Anaphylaxis and angioedema

Angioedema is "abrupt nonpitting swelling of the skin, mucous membranes, or both". It is essentially the same as urticaria, even though it can occur in the absence of urticaria (eg. in the context of ACE inhibitor therapy or hereditary C1 esterase deficiency). Anaphylaxis is a severe systemic allergic reaction, which usually features angioedema and urticaria. The natural course of this reaction is a progression to shock, airway obstruction and death. 

Question 17 from the second paper of 2006 presented the candidates with the story of a nine year old boy who developed hypotension and rash thirty minutes into his facial reconstruction surgery. The age was unimportant (all paeediatric uniqueness of this case was ignored completely in the college answer), hence this SAQ being shoved into the "cardiac arrest and resuscitation" category. 

Probably the best source for this sort of this is the ARC website, specifically Guideline 9.2.7 (Anaphylaxis). That is a good guideline for first responders. An "Advanced" anaphylaxis recommendation package aimed at emergency department staff and prehospital carers is also offered by the ASCIA people (2016 Guidelines). With greatest relevance to Question 17 from the second paper of 2006, ANZCA have a guideline for the management of perioperative anaphylaxis. The latter serves as an excellent jump-off point to read about this topic- not only is it relevant to the SAQ, but it also happens to be a guidelines statement by a major local institution, and their reference bibliography has links to full-text articles.

Classification

These conditions can be separated loosely into categories which help you decide on management. One such classification is offered by Moellman et al (2014).

• Anaphylaxis with or without angioedema, which makes you reach for the adrenaline
• Histaminergic angioedema without anaphylaxis, which makes you grab the antihistamines
• Non-histaminergic angioedema which stimulates phone calls to immunology and which is managed with some combination of FFP, tranexamic acid and icatibant instead of adrenaline or steroids.

Clinical features

Presenting symptoms and signs in the awake patient may consist of the following:

• Stridor
• Abnormal speech
• Swelling of tongue
• Hoarse voice
• Wheeze
• Vomiting
• Abdominal pain

Specific findings on examination:

• Tachycardia and hypotension with hyperdynamic circulation
• Flushed appearance
• Urticarial rash
• Swelling of the lips or face

In the intubated ICU or anaesthetised theatre patient, the following features might alert you to the possibility that a drug you just gave is causing an anaphylactic reaction:

• Hypotension
• Tachycardia
• Widened pulse pressure
• Failure to respond to routine doses of metaraminol
• Suddenly decreased EtCO₂ (suggesting decreased cardiac output)
• Sudden change of EtCO₂ waveform, suggesting bronchospasm
• Sudden increase in the peak airway pressure

Management of anaphylaxis

Intramuscular adrenaline

The dose of adrenaline quoted by www. allergy.com.au is 10 mcg/kg IM.

For those unable to titrate 10-mcg dose increments, the guidelines recommend 0.5mg (i.e. one half of a standard 1:1000 ampoule) for an adult, or 0.1mg for a child under the age of 1.   This is one of the few conditions where adrenaline is indicated for IM administration; the specific reason being the extreme vasodilation of muscular beds and the excellent absorption of IM medications due to this enhanced vascularity. Conversely, exploring for a vein or loading the IO gun might be a waste of precious minutes. This intramuscular dose may be repeated every 5 minutes, but should theoretically provide up to 40 minutes of infusion-like effect (Simons et al, 2001)

The recommendation for IM adrenaline is limited to the periarrest setting. The anaphylaxis patient in PEA is not perfusing their luxuriously vasodilated muscles, and all those adrenaline depots in their muscles are going to go to waste. IV adrenaline is indicated if the patient is in cardiac arrest.

Intravenous adrenaline

Wherever decent monitoring is available and IV access is established, an adrenaline infusion is preferable to repeated IM dosing. One can see the merit of this if one is in some sort of nightmarish quintessentially Australian prehospital setting, where the retrieval helicopter is three hours away. A five-minute IM administration cycle will have you stab the patient sixty times before they arrive. Ergo, infusion. The French guidelines recommend 0.05-0.1 mcg/kg/min, but realistically you're going to titrate it to effect.

Vigorous fluid resuscitation

The way these people progress to cardiac arrest is by rapidly leaking their whole circulating volume out of their capillaries. Within the first 10-15 minutes, up to 70% of the circulating volume may end up extravascular (Clarke et al, 2011) With greatly diminished preload there is eventually nothing to pump, and a tachycardic PEA ensues. 

