So you’ve done it. In a dazzling display of medical prowess worthy of a medical TV drama you have restored the dead to life!
With scrupulous airway management, high quality chest compressions (assisted by some waveform capnography, of course), careful administration of adrenaline boluses (I know, I know) and liberal application of Direct Current, your Out Of Hospital Cardiac Arrest (OOHCA) patient now has a pulse. Good job! You’re feeling pretty good about yourself aren’t you?
You haven’t really fixed the problem have you? You’ve been hard at work managing the cardiac arrest and its aftermath, but this person’s heart didn’t just go and stop for no reason did it?
Unless there’s a really obvious cause here, the chances are your patient has had a heart attack.
So lets get cracking! Pour some aspirin down that NG tube and start some heparin, cause that’s how we treat our ACS patient right?
What about our STEMI patients? In regular patients, the sight of that magic ST-Elevation (STE) in adjacent leads means a rapid trip to the cath lab for the friendly cardiologists to unblock whatever lesion lies beneath. Because we know this works for our normal, every day, heart still beating kind of STEMI, we have extrapolated this to our OOHCA patients, and we probably all would send patients who had STEMI post ROSC to the lab (with cooling in progress of course!).
Retrospective analysis tends to agree. STEMI predicts lesions on PCI and successful PCI is associated with better outcome.
[learn_more caption="Before we get too carried away..."] Due to the nature of our study population (patients suddenly who get as poorly as it is a possible to be without warning) randomised trials are hard to do, and the evidence is mostly observational. This means that we have to be very careful when assuming any causative link between the interventions and outcomes. So for the rest of the post, whenever I say “PCI is associated with better outcomes” what I really mean is “being in the group of patients whom were selected to go for PCI is associated with better outcomes”. Further more, the outcomes measured are often what we would consider ‘soft’ outcomes. Many of the papers use ‘acute coronary lesion’ or ‘evidence of recent coronary occlusion’ as an outcome. We then assume that this is a lesion that PCI can treat, and then assume that doing so helps the patient. A lot of assumptions there. Please don’t think I am criticising any of the excellent work I am about to reference, but merely illustrating the limitations of doing research in this population[/learn_more]
So far, so good.
But are we doing enough? It’s fair to say that if your ACS was serious enough to actually stop your heart, then it was a pretty major event. Is is really fair to lump all of the OOHCA NSTEMI patients in with the regular “I’m fine but my 12 hour trop is slightly raised” NSTEMI’s? Could any more of the OOHCA patients benefit from a trip to the cath lab?
The evidence from OOHCA patients suggsts that using STEMI as the sole criteria for post ROSC reperfusion therapy may not be adequate.
Dumas et al published a major paper on this subject in 2010. In a cohort of 435 patients (OOHCA with no obvious non cardiac cause) they found that 58% of patients with no STEMI had an acute coronary artery lesion, with a successful PCI in 26%. They worked out that using STEMI as the test for acute coronary lesions gave a PPV of 0.96 (great!), but a NPV of 0.42 (not so great). A paper by Zannutuni et al this year found similar results, with a PPV of 0.85, but a NPV of 0.67.
A successful PCI may be beneficial to these patients. The Dumas paper also found that a successful PCI was associated with survival, even in those without ST elevation (47% survival with successful PCI versus 31% without, P = 0.001). Hollenbeck et al published this year on this subject, and found that PCI was associated with survival to discharge in OOHCA without STEMI: 65.6% versus 48.6% (p = 0.017).
Remember of course, as we discussed above, that ‘PCI was associated with’ really means ‘being in the group that got PCI is associated with’. The observational nature of these trials means we should not attribute causation to any one intervention, no matter how tempting it might be.
It’s worth noting that both the Dumas and Zannutuni cohorts both had a high rate of VF/VT arrests, 68% and 77% of patients respectively, and the Hollenbeck trial included VF/VT arrests only.
So it seems that by using STEMI as our sole criterion here we may be denying some patients a beneficial treatment, at least in those who present in a shockable rhythm.
