3

Results

A total of 3576 patients over 75 years old, undergoing pPCI for STEMI were finally included in this registry and among these, 1830 (51%) met the criteria for the presence of MVD. The clinical, angiographic and procedural characteristics of these patients compared to those with single vessel disease (SVD) are shown in Table 1 . There were significant differences for many variables between both groups.

The management of the 1830 patients with MVD is shown in Fig. 1 . In 847 patients (46%) multivessel revascularization was done, either during pPCI (51%), in a staged procedure (44%) or in both procedures combined (5%). Among the 847 patients who underwent multivessel PCI, in 566 (67%) a complete revascularization was achieved. Therefore, at the time of discharge incomplete revascularization was observed in 1264 patients (69%), including the 983 patients with culprit-only PCI and the 281 not achieving complete revascularization after multivessel PCI. Noteworthy, we observed a non-significant trend for more multivessel PCI done in most recent years. 

Open full size image

Study flow chart.

This figure shows the management of patients with multivessel disease according with the revascularization strategies applied before discharge.

We analyzed the independent predictors for performing multivessel revascularization ( Table 2 ). Less older patients, males, those with previous infarction, no chronic renal failure, lower ejection fraction, in Killip class I or II, treated with drug-eluting stents or treated with IIb-IIIa inhibitors were more prone to be revascularized. With regards to gender, women were older (81.7 ± 4.7 years vs. 80.8 ± 4.1 years, p < 0.001), but they were less likely to get multivessel PCI even after adjustment by age. Noteworthy, some of the predictors of multivessel revascularization infer reverse causality, for example IIb-IIIa inhibitors or drug-eluting stents may have been used because multivessel PCI was performed and not vice versa. 

In Table 2 are listed as well the independent predictors for the timing approaches for multivessel revascularization. Patients with previous myocardial infarction, anterior location of infarction, longer time delay, advanced Killip class, no 3-vesel disease, use of femoral access site and treated with bare metal stents were more preferentially treated in the pPCI procedure.

The clinical outcomes for different subgroups are shown in the Figs. 2–4 . The follow up was completed in 97.4% of patients. The survival free of all cause death + infarction or cardiac death + infarction at 2 years was better for those undergoing multivessel PCI ( Fig. 2 ). This favorable difference was observed both in patients in Killip classes I–II and classes III–IV ( Fig. 3 ). The achievement of a complete revascularization status made no difference in outcomes among those patients undergoing multivessel PCI ( Fig. 4 A). In relation with timing of non-culprit lesions PCI, the best prognosis corresponded to those with staged procedures ( Fig. 4 B). 

Open full size image

Survival free of death/cardiac death and infarction in patients with MVD.

The survival free of death and infarction (A) and cardiac death and infarction (B) at 2 years was significantly better for those patients with MVD undergoing multivessel PCI.

Open full size image

Survival free of cardiac death and infarction in patients with MVD according to Killip class.

In this figure survival free of cardiac death and infarction is shown for those in Killip class I or II (A) and Killip class III or IV (B). Patients undergoing multivessel revascularization showed better prognosis in both subgroups.

Open full size image

Survival free of cardiac death and infarction in patients with MVD according to completeness of revascularization and timing of non-culprit lesions PCI.

A). The achievement of an angiographically defined complete revascularization status made no difference in outcomes among those patients undergoing multivessel PCI. B) The best prognosis corresponded to those patients undergoing non-culprit lesions PCI in staged procedures.

In the analysis of independent predictors for cardiovascular death and infarction at 2 years ( Table 3 ), multivessel PCI lowered the incidence of these adverse events (HR 0.60, 95% CI 0.40–0.89: p = 0.011). Entering separately the variables of multivessel PCI during pPCI or in staged procedure, only the latter remained predictor for better outcomes, with the former showing a slight positive trend. When entering the variable of complete revascularization instead of the one of multivessel PCI in the multivariant model, the former resulted an independent predictor of major adverse events, but it was no longer the case in the subgroup of patients undergoing multivessel revascularization ( Table 3 ). 

Regarding bleeding events during admission, the rate of bleeding BARC > 1 in multivessel PCI group was 2.7% and in the culprit-only PCI group was 2% (p = 0.4). The incidence of definite or probable thrombosis in 2 years follow up was 3.8% in multivessel PCI group and 2.9% in the culprit-only PCI group (p = 0.35). The incidence of revascularization after discharge in 2 years follow was 9.3% and 8.4% respectively (p = 0.53).

The propensity score matching analysis yielded two groups of 514 patients each showing well balanced in all features (standardized difference < 10%). The survival free of death and infarction for the same subgroups considered in the overall analysis is shown in Fig. 5 . Again, a significant survival advantage was seen in favor of those undergoing multivessel PCI, especially when staged, and without differences according to completeness of revascularization. 

