This was a multicenter retrospective descriptive correlational study based on the Portuguese National Registry on Acute Coronary Syndromes (ProACS) that included all patients diagnosed with STEMI treated by PCI and P2Y12 inhibitor LD between October 1, 2010 and September 19, 2017. Two groups were established: patients treated by LD before PCI (LD-PRE) and patients treated by LD during or after PCI (LD-CATH). We identified predictors of which strategy was used, and assessed the resulting impact on prognosis and safety.

All patients aged ≥18 years, registered in ProACS between October 1, 2010 and September 19, 2017, diagnosed with STEMI and treated by PCI were included. The diagnosis of STEMI was based on the presence of chest pain or angina equivalent in the previous 48 hours and ST-segment elevation in contiguous leads, according to the definition in the European guidelines.1

Patients missing data required to determine the timing of P2Y12 inhibitor LD administration were excluded.

Data were collected on demographic characteristics, relevant personal history, cardiovascular risk factors, location of STEMI, means of transport to the hospital, hospital of admission, form of admission, and admission to hospitals without a cath lab. The following timings were calculated: symptoms to first medical contact (FMC), FMC to balloon, symptoms to balloon, and door to balloon. Data were also recorded on admission characteristics, coronary angiography, left ventricular ejection fraction (LVEF), and medication prior to and during hospitalization and at hospital discharge. Adverse events during hospital stay (mortality, non-fatal reinfarction and stroke) were recorded, as well as the composite endpoint of these three major adverse events (MAE), major bleeding according to the GUSTO definition, need for transfusion or Hb drop>2 g/dl, and the composite bleeding endpoint of these three events.

The main aim of the study was to assess the impact on prognosis and safety of using a P2Y12 inhibitor LD before (LD-PRE) versus during or after PCI (LD-CATH). The prognostic endpoint was the composite endpoint of MAE and the safety endpoint was the composite bleeding endpoint.

A descriptive analysis was performed of the variables selected for characterization of the sample. Continuous variables are presented as means and standard deviation and categorical variables are presented numerically as percentages. The two groups were compared in terms of the selected variables, using the chi-square test for categorical variables and the Student’s t test for continuous variables. The threshold for statistical significance of correlations was set at 95% for p<0.05. Logistic regression analysis was used to identify predictors of the choice of treatment strategy, with statistical significance set at p<0.05, and to identify predictors of prognosis and safety, in which an odds ratio (OR) >1 indicated that LD-CATH was a predictor of this variable and an OR<1 indicated that LD-PRE was a predictor.

The logistic regression model was adjusted for the differences between groups.

A total of 6757 patients with STEMI, 5238 of them treated by PCI, were initially selected for the study. Of these, 54 (1.0%) were excluded because they had not received a P2Y12 inhibitor LD, three (0.1%) due to lack of information on P2Y12 inhibitor LD administration, and 1058 (20.2%) due to lack of information on timing of P2Y12 inhibitor LD administration. The final sample thus comprised 4123 patients, 2774 (66.3%) in the LD-PRE group and 1349 (32.4%) in the LD-CATH group.

The characteristics of the two study groups are displayed in Table 1.

More patients in the LD-CATH group were male, but mean ages were similar in the two groups. In terms of personal history and cardiovascular risk factors, a history of MI, PCI, angina and smoking and a family history of coronary artery disease (CAD) were more frequent in the LD-CATH group, while a history of malignancy was more frequent in the LD-PRE group. Most patients in the LD-PRE group arrived at the hospital by their own means, while most in the LD-CATH group arrived by medical emergency response vehicle (MERV). The main location of hospital admission in the LD-PRE group was the emergency department (ED), but was the coronary intensive care unit or the cath lab in the LD-CATH group. Arrival via the coronary fast-track system was more likely to be associated with LD-CATH, while LD-PRE was more often associated with admission to a hospital without a cath lab or to the ED.

MI of undetermined location, together with new-onset left bundle branch block (LBBB) and failure to identify the culprit vessel on coronary angiography, were more common in the LD-PRE group. Delays before PCI were shorter in the LD-CATH group. Chest pain as the main symptom was more prevalent in the LD-CATH group. The proportion of patients in Killip class II or higher was greater in the LD-PRE group, while systolic blood pressure (SBP) was lower in the LD-CATH group. Multivessel disease and aspiration thrombectomy were more frequent in the LD-PRE group, and radial access and use of glycoprotein (GP) IIb/IIIa inhibitors were more common in the LD-CATH group. Rates of PCI success were similar in the two groups. Regarding medication both during hospital stay and at discharge, ticagrelor was used more by patients in the LD-CATH group, while previous therapy with ticagrelor and vitamin K antagonists was more common in the LD-PRE group.

Predictors of the prehospital use of P2Y12 inhibitor LD were age <45 years, MI of undetermined location with new-onset LBBB, Killip class >1, LVEF<50% and symptom-to-balloon time >120 min. Predictors of the use of P2Y12 inhibitor LD in the cath lab were a history of smoking, angina, MI and dementia, anterior STEMI, chest pain as main symptom, heart rate (HR) >100 bpm, and SBP<90 mmHg (Table 2).

