Article Text

Download PDFPDF

Suspected thyroid storm in a cat anaesthetised for bilateral thyroidectomy
  1. Joanna Jane Potter1,
  2. Jennie Cook2 and
  3. Lee B Meakin3
  1. 1Veterinary Anaesthesia, University College Dublin School of Agriculture Food Science and Veterinary Medicine, Dublin, Ireland
  2. 2Milfeddygon Dolgellau Vets, Dolgellau, UK
  3. 3Small Animal Surgery, University of Bristol Faculty of Medical and Veterinary Sciences, Bristol, UK
  1. Correspondence to Dr Joanna Jane Potter; joeipotter82{at}


A 14-year-old female neutered domestic short hair cat was referred for assessment of uncontrolled hyperthyroidism. The owners were advised that a bilateral thyroidectomy was required due to the presence of comorbidities that rendered radioiodine therapy inappropriate. During the procedure the cat exhibited unexpected extreme tachycardia, hypertension, hypercapnia, severe respiratory and metabolic acidosis and ventricular arrhythmias, which all resolved after removal of the thyroid glands. Recovery was complicated by the development of congestive heart failure and hypertension which were successfully managed and the cat was discharged five weeks after presentation. Thyroid storm was considered a possible explanation for the events that occurred during anaesthesia as well as the subsequent congestive heart failure and hypertension. Thyroid storm during anaesthesia has not been reported in animals. The similarities between the reported case and thyroid storm events in humans are discussed.

  • small animals
  • anaesthesia
  • hyperthyroidism
  • cats
  • thyroid storm
  • thyroidectomy

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, an indication of whether changes were made, and the use is non-commercial. See:

View Full Text

Statistics from


Anaesthesia in cats with poorly managed hyperthyroidism/thyrotoxicosis is known to be high risk due to the often comorbid presentations of these (usually) geriatric patients and the severe cardiac, respiratory and metabolic complications that can arise.1–3 Thyrotoxicosis is defined as the clinical syndrome resulting from increased concentration of free thyroxine (T4) in the serum. One of the possible causes of elevated free T4 is hyperthyroidism, which is defined as increased serum thyroid hormone levels resulting from an overproductive thyroid gland.4 The definition of thyroid storm is based on the expression of extreme thyrotoxicosis leading to multiorgan decompensation, originally called the ‘crisis’ of hyperthyroidism.5–10 The ‘storm’ event is indicated by the acute onset of hyperthermia, central nervous system, cardiovascular and gastrointestinal or hepatic signs, along with a precipitating cause, such as surgical stimulation or other stressors like concurrent disease.7 9 11 12 However, while thyroid storm is well recognised in human medicine, including during anaesthesia,13–16 the progression and diagnosis of thyroid storm is less clear in feline medicine, indeed, some clinicians think the likelihood of true thyroid storm occurring in cats is minimal.17

Thyroid storm has not yet been described in the veterinary literature, and thus the mortality rate is not known, nevertheless, it seems to be a recognised condition and diagnostic criteria have been published.5 12 18 However, without any prior case reports or case series where the criteria have been used as a diagnostic instrument, its utility in identifying storm events must be exercised with restraint. Thyroid function tests and serum hormone analyses are often no more abnormal in storm patients than in those with uncomplicated hyperthyroidism, and therefore the diagnosis, even in humans, is based on clinical scoring systems which are unvalidated.7 11 19 The exact pathophysiology of thyroid storm has yet to be elucidated, and the diagnosis remains a clinical one. In cats and humans however, it is accepted that stabilisation of the hyperthyroid state, that is, to achieve euthyroidism, is necessary before elective surgical procedures.3 20 21 Although relatively rare, thyroid storm is documented in 10 per cent of human patients admitted for thyrotoxicosis and carries a mortality rate of 80 per cent–100 per cent if untreated.6 9 22 A variety of complications both during and post thyroidectomy are described in the veterinary literature, however thyroid storm is not listed.1 23–25 Reports of surgical mortality in cats undergoing thyroidectomy do not include thyroid storm as a possible cause of death23 even in reports where cardiac abnormalities and sudden death are reported during surgery.26 It is not known if this is because it is under-recognised or simply that any symptoms observed were not identified as a thyroid storm event. Despite this, more recent publications have listed thyroid storm as a possible complication in cats with hyperthyroidism and recommend rapid identification and treatment to improve survival.12 The authors report a case of suspected thyroid storm in a cat anaesthetised for bilateral thyroidectomy.