Open full size image

Survival free of death and infarction in matched groups.

A significant survival advantage was seen in favor of those undergoing multivessel PCI (A), especially when staged (B) and without differences according to anatomically defined completeness of revascularization (C).

4

Discussion

The main results of this study could be summarized as follows: a) Half of elderly patients (> 75 years old) with STEMI undergoing pPCI present with MVD; b) Nearly half of them underwent multivessel revascularization, with a similar rate of acute or staged procedures and under factors clearly influencing those decisions; c) Multivessel PCI is related with better 2 years outcomes and the benefit was greater if done in staged procedures; d) No harm was found for multivessel PCI during acute procedures; e) Among those patients undergoing multivessel PCI, the achievement of anatomically defined complete revascularization did not have influence on 2 years outcomes.

Routine revascularization of non-culprit lesions before hospital discharge has been supported by the recently published 2017 ESC guidelines in which a class IIa recommendation is given for this strategy [26] . The recommendation for this strategy was class IIb in the 2015 ACC/AHA guidelines [27] . These recommendations are based on the evidence from randomized trials.

The first trial included only 214 patients and after a mean follow-up of 2.5 years, patients allocated to culprit lesion angioplasty-only had more major adverse cardiac events [6] . After this study, four larger randomized clinical trials have been published. 78910 Multivessel PCI was done either during the index procedure [710] , staged during hospital admission [9] , or any time before discharge (immediate or staged) [8] . Indication of non-culprit PCI was angiography-guided in lesions with ≥ 50% stenosis [7] , > 70% stenosis [8] , or fractional flow reserve-guided [910] . Primary outcome (composite of different endpoints) was significantly reduced in the multivessel revascularization group in all four trials but mortality was not statistically different in any of these four trials. The pooled analysis of these trials shows that multivessel PCI at index or staged procedures is associated to reduction in MACE due to reduction in urgent revascularization. 11121314 being the risk of all-cause mortality and spontaneous reinfarction not different. As the optimal timing of revascularization (immediate vs. staged) has not been adequately investigated, no recommendation in favor of immediate vs. staged multivessel PCI can be formulated.

In the particular case of patients with MVD and cardiogenic shock a IIa recommendation has been given for multivessel PCI during index procedure in the 2017 ESC guidelines based on a level C evidence [26] . A meta-analysis of nonrandomized studies suggests that there may be no significant benefit with single-stage multivessel-PCI compared with culprit only-PCI. [28] Furthermore, this recommendation has been definitely challenged by the recently published CULPRIT SHOCK trial. [29] In this trial immediate multivessel PCI increased the 30-day risk of a composite of death or severe renal failure leading to renal-replacement therapy and the relative risk of death.

In clinical practice there are several factors to consider for performing multivessel PCI at the time of primary PCI including patient stability, ability to accurately assess lesion significance in the setting of acute coronary syndromes, likelihood of procedural success, risk of the intervention, and the potential benefit of revascularization. Furthermore, the presence of unstable morphology in angiography or intracoronary imaging, the use of pressure wire in lesions with stenosis 50–75% and the location of the lesion (if proximal or in a vessel that subtends a significant area of myocardium) play an important role.

Patients over 75 years present particular characteristics that limit the validity of the application of results derived from general population. Thus, evidence is lacking to support any recommendation specifically aimed to the elderly with STEMI and MVD. Advanced age could induce a more conservative approach. In our registry, the proportion of patients with multivessel disease that were revascularized in hospital was 46% and ageing decreased likelihood of non-culprit intervention independently. Female sex was indentified as well as a variable promoting a culprit-only lesion revascularization approach, regardless of the age. Patients in worse clinical condition (Killip class III or IV) were less likely to undergo multivessel revascularization. The lack of supporting evidences at that time and prior negative experiences made the operators less prone to perform multivesel PCI in this setting. This conservative strategy has been proven to be more adequate just recently. [29] In fact, the survival benefit seen in the subgroup of patients with shock undergoing multivessel PCI in this registry support a careful selective indication for multivessel PCI in this critical scenario.

The timing of non-culprit interventions was associated as well with certain variables. Patients with history of previous infarction, anterior infarction, Killip class III–IV or later presentation underwent more frequently multivessel PCI in the pPCI procedure. On the other side, a more extensive disease (three vessel disease) and the use of radial access and drug-eluting stents favored staged procedures. The apparent discrepancy for the Killip class III-IV in both analysis can be explained by the very high mortality of these patients in the first 24–48 h of presentation, which made very difficult for them to be referred for staged procedures, which accounted for almost half of non-culprit PCI procedures. With respect to the higher use of drug-eluing stents in staged procedures this is explained because staged procedures were related with more complex lesions, 3-vessel disease and with patients with Killip class I or II (higher expectancy of long term survival).

