Deep hypothermic circulatory arrest

Deep hypothermic circulatory arrest is the delicate art of draining the patient into a bucket, cooling them down to 15-20ºC, and stopping their heart so the surgeon can operate on a bloodless aortic root. This topic has only appeared once, as Question 16 from the second paper of 2021. The specific elements asked in this SAQ were the rationale  indications and perioperative complications

Definition of deep hypothermic circulatory arrest

The practice of DHCA has lots of permutations and there is no one specific definition, nor a specific boundary that separates "deep" from "not deep". Also, though the name seems to exclude techniques where some element of the circulation is preserved (eg. anterograde or retrograde cerebral perfusion ) even though they might still involve cooling the patient down to 18-22º C. A standard technique, described as "straight" DHCA, is summarised in an excellent article by Ziganshin et al (2014):

• First, cardiopulmonary bypass is established. Seeing as usually you will be doing some sort of aortic surgery, this usually involves a right atrial drainage cannula and femoral arterial return. 
• Topical cooling of the patient's head with ice may be involved in some centres
• Systemic cooling down to 18-20º C is then commenced, and this usually only takes 30-40 minutes.
• Alpha-stat management of the acid-base balance is used during this period
• Circulatory arrest  is then initiated, the aorta is unclamped, and the surgical team can do what they need to do in an essentially bloodless field (as there is no blood flow during this period). This usually takes about 30 minutes.
• Circulation is reestablished after the aortic root is repaired, and carries on for some minutes before  you can rewarm the patient - apparently to reduce the risk of raised intracranial pressure.
• Rewarming is then commenced, and this is usually a much longer process - perhaps 60 minutes or longer. The slower the better. Fast rewarming may result in a situation where the cerebral metabolic rate increases faster than the oxygen delivery to the brain, increasing the risk of brain damage.

Rationale for deep hypothermia and circulatory arrest

The rationale for each is a little different, and at a really basic level you could say that deep hypothermia is necessary for the patient to tolerate circulatory arrest, and the circulatory arrest is necessary for the surgeon's convenience. 

Rationale for hypothermia is cerebral protection. Well, organ protection, in general, which is the result of hypothermia. All sorts of famous graphs track the reduction in cerebral metabolic rate as the temperature drops, and all of them are usually some variation on this one below. It comes from Kirklin/Barratt-Boyes Cardiac Surgery (the authors' one is the 4th edition from 2012). 

Nobody ever seems to comment on this, but the origin of these data points was actually a paper by Croughwell et al (1992). The authors were trying to create an equation that predicted the maximum duration of poor or absent flow at different temperatures. From this data set, it is apparent that at 18º C, cerebral metabolic rate decreases to something like 10-15% of the normal rate seen at 37º C. The basic theory underlying DHCA is that with hypothermia, there might be a safe duration of completely absent flow which might be well tolerated. 

So, why do you need this protection against absent flow? Because the surgery requires for there to be no flow, for a rather sustained period of time. 

Rationale for circulatory arrest is surgical access. The surgical approach to the replacement or repair of the aortic arch and aortic root requires that at some stage the aorta be incised and open to air. To have blood flowing through it at this stage would be fairly uncivilised- impractically bloody and embolism-prone. The alternative (to bypass the aortic root and somehow perfuse everything that needs to be perfused without stopping the circulation) is still possible, and some authors hold that this method is protective against some of the worst side effects of DHCA. At this stage there does not appear to be any major advantage of one method vs. the other, and defenders of "straight" DHCA point to its long and proven track record.

Indications for deep hypothermic circulatory arrest

So, when might you need to do this to somebody? The following comprehensive list was stolen directly from Chapter 62 of Anaesthesiology Core Review:

• Cardiac surgery:

  • Aortic arch reconstruction (aneurysm, rupture, dissection)

  • Pulmonary thromboendarterectomy

  • Repair of complex congenital heart defects (transposition of the great arteries, total anomalous pulmonary venous return, hypoplastic left heart syndrome)

  • Vascular reconstruction during cardiac transplant

• Non-cardiac surgery:

  • Surgery on the thoracoabdominal aorta

  • Repair of giant cerebral aneurysms

  • Resection of cerebral arteriovenous malformations

  • Resection of renal cell carcinoma with caval invasion

  • Resection of other tumors with caval invasion

How low can you go?

And how long can you go, that low? The duration of hypothermic arrest is actually more controversial than the depth of hypothermia. You could theoretically reduce the temperature down to 0º C, as there really does not appear to be any technological barrier to this. However, even at extremely low temperatures, the metabolism of the brain does not completely cease, i.e. one can never get down to zero percent metabolism. Ergo, there is a finite time span that could be safely tolerated, and beyond which some level of brain injury would develop. Moreover, blood stasis tends to result in regional microvascular dysfunction which leads to decreased perfusion after the blood flow is restored again, and this is also something that seems to be time-related. As the result, it is generally believed that

• Thirty minutes of DHCA is well tolerated by most people
• Forty minutes of DHCA can be well tolerated, but the evidence for the safety of this practice is not very robust
• Sixty minutes of DHCA is poorly tolerated, and there is plenty of robust evidence for that.