Case presentation

A 14-year-old female neutered domestic short hair cat weighing 3.27 kg presented for assessment of uncontrolled hyperthyroidism of 18 months duration and to determine suitability for radioiodine therapy. At the time of diagnosis with the referring veterinarian her serum total T4 was 179 nmol/l (reference interval (RI) 15–60 nmol/l), and she had been prescribed thiamazole (2.3 mg/kg twice daily orally, Thyronorm, 5 mg/ml, Norbrook, Northamptonshire UK). It was suspected that owner compliance in administering this medication was poor, or that there may have been a problem with absorption of the medication. The cat presented with a recent history of daily vomiting and diarrhoea, weight loss, polyphagia and tachycardia. She had also developed an occasional cough, occasional forelimb tremors and chronic overgrooming. She was admitted, and after a period of acclimatisation her demeanour was pleasant; however, she was resistant to handling and easily distressed. Her initial clinical exam revealed two enlarged thyroid glands, a body condition score of 3/9 and grade II/VI systolic, parasternal heart murmur, with a heart rate of 160 beats per minute and respiratory rate of 56 breaths per minute. Her systolic blood pressure recorded using the Doppler technique was 148 mmHg.


Haematology, biochemistry and urinalysis revealed changes consistent with IRIS stage 1 chronic kidney disease and the patient’s serum total T4 levels were markedly elevated (218 nmol/l, RI 15–60 nmol/l, urea 11.8 mmol/l, RI 0.5–10.5 mmol/l, creatinine 88 μmol/l, RI 133–175 μmol/l, urine protein:creatinine ratio 0.84, RI <0.4, urine-specific gravity 1.037, RI 1.020–1.040).

ECG demonstrated marked left atrial enlargement (left atrium-to-aortic ratio 2.14, RI <1.527 28) with left ventricular dilation and mild left ventricular hypertrophy consistent with hyperthyroid induced cardiomyopathy. Thoracic radiographs confirmed cardiomegaly. Abdominal ultrasonography confirmed changes consistent with chronic kidney disease and liver nodules. The sonographic appearance of both adrenal glands was unremarkable. A retinal exam revealed diffuse, multifocal, hyporeflective patches consistent with prior bullous detachment and one pinpoint area of haemorrhage, typical of hypertensive retinopathy.


It was decided that this patient was not suitable for radioiodine therapy. Before starting radioiodine, antithyroid treatments must be withdrawn, and in this case, there was a significant risk of heart failure if her T4 levels were to surge. Additionally, restricted handling protocols would make overall management of her other abnormalities challenging. Therefore, the patient was listed for a bilateral thyroidectomy, pending the normalisation of serum total T4. The patient was prescribed 3 mg/kg thiamazole (Thyronorm, 5 mg/ml, Norbrook) by mouth twice daily and oral cobalamin (Cobalaplex, Protexin Veterinary, Somerset, UK), half a capsule once daily with food.

Eight days after her admission to the hospital, the patient’s serum total T4 had normalised to 22.3 nmol/l and she was scheduled for bilateral thyroidectomy the following day. The vomiting and diarrhoea episodes also resolved during hospitalisation.

Anaesthetic management

After a thorough physical examination and detailed assessment of the history, the cat was deemed to be an American Society of Anaesthesiologists (ASA) grade 4. Methadone (0.3 mg/kg, Comfortan, 10 mg/ml, Dechra, Shropshire, UK), midazolam (0.2 mg/kg, Hypnovel, 5 mg/ml, Roche, Hertfordshire, UK) and alfaxalone (2 mg/kg, Alfaxan, 10 mg/ml, Jurox, West Sussex, UK) were administered intramuscularly for premedication to minimise stress during handling for intravenous cannula placement. An intravenous cannula was placed in the right cephalic vein, and anaesthesia was induced with 10 mg (3.1 mg/kg) alfaxalone administered slowly intravenously. Tracheal intubation was achieved using a 3.5 mm cuffed endotracheal tube after desensitising the larynx with lidocaine (Intubeaze, 20 mg/ml, Dechra). Anaesthesia was maintained with isoflurane (Isoflurane-Vet, 100 per cent w/w, Boehringer Ingelheim, Berkshire, UK), delivered to effect in oxygen using a t-piece breathing system (Infant t-piece, Intersurgical, Berkshire, UK). While the patient was prepared for surgery a fentanyl infusion (10–20 µg/kg/hour, Fentadon, 50 µg/ml, Dechra) was started, with additional boluses of fentanyl (25 µg, 7.65 µg/kg) administered before the surgical incision. The cat received intravenous isotonic crystalloids (Vetivex 11, Dechra) for the duration of the procedure, which was titrated along with the volumes of the other infusions to a total fluid volume of 3 mL/kg/hr.

