Pretreatment with heparin in patients with ST-segment elevation myocardial infarction: a report from the Swedish Coronary Angiography and Angioplasty Registry (SCAAR)
Reported from ESC Congress 2022
Daniele Giacoppo analyses the results of the SCAAR registry which compared patients with STEMI treated by primary PCI in Sweden from January 2008 to December 2016 according to the UFH pretreatment status...
In patients with ST-segment elevation myocardial infarction (STEMI) undergoing primary percutaneous coronary intervention (PCI), unfractionated heparin (UFH) before arrival at the coronary catheterization laboratory is common practice. Early administration of UFH is thought to improve spontaneous reperfusion and reduce clot burden before primary percutaneous coronary intervention (PCI). However, there is a paucity of high-quality randomized clinical trials and large-scale observational studies supporting the advantage of pretreatment with UFH in contemporary primary PCI.
In a Late Breaking Trials session at the European Society of Cardiology 2022 annual congress, Oskar Emilsson presented the results of a cohort study from the Swedish Coronary Angiography and Angioplasty Registry (SCAAR) nationwide register comparing patients with STEMI treated by primary PCI in Sweden from January 2008 to December 2016 according to the UFH pretreatment status. Chronic total occlusions, missing information regarding UFH pretreatment, and rescue PCI following thrombolysis were among the exclusion criteria. Clinical and angiographic variables prospectively collected in the SCAAR registry were complemented with additional information extracted from the National Patient Register. The outcomes of interest were coronary artery occlusion at angiography, 30-day mortality, and in-hospital major bleeding. The outcomes were adjusted for multiple potential confounders primarily by adjusted Poisson regression with different numbers of covariates (three models), and secondarily, as sensitivity analyses, by propensity score matching, with and without multiple imputations, and inverse probability of treatment weighting.
A total of 41631 patients were included in the study, 16026 (38%) receiving pretreated and 25605 (62%) not receiving UFH pretreatment. Overall, the mean age was 67 years, 71% were male, 15% had diabetes, 14% had a prior myocardial infarction, and 10% had prior PCI. Pretreatment with UFH varied between 1% and 85% across the PCI centres, was more frequently associated with preprocedural administration of aspirin (standardized mean difference [SMD] = 0.49, p < 0.001) and ticagrelor (SMD = 0.49, p < 0.001), and less frequently with preprocedural administration of clopidogrel (SMD = -0.20, p < 0.001), warfarin (SMD = -0.17, p < 0.001), fondaparinux (SMD = -0.34, p < 0.001), low molecular weight heparin (SMD = -0.33, p < 0.001), and thrombolysis (SMD = -0.20, p < 0.001). No antithrombotic medication was given before PCI in approximately 17% of patients not pretreated with UFH, while the totality of patients pretreated with UFH received at least an antithrombotic medication (SMD = -0.20, p < 0.001). Patients with UFH pretreatment underwent more frequently transradial PCI than patients without UFH pretreatment (SMD = 0.56, p < 0.001). During PCI, patients pretreated with UFH received less frequently aspirin (SMD = -0.56 p < 0.001), clopidogrel (SMD = -0.49, p < 0.001), UFH (SMD = -0.59, p < 0.001), prasugrel (SMD = -0.14, p < 0.001), and ticagrelor (SMD = -0.28, p < 0.001).
Crude models showed that pretreatment with UFH was associated with a reduced risk of coronary artery occlusion at angiography (risk ratio [RR] 0.90 [0.89-0.92]), 30-day mortality (RR 0.77 [0.71-0.83]), and in-hospital major bleeding (RR 0.85 [0-75-0.96]) compared with no pretreatment. Regardless of the number of covariates included in the model, multivariate Poisson regression concluded showed consistent results in terms of coronary artery occlusion at angiography (model 1: RR 0.92 [0.90-0.93]; model 2: RR 0.89 [0.88-0.91]; model 3: RR 0.89 [0.87-0.90]) and 30-day mortality (model 1: RR 0.80 [0.72-0.88]; model 2: RR 0.88 [0.72-0.99]; model 3: RR 0.87 [0.77-0.99]). However, adjusted risk estimates for in-hospital major bleeding were not significant (model 1: RR 0.91 [0.79-1.04]; model 2: RR 1.00 [0.86-1.17]; model 3: RR 1.01 [0.86-1.18]).
