimal adherence to ATDs. True ATD resistance as the cause of inadequate control of hyperthyroidism is rare and largely limited to case reports (25). There is no clear definition of ATD resistance, but reports have suggested that this diagnosis can be made in patients whose thyroid hormone levels remain elevated despite compliance to high doses of ATD (26) or when ATD doses exceeding that of standard dose ranges (up to 30mg/day for MMI and 300mg/day for PTU) are required to achieve euthyroidism (27). Risk factors for ATD resistance include large goitres, iodine excretion of ≥ 100 µg/g creatinine, high pretreatment thyroid hormone levels and presence of elevated TSH receptor antibody levels (28). Whilst exact mechanisms behind resistance to ATDs are poorly understood, possible aetiologies include malabsorption, rapid metabolism of the drug, anti-drug antibodies and defects in intrathyroidal accumulation or action (1). Elevated levels of MDR-1 multi-drug resistance gene may also be associated with disease activity and treatment resistance in patients with Graves disease (29). In patients with ATD resistance, doses of up to 150mg of MMI have been prescribed in combination with lithium, steroids and inorganic iodine prior to definitive therapy (26).

It is precisely the above group of patients who require definitive treatment of their hyperthyroidism due to intolerance to or inadequate treatment with ATDs. Total thyroidectomy would be the procedure of choice in certain patient populations (30). It is also not uncommon that patients with thyrotoxicosis require surgery for an unrelated condition. There is currently no risk stratification tool to predict the risk of complications such as thyroid storm, cardiovascular morbidity and mortality in hyperthyroid patients undergoing surgery. Patients who had a diagnosis of hyperthyroidism, presence of antithyroglobulin antibody and high thyroglobulin levels (>150 ng/mL) were identified to be at higher risk of surgical complications such as damage to the recurrent laryngeal nerve, damage to the parathyroid glands, and postoperative hematoma (31). However, thyroid storm, cardiovascular morbidity and mortality were not analysed in this paper. In a study done in a cohort of older orthopaedic patients, postoperative complications (e.g. cardiovascular events, cardiac failure, infections, stroke, and neurological complications) appeared to be higher in patients who were biochemically hyperthyroid, although incidence of thyroid storm was not reported (32).

Notably, there are no specific pre-operative biochemical cut-offs (8, 33) or clinical parameters that can reliably predict the prevention of thyroid storm. If urgent surgery is indicated, achieving a normal thyroid function before surgery may not be feasible. TSH values may remain suppressed even in patients with normalized T4 and T3 levels. Nonetheless, pre-operative treatment can be undertaken to minimise the risks of perioperative complications from thyrotoxicosis as evident from the studies included in this systematic review. Following pre-operative treatment, there was a mean percentage reduction of fT4 and fT3 by 52.6 ± 8.2% and 68.1 ± 9.3% respectively in the included studies ( Table 3 ) and there were no events of perioperative thyroid storm reported despite a lack of normalization of thyroid hormone levels in some patients prior to thyroidectomy. Studies by Ali et al. and Builes-Montaño et al. did not report any events of perioperative thyroid storm despite just above 40% and 50% achieving normal fT4 and fT3 respectively prior to thyroidectomy. Keklik et al. reported a mortality from myocardial infarction complicated by tachyarrhythmia. This patient was older (60 years vs. median age of 47 years in the study) and had higher pre TPE thyroid hormone levels (pre-TPE fT4 75.2pmol/L vs. 44.2pmol/L and pre-TPE fT3 22.3pmol/L vs. 20.4pmol/L); though there were no details provided on patient’s baseline cardiac status and events surrounding the patient’s demise. Whilst Calissendorff et al. suggested a heart rate target of < 80 prior to thyroidectomy, optimal heart rate targets are uncertain. Elevated preoperative resting heart rate (> 87 beats/min) appeared to significantly correlate to postoperative myocardial impairment and mortality in patients who underwent non-cardiac surgery, although this was not specifically studied in a thyrotoxic population going for thyroidectomy (34). Existing guidelines suggest initiation of beta-blockers when heart rate exceeds > 90 to aim for a normal pulse rate (8) and higher rates of intraoperative tachycardia may be seen if normal pulse rates were not achieved preoperatively (35).

