May 19th, 2014
Lytics in STEMI: A New Analysis of Data from FAST-MI
Nicolas Danchin, MD, PhD and John Ryan, MD
In a new paper in Circulation, Nicolas Danchin and colleagues analyze data from FAST-MI and conclude that a fibrinolysis-based strategy may be valid for some patients with STEMI. CardioExchange’s John Ryan asks Nicolas, Professor at the Department of Cardiology, Hôpital Européen Georges Pompidou, for further perspective on the findings and the settings in which they apply.
Ryan: Your paper is surprising in that most people, based on the trials, would expect that patients receiving primary PCI would do so much better than those receiving lytics. How do you explain the contrast between your real-world results and the trials?
Danchin: Among Western European countries, France has continued to use lytics in a substantial proportion of STEMI patients, although the figure has declined to 14% in the most recent survey, carried out at the end of 2010 (FAST-MI 2010). This can probably be explained by two factors . First, some patients don’t live near cath labs, and the expected transportation times are sometimes well beyond 60 minutes; in the FAST-MI 2010 survey, median time from diagnostic ECG to primary PCI was 110 minutes, and about 25% of the patients received primary PCI more than 3 hours after ECG . Second, France has a very developed system of emergency ambulances, which are physician-staffed (the SAMU, “Service d’Aide Médicale Urgente”), and emergency physicians have gained experience over the years with the selection of patients for lytic therapy and the use of lytics in the prehospital setting. Beyond this particular situation, however, it must be recognized that the initial trials comparing intravenous fibrinolysis with primary PCI, which form the basis of current guidelines, used fibrinolysis as a standalone treatment. Since then, several trials have shown that this should no longer be the case. The REACT trial showed the benefit of rescue angioplasty in patients without signs of early reperfusion after lytic treatment, and several recent trials have shown that a pharmaco-invasive strategy (i.e., initial IV fibrinolysis, followed by routine angiography with PCI when necessary) was beneficial compared to stand-alone lytics even in patients who had signs of early reperfusion with lytics. Recently, the STREAM trial compared primary PCI with a pharmaco-invasive strategy in STEMI patients seen early after the onset of chest pain, and in whom PCI was not expected to be feasible in less than 1 hour. Although the time from randomization to PCI was rather short (median, 77 minutes) in the primary PCI arm, 1-month and 1-year results were similar for primary PCI and the pharmaco-invasive strategy.
Ryan: Do you think these findings are unique to France?
Danchin: All in all, both the observational data from France and the results of the STREAM trial suggest that a pharmaco-invasive strategy is a valid alternative to primary PCI, particularly when the time from diagnosis to expected PCI is likely to be long. Such favorable results suppose, of course, that a careful selection of the patients is made, in order to avoid any contraindication and to minimize the potential bleeding risk associated with lytics. Although the results of the pharmaco-invasive strategy were especially favorable in patients receiving prehospital lytic therapy in the French experience, the results of STREAM strongly suggest that administering lytics in non-PCI hospitals then transferring the patients to an institution with 24/7 PCI availability would be equally favorable. In the U.S., the experience of the Mayo Clinic network, as well as that of Minnesota community hospitals, also indicate that excellent results can be achieved with this approach. Likewise, a similar pharmaco-invasive protocol has been implemented in Norway in areas with long transfer distances. In short, these results are certainly not specific to the French context, and similar results should be expected elsewhere, provided that physicians providing initial care can properly select those who will benefit most from the pharmaco-invasive strategy: mostly younger patients, without contraindication to lytics, seen early after the onset of chest pain.
Ryan: I am in Utah, and we have a lot of patients living in the Rockies to whom we give lytics because of their remote location. Are there similar isolated areas in France where lytics are the norm?
Danchin: The situation in France is not so dissimilar to yours. Lytics are virtually no longer used in large cities where cath labs are easily available 24/7. But time delays for delivery of primary PCI, even in large cities with a number of PCI-capable institutions, are often longer than we usually think: In the greater Paris region, median time from ECG to primary PCI was 105 minutes in the FAST-MI 2010 survey! Lytics continue to be used for patients living (or staying) in the countryside and in regions not easily accessible (e.g. in some villages or small towns in the Alps, which must be a situation quite similar to that you have in the Rockies, although distances are usually shorter in France).
Has anyone looked at ‘lyrics in those “subtle, non-ST elevation” markers of acute coronary occlusion? I suspect that we can agree on true posterior. But, how about the isolated STE in avl, or the inferior or lateral ST elevation that is easily identified but doesn’t meet the 1 mm criterion or is seen better in the reciprocal ST depressions, or the “hyperacute” T wave without ST segment elevation. In some settings, it is common for the ER Docs and interventionists to agree on getting those to the cath lab as “STEMI equivalents”, but I haven’t stumbled across any discussion of thrombolytics for those cases.
It’s interesting. Looking at the data from our institution, which goes back to 2002, the mortality from STEMI treated by in-hospital thrombolysis was around 10%. Then we moved to a phase of pre-hospital thrombolysis where the mortality (of those who received it) was around 2.5%. Now we have a primary PCI service and the mortality is around 5%. We have a semi-rural population and some patients have quite long transfer times.
Clearly there are many potential biases, but it also suggests that pre-hospital thrombolysis is not so bad. An analysis of the UK MINAP database would be interesting and I think is required. It would not conclusively answer the question but it may generate some interesting data.
In the UK there has been a wholehearted move towards Primary PCI because of national policy. If I lived within a short distance of a centre with a primary PCI service that had a good record of delivering PCI in a timely fashion then that is what I would choose. If I lived a bit further out I’m not so sure. I am sure there is a window in which primary PCI is optimal and beyond that pre-hospital thrombolysis.
I don’t believe there has been a randomised trial of primary PCI versus pre-hospital thrombolysis but I may be mistaken.
Thanks for the great interview. I wonder if there is another component here (which has never occurred to me before) – are pre-hospital lytics followed by routine angio, with PCI if necessary, more patient-centered. If I was a patient and had a low risk of bleeding, I might actually prefer lytics and potentially avoid a stent if deemed unnecessary by cath. Your outcomes data suggest that this would be a reasonable approach, even for those in close proximity to a PCI-capable hospital.
our practice stands vindicated with the evidence from STREAM trial, we in india can’t offer Primary PCI as a norm, for many reasons. we practice a lytics as standlone therapy in most of the cases. of these many patients then move to better centre some 5-7 days later get a routine angio and a definitve treatment, hence forth.
the pharmacoinvasive is a better way to manage most acute MI’s.
Professor, thank you for the great blog post. Supplementing my education through reading about the applications of evidence based medicine has been a great help. I look forward to reading more of these blogs and articles to improve my foundation of knowledge as I continue through my training.