The patient was monitored continuously using pulse-oximetry, end tidal capnography and agent monitoring, spirometry, ECG, heart rate, respiratory rate and direct arterial blood pressure measurements (via the left dorsal pedal artery). These were recorded at 5 minutes intervals. Oesophageal temperature was recorded intermittently. During the surgical preparation, the arterial blood pressure dropped from a mean arterial pressure (MAP) of 100 mmHg to 70 mmHg. The heart rate also decreased from 180 to 110 beats per minute, and hypotension was addressed using phenylephrine at 0.5 µg/kg/min (phenylephrine, 0.1 mg/ml, Martindale, Cumbria, UK).29 Phenylephrine was chosen due to its selective α1-adrenoceptor agonist effects, thus producing an increase in MAP without causing an increase in heart rate. Given the pre-existing hypertension, it was decided to maintain the blood pressure nearer to the patient’s preoperative levels to prevent loss of autoregulatory organ function in the face of relative hypotension; an MAP of 90 mmHg was considered appropriate, with a systolic pressure of 150 mmHg.

Physiological parameters during the surgical approach to the thyroid gland were unremarkable, except for occasional ventricular premature complexes (occurred three times in 30 minutes), but these were attributed to the cardiomyopathy that had been diagnosed. Surgical manipulation of thyroid tissue resulted in a very unexpected and rapid increase in the heart rate, blood pressure and end tidal carbon dioxide (ETCO2), without a change in the plane of anaesthesia (assessed using jaw tone, palpebral reflexes, eye position and spontaneous respiratory rate). The phenylephrine infusion was stopped immediately. The heart rate increased to 260 beats per minute (from 90 beats per minute) and increased frequency of isolated ventricular premature complexes was observed on the ECG. The MAP increased to 180 mmHg (from 95 mmHg) and the ETCO2 partial pressure increased to 61 mmHg (from 47 mmHg). There was no muscle rigidity or hyperthermia observed (oesophageal temperature was 34.6°C at the time of the incident). Arterial blood gas analysis demonstrated severe respiratory and metabolic acidosis (pH 6.95 RI 7.35–7.45, partial pressure of arterial CO2 (PaCO2) 104 mmHg RI 36–40 mmHg, PaO2 282.1 mmHg, RI: 500–600 mmHg, base excess (BE) −10.7 mEq L-1 RI −4 to +4 mEq L-1, HCO3 22.7 mmol L-1 RI 20–24 mmol L-1, lactate 0.18 mmol L-1 RI<2.5 mmol L-1). At the time of the event the end tidal isoflurane was recorded as 1.3 per cent. To ensure adequate anaesthesia and analgesia, the patient was given a bolus of 25 µg of fentanyl intravenously and the fraction of inhaled isoflurane was increased.

Mechanical ventilation was started (peak inspiratory pressure of 11 cmH2O, tidal volume of 8 mL kg-1 and respiratory rate 15 breaths per minute, positive end expiratory pressure of 3 cmH2O). The end tidal isoflurane was increased to 2 per cent with the aim of causing vasodilation and reducing blood pressure. Approximately 7 minutes after the hypertensive and tachycardic episodes had started the surgeon advised that all blood vessels to the thyroid gland had been ligated. Within 5 minutes the heart rate returned to 80 beats per minute, the MAP reduced to 85 mmHg, and the PaCO2 returned to 60 mmHg. The inhaled anaesthetic gas mixture was adjusted to return the exhaled isoflurane to 1.3 per cent. A mixed acidosis was still present; however, the pH had improved to 7.1 and the PaCO2 had decreased by over 40 per cent. While the surgeon completed the procedure the blood pressure dropped to a MAP of 70 mmHg and the phenylephrine infusion was reinstated. The total anaesthesia time was 140 minutes and surgical time was 80 minutes. Thus, the total volume of fluid administered was 22.8 mL.