What do our guidelines say? The UK Resuscitation Council ALS guidance does suggest that some patients without STEMI may benefit from PCI:
“(PCI) should be considered in all post-cardiac-arrest patients who are suspected of having coronary artery disease as the cause of their arrest” (page 77)
So encouraging, but not very specific. What we need are some pointers, some features that suggest the patient in front of us may benefit from going for PCI. What kind of things should we be looking at?
[learn_more caption="Rhythm?"] The trials we discussed above including mainly patietns presenting in VF or VT. So it would seem reasable to suggest that in these patients, there may be a culprit lesion that the cardiologists can treat. This idea is further supported by work by Sideris et al, who retrospectively reviewed 165 OOHCA patients (50% VF/VT), who had no obvious non-cardiac cause. They were looking for features that predicted acute MI (as defined on angiography, 36% of patients). They found that presenting in VF predicted MI, with an odds ratio of 7.4 for MI, and presenting in asystole made it much less likely, with an OR or 0.17.[/learn_more]
[learn_more caption="Symptoms?"] What about the whole ‘taking a history’ thing. Seems a little quaint when we’ve got all this high tech stuff going on right? Often neglected in these patients I think, but can yield some valuble information. I’m thinking about the patient who complained of a crushing chest pain, before keeling over in a VF arrest. Surely he’s having an MI, even if there is no STEMI on the ECG? Can we send him for PCI? Well maybe. Spaulding et al did some of the seminal work in this area. They looked at 84 OOHCA patients, with no obvious non cardiac cause. 83% were VF/VT arrests, and 48% had evidence of coronary occlusion on angiography. In amongst their findings they found that a pre-arrest complaint of chest pain was predictive of acute coronary occlusion – to the tune of an OR of 4.0.[/learn_more]
[learn_more caption="ECG changes?"] Finally, can the post arrest ECG be used to identify those patietns for PCI? We know about the STEMI patients, but what about the LBBB’s , or the ST depressions or T wave inversions? This is more difficult to tease out, as many patients will have ECG changes as a result of the cardiac arrest that are not due to an underlying MI, and these changes may be temporary. The study by Sideris et al mentioned above attempted to define a set of ECG criteria to predict the need for PCI. In their 165 patient cohort thewy found that using ST elevation alone as a predicotor of MI had a PPV of 76% and a NPV 92%. Not too bad, but still missing some patients. They looked at a few different ECG criteria and came up with this: ST elevation or depression and/or LBBB and/or non specific QRS widening. Using these criteria they found a PPV of 52%, but a NPV of 100%. So no patients missed, but lots of unnesscary angiograms. In real terms, using this criteria meant taking another 7 patients to the lab to find one MI. Worth it? Sounds like it, assuming (there’s that word again…) that finding that lesion is beneficial to the patient.[/learn_more]
[learn_more caption="Cardiogenic Shock?"] This is a little tricky. We know that a significant number of patients will develop cardiogenic shock after ROSC. The well known SHOCK trial suggests that angiography/angioplasty is beneficial in patients suffering from cardiogenic shock who have not arrested. The confounding factor here is the idea of cardiac ‘stunning’. This is a cardiac dysfunction that occurs quite commonly post ROSC, but often resolves without intervention. Laurent et al found that 49% of post arrest patients developed myocardial dysfunction within hours of ROSC. However, this spontaneously improved over the next 24-72 hours. This suggests to me that cardiovascular instability can’t be used as a predictor of the need for early PCI, at least not without another suggestive feature.[/learn_more]
This has turned into a little bit of an epic….
To sum up:
It appears from what observational evidence we have that there is a group of OOHCA patients who will benefit from emergent angiography/angioplasty/stenting even if they do not have STEMI on their post ROSC ECG.
I think the evidence kind of backs up what we felt already. Personally I’m going to be calling the cardiologist if my patient:
Has had a VF/VT arrest without an obvious non-cardiac cause
Has complained of cardiac chest pain before the arrest
Of course, what the cardiologist decides to do is another matter, but at least now I can back up my arguments.
Would be interested to know if anyone out there works with a PCI protocol that includes post-arrest NSTEMI patients, let me know!