AVOID: Oxygen Potentially Problematic for Normoxic Patients with STEMI

AVOID: Oxygen Potentially Problematic for Normoxic Patients with STEMI

November 4, 2019 3 By William Morgan


(dramatic music) – I’m Karl Kern from the
University of Arizona, and I’ve been asked to interview and comment on the AVOID
trial, a very interesting, provocative trial about
the use of oxygen in acute myocardial infarction. Dion Stub, the principal
investigator is with me. Dion, welcome. We’re glad that you had
time to come down from Canada to the warm weather of Chicago. – Thanks, Dr. Kern, pleasure to be here. – Why don’t you just begin by telling us the background of this study,
and what it’s all about. – As you know, I think,
for many years, it’s been international practice
that we commonly give, almost as a first response,
oxygen to patients we suspect that myocardial
infarction or acute coronary syndrome. Our first aiders do it,
paramedics do it, nursing staff in the cath lab, it’s
almost this ubiquitous carry-over response, I
suppose, to treatment from over 50 years ago, that
really hasn’t been put under the scrutiny of modern day
randomized controlled trials. And, maybe more concerningly
is that there’s increasingly physiological evidence that,
if we do give oxygen to patients who are normoxic, we can quickly render them hyperoxic, or give them significantly
high levels of oxygen, and this, in turn, can lead to increasing coronary vascular
resistance, reduction in coronary blood flow,
production of free radicals, microcirculation impairment, and that all may contribute
to rate profusion injury. So, there’s, I suppose,
growing concern that maybe this common practice is
not the best thing for patients with AMI. And, that was really the
hypothesis that the AVOID trial tried to answer. Does the common practice of
giving oxygen to patients without hypoxia cause
increased myocardial injury in patients with STEMI. – So this was an
investigator-initiated project, as I understand it. – Yeah, so the trial was
coordinated by Ambulance Victoria and their research division
in Melbourne, Australia, and, really, in collaboration with nine of the big tertiary STEMI
centers in Melbourne. And, with support from
not-for-profit organizations, but, yeah, it was very much a pragmatic
investigator-initiated trial to try and answer this
very common, fundamental first aid practice. – Well, give us kind of
the nitty-gritty of how you did it, and what you found. – Patients could only be
randomized by Victorian-trained paramedics, and they
included all patients who described the STEMI symptoms
for less than 12 hours, had a pre-hospital 12-lead
ECG that was suggestive of ST elevation MI, and importantly, were not hypoxic on a pulse oximeter, and that was defined for
this study as an oxygen saturation greater than or equal to 94%. They had a preserved conscious state, and at that point, the
ambulance paramedics would open randomization opaque envelopes, and that would instruct
them either to not give any oxygen, or to give oxygen at 8 liters per minute via a mask. And, that oxygen therapy
continued through to the emergency department, the cath lab, and then when the patient
was stable on the ward. – And what’d you find? – So, the co-primary
end point of the study was myocardial infarc size
on the routine biomarkers of CK and troponin I, and we found a significant 25%
increase in creatinine kinase released, both, that
was mean peak as well as area under the curve, and that
was with a P value of .01. And, that was in the oxygen
group, so it looked like increased myocardial
injury in those that were given oxygen. And the troponin result
was not as significant. There was no significant difference, though the curves look
very similar to CK with this non-significant 20% trend, but the confidence
intervals were too large. – And the total number involved? – So, we randomized 638
patients pre-hospital on that, because of the pre-hospital
non-confirmed STEMIs. That ended up being 441
patients with STEMI. And we then went on and
offered all the patients cardiac MRIs at six months as a measure of final infarc size. And 139 patients ended
up having a cardiac MRI, and that was interesting,
again, you see more a trend to increased
myocardial injury with patients that were given
oxygen, with, again, this 30% increase in light
gadolinium enhancement. – And, my understanding was that was a significant difference, also. – Yes. – So, that’s actually
probably the bottom line, is that, by infarc sizing,
it looks like oxygen therapy for the normoxic
acute MI patients may not be necessary and perhaps even harmful. – Yeah, it’s a relatively small trial. I think it’s important that it was done in a pre-hospital and in-hospital settings. It would be nice for
it to be confirmed With hard clinical end points
with a larger study that, you know, our Swedish colleagues are actually undertaking at the moment. But, yeah, as a clinician,
I’d have concerns about giving high-flow oxygen
to patients with normal oxygen saturations. – Do you think MONA’s finally dead? – It’s possible. – Yes. Thanks very much, we’ve
enjoyed being here at scientific sessions in Chicago.