Anaesthetic recovery was uneventful, apart from being markedly hypothermic (33.2°C) and tachycardic (200 beats per minute) and the cat was placed in an incubator until normothermia was achieved. Postanaesthetic monitoring consisted of invasive blood pressure and pulse rhythm, ECG, respiratory rate and rectal temperature. No further ventricular premature complexes were noted, and MAP was stable at 100 mmHg without pharmacological intervention. Fentanyl was discontinued and buprenorphine (0.02 mg kg-1 intravenously every 8 hours, Vetergesic, 0.3 mg mL-1, Ceva) was administered for postoperative analgesia.

Outcome and follow-up

Over the initial 12–24 hours postoperatively, the cat developed suspected congestive heart failure (elevated respiratory rate and effort with a gallop rhythm) which was managed with 3 mg/kg spironolactone orally, once daily (Prilactone, 10 mg tablets, Ceva). Given the reports in the human literature of congestive heart failure secondary to thyroid storm,30 it was suspected that this was the cause of the postoperative cardiac signs in the patient. During her remaining time in hospital she developed several additional complications, these included persistent hypocalcaemia, a hypertensive crisis, sick euthyroidism and hyperkalaemia with tremors and hyperaesthesia. Hypocalcaemia due to iatrogenic hypoparathyroidism is a known complication after bilateral thyroidectomy, as is sick euthyroidism.31 Hyperkalaemia in a patient with renal disease and imbalances in the renin-angiotensin-aldosterone system due to hyperthyroidism is also an accepted risk of treatment with spironolactone.

The hypertensive crisis developed three days after the procedure with systolic blood pressures recorded from 240 to 270 mmHg, measured using Doppler and oscillometric techniques. Peripheral vasoconstriction after treatment of hyperthyroidism is a known complication of thyroidectomy and antithyroid treatments,32 and it is also known that chronic antihypertensive management may be required after a thyroid storm.33 The observed hypertension caused a partial retinal detachment in the patient; however, it was successfully managed with amlodipine (Amodip 1.25 mg tablets, Ceva) and the retinas reattached with no loss of vision.

The patient was discharged five weeks after presentation with her diagnostic list indicating resolved hyperthyroidism, (currently hypothyroid), hypocalcaemia due to iatrogenic hypoparathyroidism (controlled), improved cardiac disease, chronic kidney disease (IRIS stage 1 or 2, hypertensive but with resolved proteinuria), improved enteropathy with weight gain observed in hospital and no recent episodes of vomiting or diarrhoea, hypertension (controlled) and mild anaemia (PCV 20 per cent–24 per cent in the week before discharge, RI 30 per cent–45 per cent), possibly caused by chronic disease but also may have been affected by repeated blood sampling.


Thyroid storm scoring systems and diagnosis

Hyperthyroidism in cats is commonly caused by benign adenomatous hyperplasia.1 Thyroid storm is the extreme presentation of hyperthyroidism and is a diagnosis that is made on clinical evaluation.5 7 9 12 Scoring systems to assist the diagnosis of storm events are based on clinical signs and exist in both the human and veterinary literature, although neitheris validated.7 11 12 While accepting the drawbacks of using a human scoring system and applying this to animals, and in the absence of any validated system in veterinary species, criteria supporting thyroid storm in the patient using the Burch and Wartofsky7 system included pre-existing gastrointestinal dysfunction (score 10—vomiting and diarrhoea), a precipitant history (score 10—positive diagnosis) and moderate agitation (score 10) (figure 1). The scale includes cardiac abnormalities relevant for humans, and if the authors were to apply these somewhat loosely to cats, then 10 points would be added for the changes documented on ECG. Thus, the total score achieved was 40, which in humans would highly suggest impending thyroid storm.7

Figure 1

Burch and Wartofsky scoring system for identifying thyroid storm in human patients (reprinted with permission from Elsevier).

