Author + information
- Keith G. Oldroyd, MBChB, MD(Hons)∗ ()
- West of Scotland Regional Heart and Lung Centre, Golden Jubilee National Hospital, Glasgow, United Kingdom
- ↵∗Address for correspondence:
Dr. Keith G. Oldroyd, West of Scotland Regional Heart and Lung Centre, Golden Jubilee National Hospital, Agamemnon Street, Clydebank, Glasgow G81 4DY, United Kingdom.
When treating a patient with ST-segment elevation myocardial infarction (STEMI), there is nothing worse for an interventional cardiologist and his/her patient than to establish Thrombolysis In Myocardial Infarction (TIMI) flow grade 3 in the culprit artery by balloon angioplasty with prompt resolution of the patient’s pain and ST-segment elevation, only to then recreate the infarction by inducing no-reflow after stent implantation. No-reflow has been reported in up to 30% of STEMI cases, and multiple observational studies have confirmed that it is associated with an adverse prognosis (1). When it does happen, various potent vasodilators are usually administered directly into the coronary artery, but there is little evidence of benefit (2). Consequently, avoiding no-reflow in the first place is of crucial importance.
One strategy shown to minimize the risk of angiographic no-reflow in patients with STEMI is to not implant any stents at all at the time of the index procedure, allowing a period of time for pharmacological therapy to reduce the residual thrombotic burden before stenting. This is commonly referred to as deferred stenting, but is more accurately described as delayed stenting, as the vast majority of patients do eventually end up undergoing stent implantation. In a small study, the performance of which was provoked by recurrent heated discussions within our group as to the value of this approach, we randomized 100 patients to immediate or delayed stenting after restoration of TIMI flow grade 3, with a median waiting time of 9 h (range 8 to 16 h) between initial reperfusion and delayed stenting. During this period, all patients randomized to delayed stenting received full-dose antithrombotic therapy and an intravenous platelet glycoprotein inhibitor. The primary endpoint was angiographic no-reflow, assessed by a core laboratory at the Cardiovascular Research Foundation, Columbia University, New York, New York. Specific inclusion criteria were chosen to try to select a population at increased risk of no-reflow, which occurred in 29% of patients randomized to immediate stenting, but only 6% of those randomized to delayed stenting. All patients underwent cardiac magnetic resonance (CMR) studies, and there was significantly greater myocardial salvage in the delayed stenting group (3). So far, so good. However, in the much larger DANAMI-3–DEFER study (Third DANish Study of Optimal Acute Treatment of Patients With ST-elevation Myocardial Infarction), there was no benefit from delaying stenting for a median of 3 days after initial angioplasty in an “all-comers” population of 1,215 patients with STEMI on a composite primary clinical outcome of all-cause mortality, hospital admission for heart failure, recurrent infarction, and unplanned revascularization of the target vessel over 2 years of follow-up (4). DANAMI-3–DEFER also had a CMR substudy, but before going on to consider this, it is worth considering some of the potentially important differences between these 2 studies.
First, in DAMANI-3–DEFER, the investigators elected to study “all-comers,” rather than those with an increased risk of no-reflow and, accordingly, may have diluted any beneficial effect of delayed stenting and compromised their ability to detect it. Second, patients were randomized after angiography, but before initial balloon angioplasty and/or aspiration thrombectomy. This design resulted in a lot of crossover, with 22% of the patients who were randomized to delayed stenting actually undergoing immediate stenting, as the operator could not establish stable TIMI flow grade 2 to 3 with balloon angioplasty/thrombectomy alone. Ironically, of those randomized to immediate stenting, 1% received no stents at all for technical reasons or because the operators were worried about no-reflow! In clinical practice, and, I would suggest, in trials of this strategy, one would not select a patient for (potential) delayed stenting before having established that successful reperfusion had been established. Third, the main study reported an increased incidence of unplanned target vessel revascularization in the delayed stenting group: 7% versus 4%; hazard ratio: 1.7 (p = 0.03). This was mainly due to reocclusion/reinfarction in 11 patients (2%) during their “waiting” period. The investigators comment in the text of the original paper that there appeared to be a learning curve for identifying patients/lesions likely to reocclude, as only 1 of these 11 events occurred during the second one-half of study recruitment. Nevertheless, other than not selecting patients for delayed stenting in the first place, it is obviously important to balance the potential benefits of prolonged intensive antithrombotic therapy before stenting against the risk of reocclusion/reinfarction. In DANAMI-3–DEFER, glycoprotein inhibitor therapy was only used in a minority of patients (44%) and was only continued for a minimum of 4 h, whereas the median delay before stenting was 3 days. Waiting this long may have been counterproductive.
