The association of body mass index with long-term clinical outcomes after ticagrelor monotherapy following abbreviated dual antiplatelet therapy in patients undergoing percutaneous coronary intervention: a prespecified sub-analysis of the GLOBAL LEADERS Trial

Background The efficacy of antiplatelet therapies following percutaneous coronary intervention (PCI) may be affected by body mass index (BMI). Methods and results This is a prespecified subgroup analysis of the GLOBAL LEADERS trial, a prospective, multicenter, open-label, randomized controlled trial in an all-comer population undergoing PCI, comparing the experimental strategy (23-month ticagrelor monotherapy following 1-month dual antiplatelet therapy [DAPT]) with a reference regimen (12-month aspirin monotherapy following 12-month DAPT). A total of 15,968 patients were stratified by baseline BMI with prespecified threshold of 27 kg/m2. Of those, 6973 (43.7%) patients with a BMI < 27 kg/m2 had a higher risk of all-cause mortality at 2 years than those with BMI ≥ 27 kg/m2 (adjusted HR 1.24, 95% CI 1.02–1.49). At 2 years, the rates of the primary endpoint (all-cause mortality or new Q-wave myocardial infarction) were similar between treatment strategies in either BMI group (pinteraction = 0.51). In acute coronary syndrome, however, the experimental strategy was associated with significant reduction of the primary endpoint compared to the reference strategy in patients with BMI < 27 kg/m2 (HR 0.69, 95% CI 0.51–0.94), but not in the ones with BMI ≥ 27 kg/m2 (pinteraction = 0.047). In chronic coronary syndrome, there was no between-group difference in the efficacy and safety of the two antiplatelet strategies. Conclusions Overall, BMI did not influence the treatment effect seen with ticagrelor monotherapy; however, a beneficial effect of ticagrelor monotherapy was seen in ACS patients with BMI < 27 kg/m2. Trial registration The trial has been registered with ClinicalTrials.gov, Number NCT01813435. Graphic abstract Electronic supplementary material The online version of this article (10.1007/s00392-020-01604-1) contains supplementary material, which is available to authorized users.


Introduction
Body mass index (BMI) is simple to calculate and consequently used as an indicator of general adiposity [1]. Although obesity is well recognized as a major risk factor of cardiovascular disease (CVD) [2], numerous studies have demonstrated a paradoxical association between higher BMI and lower risk of adverse events in patients with established CVD, even after adjusting for confounding factors. In this phenomenon, dubbed the "obesity paradox" [3][4][5], patients with lower or even normal BMI have a higher risk of both ischemic and bleeding events after percutaneous coronary intervention (PCI) compared to those who are overweight [6]. To date, however, no tailored antiplatelet strategy has been recommended for these patients [7].
It is recognized that the efficacy of platelet inhibition due to antiplatelet therapy including novel potent P2Y12 inhibitors could be associated with a patient's BMI [8]. In other words, high or low BMI could lead to an inappropriate balance between anti-ischemic and bleeding risks [9][10][11]. Therefore, assessment of different antiplatelet strategies after PCI, stratified according to BMI, may provide additional insight into patients with a "high-risk" BMI.
The GLOBAL LEADERS trial compared the experimental antiplatelet regimen with 23-month ticagrelor monotherapy, to the reference regimen of conventional 12-month dual antiplatelet therapy (DAPT) followed by 12-month aspirin in an all-comers PCI population [12]. The superiority of the experimental strategy at 2 years was not demonstrated in the parent trial. However, non-specified secondary analyses suggested the potential efficacy of this novel experimental regimen in some specific patient subgroups [12][13][14][15]. To unravel the complex intricacies of the GLOBAL LEADERS trial, the present study aims to investigate the clinical impact of BMI on the novel antiplatelet strategy with ticagrelor monotherapy in patients undergoing PCI.

Study design
This study is a prespecified subgroup analysis of the GLOBAL LEADERS trial [16]. The GLOBAL LEADERS trial [12] is a multi-center, prospective, open-label randomized controlled trial in an all-comer population with no restriction regarding clinical presentation, complexity of the lesions or number of stents used (NCT01813435). Details of the study design and protocol have been reported elsewhere [16]. In brief, the trial randomly assigned patients before PCI to either (i) the experimental strategy with 1-month DAPT (aspirin 75-100 mg daily and ticagrelor 90 mg twice daily) followed by 23-month ticagrelor 90 mg twice daily monotherapy, or (ii) the reference regimen with 12-month DAPT [aspirin 75-100 mg daily and either ticagrelor 90 mg twice daily for acute coronary syndromes (ACS: unstable angina, non ST-elevation myocardial infarction, and ST elevation myocardial infarction) or clopidogrel 75 mg daily for chronic coronary syndromes (CCS)] followed by 12-month aspirin 75-100 mg daily monotherapy, respectively. All target lesions were treated by default with biolimus A9-eluting stents (BioMatrix, Biosensors, Europe). The trial was approved by the institutional review board at each center and followed the ethical principles of the Declaration of Helsinki. All the patients gave written informed consent prior to participation in the trial.