A second scoring system for humans, the Japan Thyroid Association Definition and Diagnostic Criteria for Thyroid Storm, could also be used, and the patient would still fulfil the criteria for thyroid storm.11 This scale is more complicated and perhaps less useful in animals. This scale also requires that the patient has demonstrated thyrotoxicosis with elevated levels of free T4 or free T3 in the non-anaesthetised patient, and offers a grade of thyroid storm, either TS 1 or TS 2, according to the presence of other symptoms. In this case, the patient demonstrated elevated total T4 levels and altered mentation (overgrooming, easily distressed), along with gastrointestinal symptoms and myocardial pathology on presentation to the hospital. This would suggest a diagnostic grade of TS 1 on the Japan Thyroid Association scale at the time of presentation. TS 1 allows the clinician to diagnose a ‘definite’ thyroid storm as opposed to TS 2 which indicates a ‘suspected’ thyroid storm, but the authors accept that cats may not align exactly with human criteria. However, we did not take any samples for analysis of free T4 or free T3 at the time of the event and the utility of this scale for a patient under anaesthesia can be questioned. As such, the value of these scoring systems for cats is, understandably, debatable, however it suggests that the patient was at risk of thyrotoxic complications. If the authors were to include the diagnostic criteria for thyroid storm that is reported in the veterinary literature,5 12 then the patient would still be considered at risk for either impending or current thyroid storm; fulfilling the criteria of altered mentation, diarrhoea, vomiting and cardiac abnormalities, along with a precipitating cause, which in this case was the surgery itself. However, an article by Ward that does list clinical signs of feline thyroid storm5 was published without peer review and the strength of the claims in this article were questioned by Peterson.17 The papers that list criteria for thyroid storm in cats5 12 focus on the conscious patient, and therefore must be interpreted with caution for this case. However, even when using these diagnostic criteria, our patient still would have been considered at risk of impending or current thyroid storm when presented to the hospital. Under general anaesthesia, thyroid storm is more challenging to diagnose and is usually comprised of marked sympathetic drive, hypermetabolism, tachycardia and hypertension, all of which were demonstrated in our patient.34 35 These signs are non-specific and other possible explanations need to be excluded.

Precipitating factors for thyroid storm

In humans many events can trigger a thyroid storm, and in the past, the most common trigger was thyroid surgery itself.7 This may be due to historical lack of pretreatment with antithyroid medication, however, the exact pathophysiology of thyroid storm is not known. Thyrotoxic patients exhibit a heightened response to T4, increased free thyroid hormones and augmented binding of these hormones to their receptors.6 7 36 However, total T4 and T3 concentrations in patients during the storm event are not necessarily higher than those in patients with uncomplicated hyperthyroidism.30 Conversely, Brooks and Waldstein documented that mean free T4 fractions are significantly higher in patients with a storm event than those with hyperthyroidism, despite similar total T4 levels between the two groups.19 One possible hypothesis for this observation is that there is a decrease in the ability of the hormone to bind to the receptor; a decrease which may be caused by a variety of triggers, such as surgery, which increases the free T4 levels and leads to a thyroid storm.19 30 While a variety of triggers are postulated to explain elevations in free T4, it is not known if these occur in cats.17 In any case, since the advent of antthyroid medications, the guidelines for human anaesthesia include that, where possible, a patient’s thyrotoxicosis be controlled before surgery.20 This is also recommended in the veterinary literature.1 3 21 It was only after admission to the hospital that the cat was able to receive thiamazole consistently (Thiamazole, 3 mg/kg, orally, twice daily, increased from 2.3 mg/kg twice daily). This facilitated a reduction in serum total T4 levels from 218 nmol/l to 22.3 nmol/l over 7 days. For rapid reduction in circulating thyroid hormones presurgery in humans, doctors also advocate administering Lugol’s solution (aqueous iodine), dexamethasone and beta-blockers to attenuate the response of adrenergic receptors to circulating thyroid hormones.37 There is currently no evidence to support the use of these agents preoperatively in cats.

Differentials for thyroid storm

Other possibilities need to be excluded to make the diagnosis of thyroid storm. Phaeochromocytoma was considered unlikely because abdominal ultrasonography identified no abnormalities of the adrenal glands. Malignant hyperthermia under anaesthesia was discounted due to the low oesophageal temperatures recorded at the time of the event (34.6°C), the lack of muscle rigidity, and the very rare incidence of this condition in cats.38 In human patients, hyperthermia is a common finding in thyroid storm. The patient was anaesthetised for almost 80 minutes when the storm event occurred and hypothermia in cats during anaesthesia is a well-documented problem, even when warming interventions are employed.39 In the case, the patient was positioned on a heat pad for the duration of the surgical preparation and the procedure, and was covered with either drapes or blankets throughout the duration of anaesthesia. It is not unreasonable to assume that the small patient size and the long interval from induction of anaesthesia to the event made the patient so hypothermic that any increase in temperature may have gone undetected.