In terms of prognosis, a simple comparison shows that patients in the LD-PRE group had lower LVEF (48±11% vs. 56±12%, p<0.001) and more reinfarction (1.0% vs. 0.1%, p=0.006), MAE (5.5% vs. 3.6%, p=0.008), Hb drop >2 g/dl (28.5% vs. 14.3%, p<0.001) and the composite bleeding endpoint (30.6% vs. 15.8, p<0.001). No differences were seen in in-hospital mortality, stroke, major bleeding or need for transfusion.

The following factors were identified as predictors of the composite endpoint of MAE: female gender (p=0.003; OR 1.22-2.69), age >75 years (p<0.001; OR 1.78-3.84), HR>100/min (p=0.014; OR 1.12-2.76); SBP<90 mmHg (p<0.001; OR 1.94-5.63), Killip class>I (p<0.001; OR 1.76-4.03), LVEF<40% (p<0.001; OR 2.68-6.33), and multivessel disease (p=0.002; OR 1.24-2.69). The timing of P2Y12 inhibitor LD administration did not predict MAE (p=0.882; OR 0.60-1.54).

Predictors identified for the composite bleeding endpoint were cardiac arrest as main symptom (p=0.005; OR 1.42-7.60), HR > 100/min (p=0.001; OR 1.23-2.12), Killip class>I (p<0.001; OR 1.36-2.33), symptom-to-balloon time >120 min (p<0.001; OR 1.37-2.65), LVEF<40% (p=0.031; OR 1.02-1.64), multivessel disease (p=0.001; OR 1.15-1.68), and prehospital LD (p<0.001; OR 0.37-0.58).

Analysis of the individual endpoints showed that prehospital LD also predicted Hb drop >2 g/dl (p<0.001; OR 0.32-0.51), reinfarction (p=0.03; OR 0.05-0.88), and heart failure (p<0.001; OR 0.39-0.66). The timing of P2Y12 inhibitor administration did not predict major bleeding (p=0.311; OR 0.77-2.25), need for transfusion (p=0.724; OR 0.46-1.71), or death (p=0.652; OR 0.64-2.02).

Considerable differences were seen in the baseline demographic and clinical characteristics of the two groups. Overall, the characteristics of the LD-PRE group appear to indicate a worse prognosis. The clinical history of patients in the LD-CATH group, such as a history of MI, PCI or unstable angina, made a diagnosis of CAD more likely. For the authors of the present paper, this was a somewhat surprising finding, since a stronger suspicion of MI should theoretically be expected to result in earlier use of a P2Y12 inhibitor. This could at least in part be due to the difficulty in establishing a diagnosis of STEMI in a patient with previous CAD and MI, who may present baseline electrocardiographic alterations.10

Concerning the operation and efficiency of the prehospital transport network, patients in the LD-CATH group were more often transported via the coronary fast-track system and by MERV and more often admitted directly to the cath lab, without requiring interhospital transfer, which conferred a prognostic benefit by enabling earlier PCI.11,12 These differences were reflected in shorter symptom-to-balloon time (total ischemic time), which was significantly longer in the LD-PRE group, resulting in a worse prognosis, as shown by their higher rates of MAE.13

Analysis of the choice of P2Y12 inhibitor shows that prasugrel was used in only one patient, while ticagrelor was used more often in the LD-CATH group. This difference may be explained by the fact that ticagrelor was not available for use in MERVs in the first years of our study. Although there was a larger proportion (without statistical significance) of patients in the LD-PRE group who had previously taken ticagrelor, presumably as part of dual antiplatelet therapy, these patients were also more likely to have received a P2Y12 inhibitor before PCI. This is somewhat surprising but is unlikely to have affected the study’s results, as they accounted for fewer than 1% of the overall population. Despite the recommendations in the guidelines to use ticagrelor or prasugrel,1,2 during this period of the study clopidogrel continued to be the most often prescribed P2Y12 inhibitor.

Data from coronary angiography reveal that radial access was used less frequently in patients in the LD-PRE group, who also more often had multivessel disease, two factors that, once again, would indicate a worse prognosis for this group; in addition, use of vascular access other than radial increases bleeding risk.14–16 Aspiration thrombectomy was used much more frequently in the LD-PRE group, which can lead to a higher rate of stroke,17 although this was not observed in our sample. GP IIb/IIIa inhibitors were used more often in the LD-CATH group, but this did not significantly increase the rate of bleeding complications.

The culprit vessel was not identified in a higher proportion of patients in the LD-PRE group (7.4% vs. 0.4%), which may be related to differential diagnosis with other conditions that have a similar presentation to STEMI, such as myocarditis and Takotsubo cardiomyopathy, but are not associated with CAD. When assessed in the cath lab following visualization of the coronary anatomy, these patients were not administered a P2Y12 inhibitor LD, which would explain the difference observed. The timing of P2Y12 inhibitor administration did not affect the PCI success rate, demonstrating that the LD-PRE group received no benefit in this regard.