Subgroup analyses based on the fully adjusted model showed that the effect on coronary artery occlusion at angiography was consistent, while some heterogeneity was present for the endpoints of mortality (significant interaction for access site and extent of coronary artery disease) and major bleeding (significant interaction for access site, body weight, and age).
The analysis of the propensity score matched datasets was consistent (coronary artery occlusion, RR 0.88 [0.86-0.89]; 30-day mortality, RR 0.79 [0.69-0.92]; in-hospital major bleeding 1.00 [0.82-1.21]). However, the results by inverse probability of treatment weighting in terms of 30-day mortality were inconsistent when results were not truncated at 95th percentile to exclude extreme weights (no truncation: RR 1.15 [0.80-1.67]; 95th percentile truncation: RR 0.93 [0.81-1.06]; 99th percentile truncation: RR 0.87 [0.77-0.98]).
Pretreatment with UFH associated with reduced risks of coronary artery occlusion at angiography and 30-day mortality?
The present nationwide analysis from the SCAAR registry shows that in patients admitted for STEMI undergoing primary PCI, pretreatment with UFH seems to be associated with reduced risks of coronary artery occlusion at angiography and 30-day mortality.
The first result may be linked to the UFH-mediated promotion of spontaneous vessel recanalization. The second, more relevant, result of a 13% to 23% relative risk reduction in 30-day mortality may be theoretically associated with reduced size of myocardial infarction. However, this conclusion remains speculative and further high-quality analyses are warranted to validate this association.
These relevant results should be interpreted against a number of considerations, In the present study, substantial differences in clinical characteristics and antithrombotic therapy were observed between groups. The authors performed multivariable adjustment, propensity score matching, and inverse probability of treatment weighting with generally consistent results. Of note, the unclear finding of a significant in-hospital bleeding risk reduction associated with UFH pretreatment observed at crude analysis was consistently disproved across adjusted analyses. However, adjustment methods rely on available data and residual confounding effects cannot be fully excluded. The large differences in the use of medications between groups (i.e., aspirin, ticagrelor, and prasugrel administration more frequent in the UFH pretreatment group; clopidogrel, fondaparinux, and low molecular weight heparin administration more frequent in the no UFH pretreatment group; 17% of patients in the no pretreatment group without any antithrombotic therapy before PCI; more patients in the UFH pretreatment group receiving thrombolysis before PCI) likely reflects substantial dissimilarities in the clinical development and diagnosis of STEMI between groups and the effects attributable to other antithrombotic agents (e.g., potent P2Y12 inhibitors) cannot be disregarded. Against this background, the results of this study require validation in a randomized setting.
In addition, large-scale registries and post hoc analyses often are affected by limitations in data availability and accuracy. In the present study, relevant information including door-to-balloon time, time of UFH administration after symptoms onset, UFH dose in relation to the body weight, and activated clotting time values, was not available or available but not further analysed. Moreover, there was a paucity of data on coronary artery disease extent and distribution as well as the main technical aspects of primary PCI. The interpretation of 30-day mortality needs to take into account differences in coronary artery disease patterns and primary PCI success.
Finally, the high between-centre heterogeneity in UFH pretreatment and variations in pharmacological and interventional strategies over time may have played a role. Mixed-effect models and time-dependent analyses could be necessary, though no statistical method can eradicate the influence of these important factors could not be eradicated.
Notwithstanding these unaddressed questions, the absence of signals of harm in this study from the SCAAR registry, the short half-life of UFH, and the periprocedural anticoagulation requirements of PCI make reasonable the systematic pretreatment with UFH of patients with STEMI undergoing primary PCI.
Latest news from ESC Congress 2022
Author
No comments yet!