Response to Letter Regarding Article, “Primary Angioplasty Versus Fibrinolysis in Acute Myocardial Infarction: Long-Term Follow-Up in the Danish Acute Myocardial Infarction 2 Trial”
Dr Fresco raises concerns regarding a secondary endpoint (all-cause mortality) in the subgroup analyses of DANAMI-2. We will try to elaborate on this issue, even though DANAMI-2 was never powered to perform such analyses. First, age is a strong predictor for mortality. In the referral hospital subgroup, there was a trend toward higher age in the fibrinolysis subgroup (median 64 versus 62 years, P=0.06), than in the group allocated to primary percutaneous coronary intervention (pPCI), and the opposite was found in the invasive hospital subgroup (median 62 versus 64 years, P=0.25). This random allocation thus may impact the subgroup analyses. Second, quality of care may differ at invasive and referral centers and could have influenced the results. Third, had Dr. Fresco focused on cardiac death, which may be more relevant, then pPCI was superior to fibrinolysis in the overall cohort (Δ3.2%), the invasive subgroup (Δ2.4%), and the referral subgroup (Δ3.6%), ie, a very consistent result.
Dr Fresco also discusses the seemingly paradoxical observation that patients successfully treated with fibrinolysis had the lowest mortality rate, and, at the same time, a higher risk of reinfarction. Based on the premise that reperfusion is beneficial in ST segment elevation myocardial infarction, it is not surprising that patients with reperfusion had a lower mortality. Reperfusion, however, has the inherent disadvantage of leading to a risk of reocclusion, and thus reinfarction, of the culprit lesion. In contrast, patients without reperfusion do not have the risk of reocclusion of the (already occluded) culprit lesion. It therefore seems logical, rather than paradoxical, that patients successfully treated with fibrinolysis had the lowest mortality rate and a higher risk of reinfarction.
In conclusion, the concerns raised by Dr Fresco are based on secondary endpoints related mainly to a prematurely terminated subgroup. We recognize the careful nature of the interpretation, but it is nonetheless our opinion that such an interpretation is scientifically flawed. The overall result of the DANAMI-2 study is that pPCI, as compared to fibrinolysis, reduced the composite end point of reinfarction and death by an absolute 6.5%, mainly driven by a 6.8% reduction of reinfarction, and secondarily by a 3.5% mortality reduction. Therefore, excluding the fact that treatment strategies for both fibrinolysis and pPCI have evolved over the years since DANAMI-2, we have no concerns about using the results to determine current strategies for the management of ST segment elevation myocardial infarction.
Michael Maeng, MD, PhD
Peter Haubjerg Nielsen
Martin Busk, MD, PhD
Steen Dalby Kristensen, MD, DMSc
Torsten Toftegaard Nielsen, MD, DMSc
Henning Rud Andersen, MD, DMSc Department of Cardiology
Aarhus University Hospital
Leif Spange Mortensen, MSc UNI-C
Danish Information Technology Centre for Education and
- © 2010 American Heart Association, Inc.