Systemic mastocytosis presenting as profound cardiovascular collapse during anaesthesia


Dr S. T. A. Vaughan Specialist Registrar in Anaesthesia, Manchester Royal Infirmary, Oxford Road, Manchester M13 9WL, UK


Mastocytosis is a rare disorder with serious anaesthetic implications. Anaesthetic management is hazardous since trauma, stress, extremes of temperature and drugs may precipitate intra-operative mast cell degranulation. Release of histamine and other mast cell mediators can lead to profound cardiovascular collapse and even death. We present a case report of a patient with mastocytosis who suffered cardiac arrest during anaesthesia. Anaphylactoid/anaphylactic shock may be delayed and lack supporting signs of histamine release such as cutaneous flushing and bronchospasm.

Systemic mastocytosis is a rare disorder occurring within a spectrum of mast cell diseases which include cutaneous mastocytosis (urticaria pigmentosa). The incidence of cutaneous mastocytosis is estimated to be between 1 in 1000 and 1 in 8000 of the population [1], of which 10% will have systemic involvement [2]. Diagnosis is often difficult and delayed due to the variety of presenting symptoms [3[4]–5], the most severe being anaphylactic shock and death [6]. This is the first report of profound cardiovascular collapse during anaesthesia in a patient with mastocytosis who did not divulge his condition pre-operatively to the anaesthetist.

Case history

A 43-year-old male van driver weighing 85 kg presented to the Accident and Emergency department having sustained a comminuted fracture to the left olecranon following a fall of 4 feet (1.3 m) from the tailgate of his van. On admission and at the pre-operative visit he denied any ill health, current treatment or allergies. He had undergone an uneventful anaesthetic 20 years previously for manipulation of a fractured left distal radius. Physical examination was unremarkable except for some brown maculopapular lesions over the thorax and upper abdomen having the appearance of benign naevi. The patient was not a local resident, so old hospital notes and anaesthetic charts were unavailable for review.

A decision was made to proceed to internal fixation of the fracture under general anaesthesia. No premedication was prescribed and the patient arrived in the operating theatre 27 h after the injury. Monitoring of ECG, pulse oximetry (SpO2) and noninvasive blood pressure was started before induction of anaesthesia. A 20 gauge intravenous cannula was sited in the dorsum of the right hand and, following pre-oxygenation, anaesthesia was induced with propofol (160 mg) and fentanyl (0.1 mg). A size 4 laryngeal mask airway was inserted and connected to a capnograph and a Lack breathing system. Anaesthesia was maintained with 70% nitrous oxide in oxygen and isoflurane 1.5%. An infusion of normal saline was started and 1 mg of morphine was given intravenously. The patient was transferred to the operating theatre where the SpO2 was noted to have decreased from 98% to 94%. Heart rate, blood pressure and end-tidal CO2 were unchanged and, as the SpO2 was corrected by temporarily increasing the inspired oxygen (FIO2) to 50%, the desaturation was attributed to a fall in the functional residual capacity in a supine anaesthetised patient.

An upper limb tourniquet was applied to the left arm and a further 2 mg of morphine was given intravenously before inflation. On starting surgery the heart rate was 60 beat.min−1, blood pressure 120/65 mmHg and SpO2 95% with an FIO2 of 35%. During the next 20 min, morphine 7 mg was administered in divided doses and this was accompanied by a gradual decline in SpO2 requiring a progressive increase in the FIO2. Twenty-five minutes after the start of surgery the SpO2 was 92%, the FIO2 100%, the pulse rate 105 beat.min−1 and the blood pressure 120/45 mmHg. Repeated examination of the patient failed to reveal evidence of cutaneous flush, oedema, bronchospasm or pneumothorax. End-tidal CO2 remained relatively constant over this time between 7 and 7.6 kPa.

Surgery was stopped, the trachea intubated after a 40 mg bolus of atracurium and controlled ventilation commenced. There was no laryngeal soiling at laryngoscopy. Immediately following intubation and ventilation with 100% O2, the patient progressed into electromechanical dissociation (EMD). External cardiac massage, intravenous adrenaline 1 mg and a rapid infusion of 500 ml of colloid were started in accordance with the European Resuscitation Council guidelines [7].

Tension pneumothorax and cardiac tamponade were excluded as a cause of EMD by bilateral intercostal chest drainage and pericardiocentesis. Since there was no reason to suspect electrolyte abnormalities, a massive pulmonary embolus was considered to be the most likely precipitating event at that time.