In their guidelines statement, ANZCA recommend the use of colloid in resuscitating the patient with anaphylaxis. This is repeated in the college model answer to Question 17 from the second paper of 2006. The references used by ANZCA to support this (Grade D) recommendation consist of Mertes et al (2011) and Tajima et al (2013). The first is a guidelines statement by the EAACI Interest Group on Drug Allergy which recommends that colloids be substituted for crystalloids after a dose of 30ml/kg has been exceeded, but gives neither a reference for this, nor even a grade of recommendation. The second is a study by a French group, exploring the haemodynamic effects of different volume expanders in a rat model of anaphylaxis.  The range of volume expanders included isotonic saline, hypertonic saline and hydroxyethyl starch. Starch helped the investigators reach their haemodynamic endpoint faster than saline, by up to 50%. However, use of starch is now forbidden in Australian humans. Apart from the other usual reasons for why this study lacks external validity (my patients are not rats, etc) is experimental methodology: anaphylaxis in these rats was actually induced by a colloid infusion (ovine albumin).

In short, the recommendation for the use of colloid in perioperative anaphylaxis rests on a poorly referenced European opinion statement and a rat study.  Opponents of colloids would go on to comment that the use of colloid more generally is not based on evidence that is any better than that. Supporters of colloids would point to the reasonably sound physiological basis supporting the use of colloid (theoretically, colloids should oppose the transcapillary escape of fluid in leaky anaphylaxis). The pragmatic CICM fellowship exam candidate will write about colloid in their answer, because it is the politically correct thing to do - the people who wrote these guidelines are likely going to be the college examiners and their friends.

Intravenous glucagon

This is mentioned in numerous guidelines statements as the management of refractory shock in those anaphylactic patients who are under the influence of beta-blockers. These patients may end up refractory to adrenaline. In other words, they should be managed as a beta-blocker overdose.

Nebulised adrenaline

If the patient happens to be not intubated, their airway will be threatened. One of the ways of dealing with this is to offer nebulised adrenaline. The dose is 5mg as a neb, i.e. 5×1ml ampoules of 1:1000.

Corticosteroids

Everybody always gives hydrocortisone or prednisolone to these people. However, there is no evidence to support this practice. A Cochrane review (Choo et al, 2013) could not find any studies which met eligibility criteria to undergo meta-analysis. 

Antihistamines

For some reason everybody seems to give promethazine to these people. Correct, those who are conscious will occasionally complain of being itchy. This is of little interest to us. They first need to survive, so that in some hypothetical future they may go on to scratch themselves. Promethazine (phenergan) will play no role in improving the concurrent anaphylaxis situation, and will do nothing to prevent future episodes of anaphylaxis. In spite of this, Alrasbi et al (2007) found numerous international guidelines (Canadian, Russian, British, American) which recommmended the use of antihistamines. The Canadian guidelines from 2003 even recommended the use of ranitidine, which is actually the wrong kind of antihistamine. In that statement, Ellis and Day base their recommendation on a 1990 paper  by Phil Lieberman, whis reviews several (contradictory) studies of patients who were premedicated with antihistamines prior to receiving a dose of their allergen. Resisting this madness, local guidelines do not recommend the use of antihistamines, except in the management of symptomatic itch.

Icatibant

 Icatibant is a synthetic bradykinin B2 receptor antagonist. It is effective in hereditary angioedema (due to C1 esterase deficiency, for example). Its mechanism of action suggests that it may also be effective in allergic anaphylaxis, but it will have no effect on the histamine-mediated aspects of the disease process, and is unlikely to control all of the consequences. Plus, it is ridiculously expensive.

Issues unique to perioperative anaphylaxis

In addition to the routine medical therapy as mentioned above, ANZCA guidelines also include the following important points:

• Anaesthesia should be maintained with minimal volatile agent until the situation is stabilised
• If a non-depolarising NMJ blocker is implicated as the cause of this event, ANZCA do not recommend using sugammadex. It has never been demonstrated to work in this setting, and there will certainly be problems associated with sudden reversal of paralysis when you are trying to ventilate this patient (bronchospasm, etc).
• Though it seems sensible to abandon the surgery when someting like anaphylaxis happens, ANZCA acknowledge that in some situations it may actually be important to continue (eg. what if you were evacuating an extradural haematoma?)
• Referral to an anaesthetic allergy testing centre is required (the ANZAAG website lists local centres).

Mast cell tryptase

Much is made of this serine protease. Specifically, one should be looking at the level of mature-β-tryptase which is a measure of mast cell activation (whereas pro-β-tryptase is more of a marker of mast cell number). 

• Half life of serum tryptase is about 2 hours
• Peak level is reached about 1-2 hours after onset of the reaction.
• The level then declines over the following 3-6 hours
• A level should be collected about 24-48 hours after the event, to make sure a baseline comparison is available (eg. in mastocytosis the level is chronically high)

Under extremely weird circumstances, tryptase can lie. For instance, it is also produced in basophils, which are less numerous than mast cells by about 700-fold. Conceivably, one can have a reaction which involves basophils instead of mast cells. Tryptase may also be elevated in severe trauma.