Finally, in terms of the prognostic and safety endpoints, the LD-PRE group had a greater percentage of events, including higher rates of reinfarction, MAE, Hb drop >2 g/dl and the composite bleeding endpoint, which confirms that the differences described above had a negative effect on prognosis in this group.

The ATLANTIC study,5 in which 1862 patients were randomized to receive prehospital (in the ambulance) or in-hospital (in the cath lab) treatment with ticagrelor, set out to determine the best timing for P2Y12 inhibitor LD administration. Efficacy, as measured by, among other endpoints, the proportion of patients with at least 70% resolution of ST-segment elevation at 60 min after PCI and TIMI flow 3 at the end of the procedure, was similar in the two groups. No differences were seen in mortality, stroke or bleeding; the only statistically significant difference was a reduction in stent thrombosis in the prehospital ticagrelor group.

Concerning the use of clopidogrel, there are only a few registries and a single small randomized trial, which suggested a benefit for prehospital LD.6–8 A meta-analysis with a total of 37 814 patients who underwent angioplasty but not necessarily PCI showed that pretreatment with clopidogrel did not change mortality or bleeding rates, but did reduce the number of thrombotic events (MI, stroke and urgent revascularization).18 Another meta-analysis, assessing pretreatment with P2Y12 inhibitors in patients with non-ST-elevation MI (NSTEMI), found no impact on mortality but a significant increase in bleeding complications,19 while in a non-randomized cardiac magnetic resonance study, pretreatment with clopidogrel was associated with lower prevalence and smaller extent of microvascular obstruction, but no difference in infarct size.9

Regarding prasugrel, a randomized trial in patients with NSTEMI found no benefit from prasugrel pretreatment in terms of prognosis or safety.20

As can be seen, multiple studies have approached this subject but their results have been inconsistent, and it remains to be proved unequivocally whether pretreatment with a P2Y12 inhibitor LD is in fact beneficial. 21 The only randomized trial in patients with STEMI did not demonstrate a benefit of this strategy.5 However, the newer P2Y12 inhibitors have a more rapid onset of action than clopidogrel,22,23 and so studies indicating a benefit from clopidogrel pretreatment may not apply to the newer drugs. A small real-world study of pretreatment with ticagrelor at least 1.5 hours before PCI showed improved TIMI flow grade after PCI but no impact on mortality, stent thrombosis or bleeding.24

In our study, which included 4123 patients with STEMI treated by PCI, multivariate analysis showed that use of a P2Y12 inhibitor LD had no impact on mortality or MAE, and that prehospital treatment increased the risk of reinfarction. The authors consider this latter finding unexpected, since the purpose of administering a P2Y12 inhibitor is to reduce thrombotic events and to improve coronary flow at the time of PCI by inhibiting platelet aggregation earlier. Although in univariate analysis there was a marked increase in MAE in the LD-PRE group, this was not seen in multivariate analysis, in which the worse prognosis in this group was related to the greater number of comorbidities, as stated above, and with longer delays before PCI. With regard to the bleeding endpoints, the composite bleeding endpoint and Hb drop >2 g/dl were more frequent in the LD-PRE group, but major bleeding and need for transfusion were not.

Considerable differences were seen between the groups in radial access, use of GP IIb/IIIa inhibitors and direct admission to the cath lab (all much more frequent in the LD-CATH group), which, as pointed out above, have important implications for prognosis and bleeding risk. It should be noted that these variables were included in the logistic regression model in order to minimize the differences between the groups and to improve the reliability of the results obtained.

An overall analysis of our results indicates that pretreatment with P2Y12 inhibitors has no prognostic benefit and increases the risk of bleeding complications in patients with STEMI undergoing PCI, implying that this strategy is of little value, which to some extent contradicts the international guidelines.1,2 Our study clearly has significant limitations, but it is in line with others that call the current recommendations into question.5,19,24 It can be seen from our analysis that when considering predictors of the study endpoints, many of the factors that increase thrombotic risk also increase bleeding risk, such as HR>100 bpm, multivessel disease, LVEF<40% and Killip class>I, and these factors were shown to be predictors both of MAE and of the composite bleeding endpoint.

Age <45 years, MI of undetermined location with new-onset LBBB, Killip class >1, LVEF<50% and symptom-to-balloon >120 min were predictors of the use of prehospital P2Y12 inhibitor LD. This strategy tended to be used in younger patients, in those with evidence of heart failure (left ventricular dysfunction and worse Killip class) and in those in whom revascularization was expected to occur later.

A history of smoking, angina, MI and dementia, anterior STEMI location, chest pain as main symptom, HR>100 bpm, and SBP<90 mmHg were predictors of P2Y12 inhibitor LD administration in the cath lab. This strategy was preferred in patients with established or probable CAD (those with a history of smoking, angina and MI, and chest pain as main symptom) and in those with anterior MI and greater hemodynamic instability.

In both cases, given the limitations imposed by the retrospective nature of the study, it is difficult to be sure whether these predictors are taken into consideration in decisions concerning the strategy to adopt or are in fact the consequence of the chosen strategy. These findings should therefore be treated with caution.