Resuscitation continued for 75 min during which time central venous access and a femoral arterial line were secured. Large volumes of crystalloid (2000 ml) and colloid (2500 ml) were administered, but the patient was heavily dependent upon adrenaline, requiring six 1-mg boluses and an infusion of 1 mg.h−1 to maintain a systolic blood pressure of 70–80 mmHg. At no time did the SpO2 rise above 85%. The patient eventually developed a pulseless idioventricular rhythm that was considered to be premorbid, so thoracotomy and internal cardiac massage were performed. This resulted in an improvement in cardiac output and rhythm, but was complicated by two episodes of ventricular fibrillation. Internal defibrillation paddles were not immediately available, so external 200-J defibrillation was performed which restored sinus rhythm.

The adrenaline infusion was increased to 2 mg.h−1 until the systolic blood pressure stabilised at around 100 mmHg. The patient was transferred to the Intensive Care Unit and underwent urgent pulmonary angiography, but this was found to be normal. A bleeding diathesis subsequently developed with an activated partial thromboplastin time in excess of 300 s and prolonged prothrombin and thrombin times. An anaphylactic reaction had been considered earlier in the course of the patient's resuscitation but lack of clinical evidence had encouraged us to reject the diagnosis. In view of the lack of positive findings for the cardiovascular collapse and the coagulopathy, an anaphylactic reaction was now reconsidered and the patient treated with steroids and antihistamines. Blood samples for plasma tryptase estimation were collected over a 24-h period.

The patient developed acute renal failure, acute respiratory distress syndrome and never regained consciousness. He died 4 weeks after admission.

Unbeknown to us, at the time of administration of anaesthesia the patient had a 15-year history of flushing attacks and headache precipitated by ingestion of alcohol and hot baths. A maculopapular rash had also developed over the preceding 10 years. Two years previously he had seen a dermatologist at another hospital who had diagnosed cutaneous mastocytosis from these symptoms and a skin biopsy. Examination of the bone marrow had, however, been normal. More recently the patient had developed new symptoms of abdominal pain, diarrhoea and palpitations and had been referred for further investigations, but failed to attend.

The previous diagnosis of cutaneous mastocytosis was obtained the day after admission to the intensive care unit from the patient's general practitioner. Systemic mastocytosis was diagnosed from post-mortem findings of increased mast cell aggregates in the bone marrow, lymph nodes, skin and in the liver. Analysis of the plasma tryptase levels revealed a raised level of 18−1 (normal values < 1−1). Extrapolation back to the initial time of the cardiovascular collapse would have given a value in excess of 20−1 (personal communication, Dr J. Watkins, Sheffield) which would be consistent with massive mast cell degranulation. Plasma IgE levels were in the normal range, implying that the elevated tryptase levels were a result of an anaphylactoid rather than a type 1 hypersensitivity (anaphylactic) reaction.


Mastocytosis was first described over 100 years ago [8]. The term describes a spectrum of diseases from the cutaneous variety (urticaria pigmentosa) to the systemic form of the disease. An urticarial rash is frequently seen in systemic mastocytosis while 10% of patients with cutaneous mastocytosis will have systemic involvement. Mastocytosis is caused by mast cell hyperplasia in response to overproduction of cytokines [1, 9]. Mast cell aggregates may be found in a variety of tissues including skin, bone marrow, liver, spleen, lymph nodes and the gastrointestinal tract [1, 10]. The systemic variety may present as episodic flushing, pruritis, headache, palpitations, abdominal pain and diarrhoea [1, 3[4]–5, 10]. Unpredictable episodes of profound hypotension are a recognised feature of the disease and are due to release of mast cell mediators including histamine [11] and prostaglandin D2 [12]. Heparin release may lead to haemorrhagic manifestations [13]. Attacks may be precipitated by toxins, venoms or drugs (acetylsalicylic acid, alcohol, morphine, codeine, d-tubocurarine, gallamine, reserpine, quinine, iodine contrast media) [1, 14]. Physical factors such as stress, trauma and extremes of temperature may also degranulate the mast cell. Tryptase released from the mast cell is a reliable serum marker of degranulation in mastocytosis and anaphylaxis [15].