Possible concurrent infection cannot be completely excluded in this patient due to the enteropathy and the risk of bacterial translocation, and must be considered because, since the implementation of presurgical hormone stabilisation, infection is now the most common cause of thyroid storm in humans.9 The authors cannot rule out the possibility that subclinical infection may have been a contributing factor for the storm event, but the patient showed no evidence of infection. Additionally, while the authors did achieve euthyroidism before the procedure, it is not known whether a longer period of thyroid hormone stabilisation would have influenced the occurrence of a suspected thyroid storm in the patient.

Events that occurred during general anaesthesia

An analysis of the events that occurred under anaesthesia is warranted. The sudden increase in blood pressure and heart rate did not occur until 30 minutes after the first incision, and it is estimated that the whole storm event occurred over a period of less than 10 minutes with a 5 minutes recovery phase. Given the length of time from the start of the procedure, it seems unlikely that the hypertension could be due to surgical stimulation alone. Additionally, the fentanyl infusion was implemented to combat any responses to surgical stimulation, and to offer some degree of anaesthetic agent sparing which would facilitate less reliance on pharmacologically assisted blood pressure maintenance. Furthermore, phenylephrine was discontinued when the blood pressure was noted to be rising. Phenylephrine has a very short half-life and therefore the ongoing pressure rise observed in the cat cannot be attributed to this agent. Studies show that as soon as the phenylephrine infusion is stopped blood pressure decreases immediately,40 but in the patient, the pressure continued to rise making iatrogenic causes of the observed hypertension extremely unlikely. Additionally, phenylephrine would also be expected to cause a reduction in heart rate due to the baroreceptor reflex triggered by increased peripheral vascular resistance,40 not tachycardia, which was observed in this case.

Despite discontinuing phenylephrine and the continuation of fentanyl administration, the heart rate and blood pressure remained significantly above normal levels for 10 minutes without a corresponding change in anaesthetic depth. The palpebral reflexes remained absent, and jaw tone was loose; the eyes were rotated ventrally throughout the event. An increase in respiratory rate or increased muscle tension and gross movement would correspond to a ‘light’ or inadequate depth of anaesthesia or malignant hyperthermia. The patient displayed neither a change in depth or rate of respiration, nor increased muscle tone. It is not unreasonable to suggest that an increase in sympathetic tone could have resulted in the thyroid storm. However, this possibility was considered unlikely because the patient did not display any reason to have increased sympathetic tone while under anaesthesia; the cat was adequately anaesthetised and there was no evidence of a pre-existing condition that could trigger such cardiovascular changes, for example, a phaeochromocytoma.

In the human medical literature, such increases in blood pressure and heart rate in hyperthyroid patients during anaesthesia would be managed with beta-blocking agents.16 In this case, the potential need for an infusion of esmolol (Esmolol, 10 mg/ml, Hikma Farmaceutica, Fervenca, Portugal) was discussed and prepared before the procedure. However, during the event itself, it was felt that the treatments implemented (stopping phenylephrine administration, administering fentanyl and increasing the fraction of inspired isoflurane) needed to be allowed time to have an effect before considering the next possible treatment.

During the event, ETCO2 levels also increased sharply, with no change in respiratory rate or tidal volume noted on spirometry. Throughout the initial part of the anaesthetic period spirometry readings were observed and the tidal volume varied between 6 and 8 mL/kg, with a respiratory rate between 6 and 15 breaths per minute (minute volume ranging from 117 to 392 mL). It is possible that this was not an adequate minute volume for this patient, however, the ETCO2 had been consistently within normal limits before this, and occasional manual ventilation as performed when connecting the patient to the breathing system in the theatre did not reveal any sudden increase in exhaled CO2 thus making inadequate minute volume less likely. At the time of the suspected ‘storm’ an arterial blood gas analysis was performed. This demonstrated a severe respiratory and metabolic acidosis with a marked PaCO2 to ETCO2 gradient (PaCO2–ETCO2=43 mmHg in the patient, RI 2–5 mmHg). It was so severe and sudden that it was felt that this patient was in a hypermetabolic state and that this was contributing to increased arterial CO2 tension.