Regarding the prognostic implications of the different therapeutic approaches. In this registry, multivessel PCI was associated with an absolute risk reduction of 5% and was an independent predictor (reduction by 40%) over 2 years in the combined incidence of cardiac death and myocardial infarction. The prognostic advantage of multivessel PCI was observed across all the Killip classes. When comparing multivessel revascularization vs. culprit-only revascularization in STEMI, the value of revascularization as an outcome is to some extent questionable and definitely less relevant. That is why a combined endpoint of cardiac death and infarction could be in our view more adequate and so was selected for this study.

With respect to the timing of intervention, the survival curves showed an early separation and this finding could justify a prompt non-culprit PCI. We found that staged PCI of non-culprit lesions was associated to better outcomes. Moreover, in survival curves it appears that patients having non-culprit PCI during pPCI showed a quite similar outcome than those with culprit lesion PCI alone. Nevertheless, though multivessel PCI in acute procedure was clearly associated to a worse patient profile, it was associated to a trend for better outcomes in multivariant analysis.

Finally, completeness of revascularization emerged as an independent predictor of cardiac death and myocardial infarction in overall population with MVD. However, among those patients who underwent multivessel PCI, the achievement of anatomically defined complete revascularization did not appear to have influence on two years outcomes. These results suggest that not all lesions have the same prognostic impact on outcomes, and perhaps only the treatment of non-culprit lesions located in proximal or large segments of main coronary arteries is prognostically relevant. Furthermore, lesions with 50–75% stenosis left untreated could be functionally nonsignificant. In this regard, a combination of angiographic and functional criteria (based on fractional flow reserve) could be more adequate no define complete revascularization.

5

Study limitations

This is a retrospective multicenter registry and, although patients were strictly consecutively recruited from each center, data collection and follow-up were not as accurate and uniform as they would have been with a prospective design. Due to the retrospective design, it was not possible to record medical therapy adherence and its eventual relationship with ischemic and hemorrhagic events. Given the nature of the study and the large number of events in this population, event adjudication was decided by the local investigators using pre-defined criteria. In case of conflicting data, the clinical information was analyzed at the coordination site and adjudication was done accordingly by two investigators.

Regarding the predictors for performing multivessel PCI, some of the predictors infer reverse causality, which means that these factors may have been used because multivessel PCI was performed and not vice versa. For the purpose of comparing outcomes between treatment groups, the registries entail limitations particularly the problem of bias secondary to known and unknown confusion factors not fully sorted out even after careful adjustments with multivariant analyses. Then, it still remains highly possible that some unmeasured confounders could favor the use of one therapeutic alternative over the other. Thus, the results of the multivariant analysis might have been somewhat affected by hidden biases, such as comorbidities omitted from the databases or biological frailty, which could in turn be linked to some therapeutic strategies more than others. These biases could be particularly relevant in the elderly population and could affect the results of the survival analysis for the different treatment groups. The steps taken in order to attenuate these limitations consisted in the use of a propensity score matching using a full set of baseline (clinical and angiographic) variables, and avoiding the inclusion of procedural factors that could bias the analysis (such as access site, stent type or antithrombotic therapy among others).

Definition of significant lesion was based on angiographic findings. We realize that a certain proportion of these lesions could not be significant by means of a functional assessment with pressure wire. Data relative to the use and results of pressure wire within the procedures was not found to be reliably collected and is not provided. Regarding completeness of revascularization, the residual Syntax score would have been of value, but this information is not available. The use of drug-eluting stents was not high, but these results reflect the real practice in recent years. Although a higher penetration could be considered currently, the exclusion of revascularization from the primary outcome diminish the influence of this variable.

6

Conclusions

In this registry around half of patients showed significant multivessel disease and almost half of them underwent multivessel PCI with similar rate for acute or staged procedures and under factors clearly influencing those decisions. Multivessel PCI was related with better two years outcomes but the benefit seems to be greater for staged procedures. Of note, no harm was detected for multivessel intervention during pPCI. Angiographically defined complete revascularization had no influence among those undergoing multivessel PCI.

Statements

No conflicts of interest to disclose. None of the authors hold any kind of financial and personal relationships with other people or organizations that could inappropriately influence (bias) this work.

This work has not been published previously (except in the form of an abstract) and it is not under consideration for publication elsewhere.

The manuscript has been approved by all authors and tacitly or explicitly by the responsible authorities where the work was carried out, and if accepted, it will not be published elsewhere in the same form, in English or in any other language, including electronically without the written consent of the copyright-holder.

No funding was required for this study.