While guidelines suggest for normalization of thyroid function prior to surgery, this may not be viable for patients with time-critical surgeries. This could also lead to delays in patients obtaining surgical treatment. In our review, patients with hyperthyroidism who were treated with second-line pharmacotherapy before surgery were able to undergo the procedure safely, even though their thyroid hormone levels did not normalize. Second line pharmacotherapy used as an alternative to or combination with ATD was safe in the studies included in this review. Side effects of rash, vomiting and finger swelling experienced with iodine were reversible with either dose reduction or stopping the offending medication. Three of 10 patients who received dexamethasone at 2mg/day for 10-14 days developed secondary adrenal insufficiency; all demonstrated normalization of adrenal function after 3-6 months. TPE was also an effective treatment option for pre-operative hyperthyroidism, achieving a mean percentage reduction in fT4 and fT3 levels of 51.7 ± 8.7% and 67.7 ± 10.3% respectively in the papers included in our review. Adverse effects of TPE in the studies were mostly mild: pruritus, rash, prolonged prothrombin time with no clinical bleeding manifestations and transient hypotension were reported. One patient suffered anaphylaxis to FFP, 2 had catheter-related infections, and 1 developed deep vein thrombosis. Based on the latest edition of American Society for Apheresis (ASFA) guidelines (9), TPE is currently indicated for treatment of thyroid storm in patients who respond poorly or have intolerance to first-line therapies (Category II, Grade 2C Evidence). We may consider broadening the criteria for using TPE to include patients who are not in thyroid storm, but require optimization of preoperative hyperthyroidism.

The NOS is a tool designed to assess the reliability and validity of findings reported in non-randomized studies included in systematic reviews. It gauges various aspects of study quality, including the selection of participants, the comparability of groups, and the assessment of outcomes. The studies included in our review generally demonstrated strong performance in the selection and outcome assessment components of the NOS. The cohorts were somewhat representative of the population of patients requiring perioperative management of thyrotoxicosis, though some studies were on patients with more severe disease requiring second-line pharmacotherapy and/or TPE. Exposure assessment was adequately conducted and outcomes were monitored across the perioperative period with complete follow up for all subjects. However, only three studies included a comparative non-exposed cohort, and none of these studies controlled for confounding factors such as thyroid hormone levels or comorbidities that could affect operative outcomes.

There are several limitations of this systematic review. Firstly, studies included in this review were either case-series or retrospective cohort studies which were largely descriptive in nature. Patient numbers are too few to extrapolate outcomes to a larger population and hence may not be generalisable. Secondly, due to heterogeneity of the studies and lack of head-to-head comparison studies, we were unable to ascertain which 2nd-line treatment option or combination was superior. Individual patient results were often not available and there were multiple variables that could have confounded outcomes. Thirdly, as studies included in this review were from various centres across the world with different reference ranges for thyroid hormone levels, we could only report the percentage change in thyroid hormone levels rather than absolute values. Thus, we were unable to identify any specific biochemical cut-offs. Further prospective studies are required to help risk-stratify patients who are at high risk of developing complications post operatively and evaluate safe threshold levels of thyroid hormone levels prior to thyroidectomy. Upcoming potential treatment options being studied including antigen specific immunotherapy e.g. ATX-GD-59 (36), monoclonal antibodies e.g. Iscalimab and TSH receptor-Blocking Antibodies e.g. IMVT-1402 could potentially be investigated as perioperative adjunctive treatment in the future (37, 38).

There is a paucity of data regarding optimal therapy when conventional treatment with ATDs is not tolerated, is ineffective or when rapid optimisation is required for urgent surgery. Patients with hyperthyroidism should be treated to reduce the risk of thyroid storm and cardiac complications, although the most optimal 2^nd line regime is still uncertain and there is no available guidance in terms of biochemical cut-offs nor clinical parameters that are currently suggested in guidelines. Second line pharmacotherapy and/or TPE in this review were shown to be effective at reducing thyroid hormone levels to allow patients to safely proceed for surgery. More studies are needed to ascertain the optimal regime of 2^nd line therapy, develop risk assessment tools for patients at higher risk of post-operative thyroid storm and determine biochemical and clinical parameter cut-offs.