Ventilation-perfusion (V/Q) mismatch as a cause of the elevated PaCO2 was considered less likely because of the speed with which the change occurred, and the fact that the patient had been stable for the preceding 90 min, including demonstrating no evidence of increased exhaled CO2 after transfer from the preparation area when a few breaths were administered manually. Using the Enghoff equation41 the ratio of tidal volume to dead space in the cat was 0.4 (RI 0.3) indicating only a minor increase in dead space. Furthermore, it would be unusual to have significant regions of lung tissue ventilated but not perfused given the extreme increase in both heart rate and blood pressure. It could be argued that once the hypermetabolic state had produced such a dramatic increase in circulating CO2 the patient was unable to increase minute volume to compensate (due to being anaesthetised), and demonstrated severe hypoventilation.

If suffering from pulmonary shunt, the likely cause in the patient would be lung atelectasis, rather than consolidation, pneumonia or lung collapse given no prior history of respiratory disease. It was not possible to accurately quantify the level of shunt in this cat without the use of invasive techniques to obtain the necessary values for the calculations. However, while it is known that shunt immediately affects PaO2, the PaCO2 does not rise until the shunt value is around 50 per cent.42 Other indices of shunt and oxygen dynamics that are widely used in human medicine as surrogates for invasive measurement include the PaO2/FiO2 ratio, where a value <200 mmHg indicates a shunt of 20 per cent.43 In this study, the PaO2/FiO2 was 282.1 mmHg at the time of the event, thus the shunt in the patient was less than 20 per cent (normal shunt fraction is 5 per cent in healthy individuals). Therefore, the observed rise in PaCO2 is very unlikely to have been caused by shunt in the patient making a hypermetabolic state more likely. In any case, the PaCO2 was addressed rapidly by increasing the patient’s minute volume using intermittent positive pressure ventilation.

Diaphragmatic dysfunction and hypoventilation have also been reported in patients with thyrotoxicosis, and this may have contributed to the degree of sudden hypoventilation in the patient.44 In humans, there is a condition known as thyrotoxic periodic paralysis which may progress to hypercapnic respiratory failure and is associated with hypokalaemia.45 The patient did not display a hypokalaemia (serum potassium 3.63 mmol/l, RI 3.3–4.2 mmol/l) at the time of the event. Therefore, it was felt that the patient’s sudden increase in arterial CO2 tension was due to sudden hypermetabolism and an inability to ventilate appropriately in the face of such a sudden change.

It is possible that the hypermetabolic state that occurs in thyroid storm is more likely to be exhibited as a metabolic acidosis.46 While this was observed in this case, the patient also demonstrated a significant respiratory acidosis which would indicate a closer relationship with malignant hyperthermia.46 However, the lack of muscle rigidity at any time during the anaesthetic period as well as no increase in temperature makes malignant hyperthermia a far less likely diagnosis.46

It is important to note that the severe hypercapnia observed can explain the elevations in blood pressure observed in the cat.47 Other causes of increased CO2 such as inadequate fresh gas supply and equipment failure were excluded using capnography. At all times, the capnogram revealed a normal trace. If there is inadequate fresh gas flow, then the exhaled CO2 is not removed from the part of the breathing system attached to the ET tube. When the patient next inhales, it draws in the CO2 still present within that part of the breathing system, thus causing an increase in the fraction of inspired CO2 (FiCO2) and the trace will not reach baseline (0 mmHg) on inspiration. It is also possible that the severely elevated PaCO2 contributed to the arrhythmias noted the time of the event, however, arrhythmias caused by thyroid storm cannot be excluded, especially given that occasional ventricular premature complexes were noticed in this case intermittently from the start of the procedure and hyperthyroidism itself is associated with arrhythmias.9 48 49 However, this patient did not demonstrate arrhythmias during her ECG, nor were they detected on auscultation.

