ABC | Volume 112, Nº2, February 2019

Original Article Somuncu et al Antiplatelet resistance in young MI patients Arq Bras Cardiol. 2019; 112(2):138-146 Statistical analysis Statistical analysis was performed using the SPSS software version 18.0 for Windows (SPSS Inc., Chicago, Illinois, USA). Visual (histograms, probability plots) and analytical methods (Kolmogorov-Smirnov test, Shapiro-Wilk’s test) were used to assess the normal distribution of the variables. Descriptive analyses are presented as means and standard deviations for variables with normal distribution, as median and interquartile range for non-normal distribution. The categorical variables are expressed as numbers and percentages. Comparisons between the groups were performed using unpaired Student’s t -test or one-way ANOVA for continuous variables with normal distribution, and Kruskal-Wallis or Mann-Whitney U test for continuous variables without normal distribution. Tukey and Tamhane’s T2 tests were used based on the equal variance assumption in binary comparisons in groups with normal distribution and more than two independent variables. Mann-Whitney U test was used for the binary comparison of multiple groups with non-normal distribution. A Bonferroni correction was employed to adjust for multiple comparisons. Categorical data were compared with the chi‑square test. Because of the statistical difference in the total model, the chi-square test was applied in binary groups to compare 3-year MACE results. The cumulative survival curve for 3-year cardiac mortality was executed using the Kaplan–Meier method, with differences assessed by log-rank tests. Multivariate Cox regression backward stepwise, that included variables with p < 0.01 on univariate analysis, was carried out to identify independent predictors of 3-year MACE. A p value less than 0.05 was considered statistically significant. Results Among the 123 patients included in the study, the prevalence of poor responders to aspirin was 16.2%, to clopidogrel 18.6%, and to dual therapy 13.0%. In other words, in young MI patients, 47.8% of resistance against one or more antiplatelet was detected. Among baseline characteristics, hyperlipidemia, presence of family history, platelet counts, and platelet aggregation were different between the groups; no other differences were detected (Table 1). At the 3-year follow-up, the difference in the primary outcome (composed of CVmortality, non-fatal reinfarction, TVR, advanced heart failure, and stroke) was statistically significant between the groups (p< 0.001). When we analyzed secondary outcomes, cardiac mortality and TVR were statistically higher in the group of poor responders to dual therapy (p = 0.002, p = 0.010 respectively) (Table 2). More MACE was observed in the group of poor responders to dual therapy and clopidogrel poor responders compared to the group of dual responders (OR: 1.875, 1.144-3.073, p<0.001; OR: 1.198, 0.957-1.499, p = 0.036, respectively) (Figure 1). In logistic regression analysis, family history, LVEF and clopidogrel aggregation time were identified as independent predictors of MACE in 3 years. Besides, we found that being a poor responder to dual therapy had 3.3 times increased odds for 3-year major adverse cardiovascular events than being in the dual responder group independent from family history and LVEF (Table 3). Moreover, the Kaplan-Meier survival plot for three‑year CVmortality in dual poor responders and responders to one or both antiplatelet drugs is presented in Figure 2. Discussion We can summarize the findings of our study as follows: (a) amongSTEMI patients under theageof 45yearswhounderwent PCI, 47.8% have a poor response to aspirin and/or clopidogrel; (b) poor responders to both aspirin and clopidogrel had a significantly higher level of MACE at 3 years follow-up compared with dual responders. Furthermore, secondary outcome analysis showed a significant difference in cardiac mortality and TVR between these two groups; (c) after adjustment for potential confounders, it was found that being a dual poor responder was one of the independent predictors of MACE. Moreover, the Kaplan-Meier survival plot for three-year CV mortality showed poor prognosis of dual poor responder patients (log rank<0.001). Antiplatelet resistance is a multifactorial phenomenon that has been studied in many populations with different methods. Therefore, the presence of variable results in the literature makes it difficult to compare our results with those of other studies. However, the lack of previous studies in young STEMI patients and long-term results of the dual antiplatelet resistance in this group make this study unique and valuable. There is no single way to initiate thrombotic events; therefore, inhibition of a single pathway does not prevent all thrombotic complications. In addition, in some patients, the sensitivity of aspirin and clopidogrel is low, resulting in clinical complications. Therefore, several studies have been conducted to determine the clinical implications of being poor responders to aspirin and/or clopidogrel. In a meta‑analysis of 1,813 patients with 12 studies examining the effect of aspirin resistance on prognosis, the mean biochemical aspirin resistance was 27% and the odds ratio for MACE was 3.8 (95% CI: 2.3‑6.1) in patients with aspirin resistance. 4 In another meta‑analysis of 2,930 patients, aspirin resistance was detected in 28% of these patients, cardiovascular events in 41% (OR 3.85, 95% CI: 3.08‑4.80), mortality in 5.7% (OR 5.99, 95% CI: 2.28‑15.72) and acute coronary syndrome in 39.4% (OR 4.06, 95% CI: 2.96-5.56). 5 In another study with patients with symptomatic peripheral artery disease, aspirin resistance was found as an independent predictor of adverse cardiovascular events with 2.48 hazard ratio. 10 In a study on non-STEMI patients, aspirin resistants were at significantly higher risk of cardiovascular death with hazard ratio of 2.6 (95% CI 1.6‑4.3) than aspirin sensitives (23.1% versus 9.6%). 11 Although all of the above studies showed an association of aspirin resistance with cardiovascular events, in our study, the increase in MACE in poor responders to aspirin did not reach statistical significance (15% versus 6%, p = 0.217). These differences may be explained by several factors. Firstly, the lack of statistical significance may have been caused by our smaller sample size. Secondly, these studies were carried out in different groups of patients using different methods. Besides, in these studies above mentioned, there was no analysis of a subgroup of young patients. We may speculate that aspirin resistance in this group of patients may not affect cardiovascular events due to different pathophysiological mechanisms. However, synergistic contribution to the increase in cardiovascular events with clopidogrel responsiveness was detected in our study. Larger studies need to clarify this conflict. 140