DHCA here is conventionally defined, i.e. temperatures of 15-20º C. That is as low as you normally go, for routine midweek DHCA. Experimenters have naturally gone further, mainly in animals. Michenfelder & Milne (1992) cooled dogs down as low as 7º C, and found that the CMRO₂ decreased even further, to something like 5% of normal (0.2ml/100g/min, down from 3.59). However, the drop from 18º C to 7º C does not appear justified. The EEG is already isoelectric by this point, and the extra 5% decrease in cerebral metabolism would not produce a marked increase in the safe duration of DHCA. But it certainly would increase the rewarming time, which is already extremely long. The result would be more complications, and more use of operating theatre resources, for no real gain.

Complications of deep hypothermic circulatory arrest

Question 16 from the second paper of 2021 specifically asked the trainees to outline the adverse effects of DHCA that may be encountered in ICU, after the patient returns from the operating theatre. These are obviously going to a bit difficult to separate from the complications associated with cardiopulmonary bypass in general, or from the complications of the major aortic surgery. Did the patient develop cognitive dysfunction from the stasis of DHCA, or from microemboli of bypass, or from the cholesterol dislodged during aortic clamping, or from some unrecognised air emboli?  

Anyway. Specific complications attributable to the overall technique of DHCA for aortic surgery are as follows:

• Respiratory complications:
  • Hypoxia due to atelectasis is the consequence of being on bypass for a prolonged period, and of having the lungs deflated. They many not reinflate very well.
  • ARDS could develop as the extreme manifestation of prolonged bypass, as the result of the inflammatory complement-activating effects of the bypass itself. Indirectly, this is the consequence of DHCA, because DHCA is the thing that requires such a prolonged bypass time (mainly waiting to slowly rewarm the patient)
• Circulatory complications:
  • Low cardiac output states are common following DHCA, partly because of the prolonged bypass time and partly because of whatever the original cardiovascular pathology had been.
  • Arrhythmias are common, especially if the patient returning from theatre is incompletely rewarmed.
• Neurological complications:
  • Stroke (mainly embolic, and mainly in those who had more than 40 minutes of DHCA); the risk is higher in aortic surgery than in normal cardiac surgery.  
  • Seizures
  • Choreoathetosis (usually develops 2-6 days following the surgery; seems to be related to temperatures below 15º C)
  • Decreased cognitive function: this data seems to mainly come from ancient studies which were mainly performed on children, but sounds relatively serious: 6% of the kids in one series had sub-normal IQ (~80) not explained by other factors (Stevenson et al, 1974). These days, everybody seems to get consented for these procedures with a caveat that one might experience a substantial drop in their cognitive abilities, but it seems that this is perhaps less of an issue for modern techniques. Percy et al (2009), surveying high-functioning survivors of DHCA, found that they were able to carry on their highly demanding jobs without any major cognitive changes (these people were doctors, business managers, teachers, IT professionals and artists, to name some of the occupations). In fact, some people actually reported better cognitive abilities, perhaps related to having a more normal cerebral blood flow. In short, some very subtle deficits may occur, and the risk definitely increases if you extend the arrest time beyond 45 minutes, but we don't tend to routinely do that, and so we probably don't see this very often.
  • Spinal cord ischaemia can also occur. Hypothermia provides some spinal protection, and the spinal cord is generally less susceptible to ischaemia than the brain is. Still, spinal injury can occur, though it is actually less frequent in procedures involving the root and arch (1.1%-4.3%) as compared to procedures where the descending thoracic aorta is interfered with (Hlaing et al, 2016).
• Renal complications
  • Acute kidney injury: There is the possibility that vulnerable kidneys will be injured by the process of DHCA and the subsequent reperfusion. This appears to be a relatively low risk when you look at the total population of DHCA patients. 
• Hepatic complications:
  • LFT derangement is rare, because the liver is surprisingly tolerant of prolonged low flow or arrest while profoundly hypothermic, 
• Haematological complications:
  • Coagulopathy is more likely in these patients, again mainly because of the longer bypass time, which leads to more platelet activation and more clotting factor depletion.

Other complications may also be seen in association with DHCA, but not due to it. For example, tthe left recurrent laryngeal nerve travels close to the arch of the aorta, and can be injured as collateral damage. According to Hlaing et al (2016), some degree of vocal cord paralysis occurs in 5% to 21.9% of thoracic aortic surgery patients. On the other hand it could actually improve: the aortic arch aneurysm may have been causing a recurrent laryngeal nerve neuropraxia. You would inevitably see this in patients following DHCA, but it is not one of the effects of DHCA per se.