The plasma tryptase levels in our patient, although elevated, are not as high as reported values in bee sting anaphylaxis [16]. The lower values in this case can be explained on two counts. Firstly, the initial blood sample was only collected 1 h after admission to the Intensive Care Unit and 4 h after the initial cardiovascular collapse. Peak tryptase values occur within 1 h of mast cell degranulation [16], implying that peak values in our patient would have been well in excess of 20−1. Second, a G5 radioimmunoassay was used to analyse the tryptase levels in this patient. The G5 assay uses a monoclonal antibody raised to an epitope on denatured tryptase. As a result it does not bind well to natural tryptase in serum and has a sensitivity of around 1−1. A new B12 assay, using mouse monoclonal antibodies raised to natural plasma tryptase, has been developed which is 10 times more sensitive than the G5 assay [17], so in reality the levels may have been in excess of 100−1. The plasma tryptase level in this patient only confirms that mast cell degranulation took place; the degree of degranulation can only be inferred from the severity of the cardiovascular collapse.

The treatment of mastocytosis is aimed at stabilising the mast cell membrane or antagonising the effects of released mediators. Antihistamines remain the mainstay of therapy and, while the use of sodium cromoglycate confers no benefit when compared to antihistamines, the frequency of gastrointestinal symptoms may be reduced [10, 18, 19]. The vasodilator effects of histamine are mediated through H1 and H2 receptors [20, 21]. H1 blockade alone will reduce the frequency of attacks but H2 blockade may be required to reduce their duration. Some patients remain symptomatic, despite dual therapy, due to the release of prostaglandin D2. Aspirin, a cyclo-oxygenase inhibitor, has been shown to be of benefit in these patients [12, 22], but treatment should be started cautiously as aspirin itself may precipitate mast cell degranulation.

In acute hypotensive episodes, systemic antihistamines and steroids are of little benefit [22, 23]. The drug of choice in anaphylaxis is adrenaline [24] which, apart from its general cardiovascular actions, may also act at mast cell β2 adrenoceptors to prevent degranulation [22, 23].

Anaesthesia is particularly hazardous for patients with mastocytosis as they are exposed to a stressful situation while, at the same time, specific drugs are administered which may provoke cardiovascular collapse. The rash alone is frequently mistaken for freckles and a spot diagnosis is difficult [11]. It is therefore vital that the anaesthetist is made aware of the diagnosis well in advance of the administration of an anaesthetic. Prior treatment with antihistamines and sodium cromoglycate can result in uneventful anaesthesia [25[26][27]–28], but profound intra-operative hypotension with flushing and bronchospasm has been documented despite pretreatment [29].

In a similar case, a patient known to have cutaneous mastocytosis developed a hypotensive crisis with flushing 90 min after intravenous morphine [30]. The administration of morphine was the most likely cause of a similarly delayed cardiovascular collapse in our patient (personal communication, Dr J. Watkins, Sheffield). The atracurium may have contributed to the collapse but it is unlikely since the patient had begun to deteriorate long before it was given. Interestingly, clinical examination did not reveal flushing, oedema or bronchospasm. These symptoms, while common in mastocytosis [4, 5], are not a universal feature of anaphylaxis [23, 31], making the diagnosis particularly difficult during anaesthesia. The management of this case included rapid intervention with adrenaline, and it is doubtful whether earlier therapy with steroids and antihistamines would have altered the outcome.

There is very little literature on regional anaesthesia and mastocytosis. It would seem reasonable to assume that a regional technique would be a safer alternative to general anaesthesia as it is easier to avoid potential histamine-releasing drugs. A cholecystectomy has been successfully performed under epidural blockade in an adequately premedicated patient [27], although there is a report of a severe urticarial reaction following a Bier's block in a patient with mastocytosis who only received an anxiolytic premedicant [32]. This illustrates the need for adequate pre-operative preparation in those patients undergoing a regional technique, as mast cell degranulation can be precipitated by stress, friction, tourniquets and extremes of temperature.

Mastocytosis is invariably diagnosed by dermatologists or physicians and, while information is readily available in anaesthetic texts [14, 33, 34], few general medical textbooks mention the anaesthetic risks of systemic mastocytosis. In our opinion this information should be more widely available to specialities outside anaesthesia, including general practice, and these patients should be strongly encouraged to inform medical staff of their diagnosis. In addition, possession of a warning card or Medic Alert bracelet may help to avoid a life-threatening situation.