It is noteworthy that the cardiovascular and respiratory system disturbances stopped rapidly when the surgeon ligated the abnormally large vessels supplying the thyroid gland. While it has been shown that surgical stimulation does not consistently increase serum thyroid hormone levels in humans,50 it is not known if this is the case in cats, and we did not submit any blood for analysis at the time of the event. However, sources have described instances where, despite correction of serum total T4 to euthyroid levels, there remains the possibility of sudden increases in free T4 precipitated by a trigger such as surgery, although not necessarily thyroid gland manipulation, causing a saturation of plasma binding ability, thus leading to increased free T4 and thyroid storm.7 19 37 It is possible that this happened in this case. Both congestive heart failure and hypertension, sequelae associated with thyroid storm in humans, occurred in this cat. One could argue that the anaesthetist should have started beta-blocker treatment at the time of the incident and continued this for the immediate postoperative period. Given that the half-life of T4 in cats is 10 hours, it is possible that subsequent development of heart failure was due to lack of appropriate medical control being initiated during the storm event and ensuing hours; in the human literature it is suggested that treatment is continued until the half-life of T4 is exceeded.51 Indeed, one reference advocates the use of beta-blockers in hyperthyroid cats before stressful events and also for acute exacerbations such as thyroid storm.5 However, beta-blocker administration is not benign, with reported side effects including bronchospasm and excessive bradycardia and hypotension during anaesthesia.52 In this case, during the suspected storm event, other steps were being taken to correct the observed physiology. However, it is felt that these took longer than desired and should the situation occur again, the anaesthetist would consider beta-blocker therapy immediately. It is important to note the difference between the risk of presurgical treatment with an oral beta-blocker in a cat with structural cardiac changes versus the treatment of an acute storm event, during which there is markedly increased cardiac workload, with very short acting beta-blockers administered intravenously. It is felt that the latter would have been an appropriate treatment choice and may have improved the immediate postoperative outcome.

Other considerations

Presurgery treatment of this patient with beta-adrenergic blocking agents may have alleviated some of the risks of developing thyroid storm, as has been shown in human patients.37 However, due to the severity of the structural cardiac changes this was not appropriate for the patient, and any sudden reduction in blood pressure may have worsened the pre-existing kidney disease before surgery was attempted. It could also be argued that blood pressure management may have been started before surgery while the cat was being rendered euthyroid. This was not done because blood pressure measurements during the presurgery hospitalisation did not indicate that it was necessary and any disruption to the cat’s current cardiovascular status was considered high risk given the severity of ECG changes observed.

It could also be suggested that this patient had severe thyrotoxicosis and the events that occurred were not a true ‘storm’ event, especially since the diagnosis is clinical and has no definitive diagnostic features described in any veterinary species. However, this cat was euthyroid before the procedure, and the events observed were sudden in onset, from baseline parameters that would be considered normal for anaesthetised patients, making a ‘storm’ a more likely diagnosis than severe thyrotoxicosis. In any case, regardless of the classification of the event, the fact remains that the hypertension, tachycardia and hypermetabolic state resolved rapidly after ligation of the vessels supplying the abnormal thyroid tissue, supporting a potential role for thyroid gland manipulation causing these storm-like signs in cats. The subsequent development of congestive heart failure and hypertension also increase the likelihood that the events observed during anaesthesia were a thyroid storm.

Learning points

To the authors’ knowledge, this is the first report of a suspected thyroid storm occurring in a cat during general anaesthesia. This patient demonstrated many symptoms that are consistent with thyroid storm reports in humans, including a precipitating event with dramatic, sudden and severe tachycardia and hypertension and a hypermetabolic state arising from what would be considered normal baseline parameters with subsequent clinical manifestations of congestive heart failure after the event. Diagnosis of thyroid storm is based on clinical signs, and therefore it is reasonable to categorise the events in this patient as a thyroid storm. Other possible causes of the events have been analysed and considered not to offer adequate explanations of the events that occurred. Learning points from this case include:

  • The necessity for vigilant monitoring of cardiopulmonary parameters in hyperthyroid cats under general anaesthesia so that storm events can be recognised, and noting that the risk of thyroid storm still exists even when cats are euthyroid and deemed to be otherwise stable.

  • The realisation that thyroid storm is a possible complication even in a euthyroid patient, and the acceptance of this into the wider body of veterinary literature for anaesthetising hyperthyroid cats so that it can be managed appropriately, as well as recognising that a storm event may be a cause of death if untreated.

  • The potential utility of the available human scoring systems to identify patients at risk of thyroid storm, and comparing this with the published veterinary diagnostic criteria, to create a more robust method of identification of at-risk veterinary patients.


The authors would like to thank Dr Angie Hibbert, the primary clinician, for the contribution to this case.


View Abstract


  • Contributors Article written by JJP and JC, surgeon involved with case and manuscript preparation was LBM and primary clinician involved with the case management was AH.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement There are no data in this work.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.