Turning now to the CMR substudy of DANAMI-3–DEFER in this issue of the Journal (5), it recruited from 1 of the 4 centers taking part in the main study. Of 235 patients in the control immediate-stenting group and 275 patients in the delayed-stenting group, just over 400 patients had paired CMR scans at pre-discharge and at 3 months. Overall, the results are consistent with the main study, showing no evidence of benefit measured as infarct size, salvage index, or incidence of microvascular obstruction. There is an interesting, but non–pre-specified subgroup analysis suggesting that in patients requiring a stent length ≥24 mm, delayed stenting was associated with a final infarct size of 13% of left ventricular (LV) mass (IQR: 7% to 23%) compared with 6% of LV mass (IQR: 2% to 18%) in patients requiring a stent length <24 mm (p = 0.006 and p-interaction = 0.005) (5). Stent length is a surrogate for lesion length, which, in turn, is a predictor of no-reflow. TIMI flow grade 0 to 1 at baseline is also a predictor of no-reflow, but was not associated with any benefit from delayed stenting in DANAMI-3–DEFER. However, again, the study population was diluted by the two-thirds of patients recruited who had TIMI flow grade 2 to 3 at baseline.
Going back to the subject of reocclusion in the delayed stenting group, the effect of this has been difficult to study in prior CMR studies, as the first scan has usually been performed after the delayed percutaneous coronary intervention (PCI). However, in DANAMI-3–DEFER, the “majority” of baseline CMR scans were performed before delayed stenting. Infarct size, expressed as a percentage of LV mass, increased in 18% of patients between the baseline and 3-month scans, but with no difference between the 2 patient groups. This is an unexpected finding, but its interpretation is complicated by the indexing of infarct size to LV mass and the fact that some of the patients who had multivessel disease were also randomized to culprit vessel-only or staged fractional flow reserve–guided complete revascularization in the parallel DANAMI-3–PRIMULTI study of complete revascularization versus treatment of the culprit lesion only in patients with STEMI and multivessel disease. So, any given patient could have had delayed culprit lesion stenting and, at the same time, staged complete revascularization. We already know that multivessel PCI generates new infarcts in some patients, but the investigators state that there was no difference in the incidence of patients with increased infarct size over time among patients randomized to fractional flow reserve–guided complete revascularization compared with culprit-only (21% vs. 16%; p = 0.47).
Ultimately, the absence of any beneficial effect of delayed stenting on clinical and surrogate CMR-derived outcomes in a randomized study of over 1,200 patients carries massively more weight than a study of 100 patients with an angiographically-derived primary outcome measure. Nevertheless, the post hoc finding of a marked benefit of delayed stenting in patients with long lesions, as defined by a requirement for a stent longer than 24 mm, is intriguing. One can speculate that these patients had a higher atherothrombotic burden in their culprit lesion, and therefore a higher risk of no-reflow, with more to gain from delaying stent implantation. Although there is clearly no justification for a routine policy of delayed stenting (and never was), there may still be room for a further study in high-risk patients with long lesions and/or TIMI flow grade 0 to 1 at baseline angiography.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. Oldroyd has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2017 American College of Cardiology Foundation
- Harrison R.W.,
- Aggarwal A.,
- Ou F.S.,
- et al.,
- for the American College of Cardiology National Cardiovascular Data Registry
- Su Q.,
- Nyi T.S.,
- Li L.
- Carrick D.,
- Oldroyd K.G.,
- McEntegart M.,
- et al.
- Kelbæk H.,
- Høfsten D.E.,
- Køber L.,
- et al.
- Lønborg J.,
- Engstrøm T.,
- Aleksov K.,
- et al.,
- for the DANAMI-3 Investigators