Role of post-resuscitative coronary angiography

The question, "does this patient need angiography", comes up frequently in the management of the post-cardiac arrest patient. Arguments can be made for and against. On one hand, coronary pathology may be revealed which needs to be dealt with, and which might have precipitated the arrest. On the other hand, the angiogram might be pointless (i.e. no lesion, or nothing stentable), or actually harmful (all that contrast). Lastly, one has to question the utility of an expensive intervention in a patient whose longevity will be determined by their neurological recovery, something more related to the duration of ischaemic downtime than to the patency of their coronaries  

Question 18 from the first paper of 2018 was the first time the college have ever asked about this issue. The SAQ was worded atypically ("what are the pros and cons of this approach" they asked, after a statement that all such patients should go to cath lab). Then, it appeared again, essentially unchanged, as Question 8 from the second paper of 2020. A more reasoned answer would probably result from a "critically evaluate" question, where the trainees might have the opportunity to discuss not only advantages and disadvantages but also the rationale, background and available evidence. It is not inconceivable that in the future the college will ask this question in that way. To prepare for that eventuality, this chapter is designed as a critical evaluation answer.

The excellent powerpoint presentation by Georg Furnau Luebeck for the European Society of Cardiology is a good starting point to look for references. Some of the best review of the most important arguments for and against angiography in unselected cardiac arrest patients can be found in the paper on the study design of the COACT trial by Lemkes et al (2016). By extension, one should also be at least passingly familiar with the actual COACT trial (Lemkes et al, 2019).

Rationale for angiography in all post-arrest patients

The main reasons given to rationalise performing an angiogram in all-comers are:

• ST changes in the ECG post arrest are difficult to interpret 
• History of chest pain may not be available (patient is comatose and unable to comment)
• Just because there is no ST elevation doesn't mean there is no coronary problem. Of the patients who had no ECG changes, Hollenbeck et al (2014) found an acute thrombotic coronary occlusion in 26%.
• To exclude coronary artery disease is an important step in the process of determining the causes of the cardiac arrest
• Patients undergoing angiography receive a "greater intensity of care" (Lemkes et al, 2016) - they are resuscitated more aggressively, get seen by more doctors, receive early anticoagulation and have more mechanical / pharmacological support, which could translate into better outcomes. Presumably, in contrast to them, the comatose out-of-hospital arrest survivor receiving conservative management in the ICU is just parked in a dark corner until it's time to have a family conference.

Evidence in support of early angiography for all post-arrest patients

There is ample evidence that reversible acute coronary artery disease is prevalent among cardiac arrest patients, and that reversal of this disease leads to better outcomes:

• Spaulding et al (1997) was the first major study of this matter, and pre-dates therapeutic hypothermia. The investigators performed angiography in 84 patients who had "no obvious non-cardiac causes" for their arrest. 48% of them had clinically significant coronary artery occlusion, and the predictive value of ECG changes or clinical features (eg. history of chest pain) was very poor. "Nine patients presented with no ST-segment elevation or chest pain and were found on angiography to have had a recent coronary-artery occlusion",  the authors complained. There was an improvement in the survival of post-angio patients.
•  PROCAT registry from Paris (Dumas et al, 2010) confirms that among the STEMI patients, 96% had significant coronary artery lesions on angiography- which is not surprising. What was surprising was the incidence of these lesions among the non-STEMI crowd - 58% of those patients also had significant coronary artery lesions. Those patients did not have any ST segment elevation on their post-ROSC ECG. This suggests that cardiac arrest itself strongly suggests cardiac pathology when there is no other immediately obvious cause.
• Hollenbeck et al (2014) in a large retrospective cohort found that patients without ST changes who ended up having an early angiogram had greatly improved survival at hospital discharge (65.6% vs. 48.6%), compared to those who had no ST changes and who were treated conservatively.

There is lukewarm society support for the practice of unselective angiography in these patients:

• The AHA vaguely supports the  practice: they say that it's "reasonable". (In full, "Emergency coronary angiography is reasonable for select (eg, electrically or hemodynamically unstable) adult patients who are comatose after OHCA of suspected cardiac origin but without ST elevation on ECG.")- Callaway et al, 2015
• The ESC/ERC are conflicted, on one hand saying that angiography should be "considered as soon as possible (less than 2 h) ... in the absence of an obvious non-coronary cause", but then they also call the practice "controversial" and back off to saying that "it is reasonable to discuss and consider" angiography in these patients (Nolan et al, 2015).

Arguments against early angiography for all post-arrest patients

• Cardiac arrest is not uniformly a phenomenon of coronary artery disease, i.e. there are many noncardiac causes, of which several (eg. SAH) would surely not benefit from the obligatory loading doses of dual antiplatelets. This is an argument against
• Angiography may exacerbate the acute kidney injury which often accompanies the post-resuscitation syndrome, mainly by means of a contrast load.
• Even where there is coronary artery disease, not all patients can be stented, and the survival benefit of angiography seems to be limited to those patients in whom stenting was successful. In about 25% of patients undergoing PCI, there is either no lesion or a non-stentable lesion, even when there are ST changes (and if there aren't, that proportion rises to 75%) according to Dumas et al (2010)
• Even where there is stentable disease, there may be no mortality benefit to stenting it, because outcome depends more on the global ischaemic damage from "down-time" than the events in local coronary territories. The SWEDEHEART study (Wester et al, 2018) certainly did not find any mortality difference between patients who had early PCI versus those who did not, even though 43% of the patients were found to have 90% stenosis in one of their vessels.
• If stenting is so good for outcomes, then stenting all the lesions should give maximal benefit - but in fact it seems the fewer stents you do, the better. The CULPRIT-SHOCK trial (Thiele et al, 2017) found improvement in mortality if the angiographer limited their post-arrest intervention to just the culprit lesion, with both mortality and risk of AKI increasing if multiple stents were deployed. 
• Most of the data in support of early coronary angiography if French, and specifically Parisian (which limits generalisability). 
• The COACT trial (Lemkes et al, 2019) found that immediate angiography following cardiac arrest without ST elevation did not improve survival at 90 days. As critics have pointed out, the study design was bizarre (haemodynamically unstable patients were excluded, and the investigators  powered the study to detect nothing less than a 40% change in mortality, which would be unheard of in critical care literature). Unlike the PROCAT registry, only 20% of the arrested Netherlanders had an acute coronary lesion (33% in the "immediate" group).   
• The TOMAHAWK trial (Desch et al, 2021) also did not find any mortality benefit at 30 days.

What's happened since the last time this appeared in 2018?

• Jentzer et al published a trial in (2018), specifically in February of 2018 (i.e. long after the examiners would have stopped thinking about this question paper, but before was inflicted on the trainees in March). Survival was much better in the angio group (56.2% vs 31%) as was the neurological outcome (28% vs 11%). However, 
• Verma (2020) performed a meta-analysis of about 3500 patients, and found little difference in mortality or outcome; rather, the 30-day mortality was more related to the presentation comorbidities 
• Song et al (2021) used the GRACE risk score and found that patients with a high score (i.e. old age, history of CCf or MI, tachycardia or hypotensive, with ST-segment depression, AKI, raise cardiac enzymes) seem to have some survival benefit from early angiography