Patients population and study endpoints
The patient's baseline BMI was calculated as weight in kilograms divided by height in meters squared collected at the time of randomization. Patients were divided into two groups according to a threshold BMI of 27.0 kg/ m 2 , which was prespecified in the design paper [16] and adopted by reference to previous publications [17,18], and also corresponds to the median value of BMI in the present population. In each BMI group, clinical, demographic, angiographic, and procedural characteristics were compared between patients who received the experimental and reference antiplatelet regimen.
The primary endpoint of this study was the composite of all-cause mortality and new Q-wave myocardial infarction (MI) up to 2 years. Deaths from any cause were ascertained without the need for adjudication [19,20]. Q-wave MI was centrally adjudicated by an independent electrocardiogram core lab and defined in accordance with the Minnesota classification (new major Q-QS wave abnormalities) or by the appearance of a new left bundle branch block in conjunction with abnormal biomarkers [21]. The secondary safety endpoint was major bleeding events according to the Bleeding Academic Research Consortium (BARC) criteria type 3 or 5 [22]. Additional endpoints included stroke (ischemic or hemorrhagic), BARC type 2 bleeding, definite stent thrombosis according to Academic Research Consortium (ARC) definition [23], and the composite of all-cause mortality, any stroke, and new Q-wave MI [16]. The composite endpoints were analyzed according to time-to-first event analysis.

Statistical analysis
Continuous variables are reported as mean ± standard deviations (SD) or median and interquartile range (IQR), and are compared using Student's t tests or Mann-Whitney U test, respectively. Categorical variables are reported as percentages and numbers and are compared using Chi-square or Fisher's exact test as appropriate.
Association between baseline BMI as a continuous variable and adverse outcomes including the primary and secondary endpoint is depicted using restricted cubic spline function from the adjusted Cox regression model. Kaplan-Meier method is used to estimate the cumulative rates of clinical events and log-rank test is performed to examine the differences between groups. The effect of BMI on the outcomes is assessed in the unadjusted and adjusted Cox proportional hazards model. The clinical outcomes were compared stratified according to both the prespecified threshold of 27 kg/ m 2 and the World Health Organization (WHO) classification: underweight (BMI < 18.5 kg/m 2 ), normal weight (BMI 18.5-24.9 kg/m 2 ), overweight (BMI 25.0-29.9 kg/m 2 ), and obesity (BMI ≥ 30 kg/m 2 ). The covariables in the adjusted model are listed in Fig. 2 and Table 2, which were selected based on previous knowledge and literature [24,25]. Variance inflation factor (VIF) of covariables are calculated to confirm the absence of multicollinearity. We also performed the receiver operating characteristic (ROC) analysis to detect the optimal cutoff value of BMI for predicting the primary endpoint according to the Youden index. The treatment effect of the experimental vs. the reference strategy between subgroups is estimated with an unadjusted Cox regression model.
Because different P2Y 12 inhibitors in the reference group were used depending on clinical presentation (ticagrelor for ACS or clopidogrel for CCS), the prespecified stratified analysis according to clinical presentation is performed. In addition, landmark analyses are reported using the prespecified time points of 1 year (at the time of the planned cessation of a P2Y 12 inhibitor in the reference strategy).
Statistical significance was considered if two-sided p value was less than or equal 0.05. All analyses were performed in SPSS Statistics, version 26 (IBM Corp., Armonk, 281 N.Y., USA) and R software version 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria).

Baseline characteristics
A comparison of the baseline characteristics between the two BMI groups is shown in Table 1. Compared to those with a BMI ≥ 27 kg/m 2 , patients with BMI < 27 kg/m 2 were older; more likely to present with ACS; had lower prevalence of diabetes, hypertension, hypercholesterolemia, renal impairment, previous MI, and previous PCI; were more likely to be smokers, and had a higher prevalence of peripheral artery disease. Patients with a BMI < 27 kg/m 2 had higher rates of PCI in the left anterior descending artery compared to those with a BMI ≥ 27 kg/m 2 . There were no significant differences in the rates of radial access between the two BMI groups.
Baseline patient characteristics were comparable and well balanced between the experimental and reference arms in each BMI group as shown in online Table 1.
The comparison of clinical outcomes according to WHO classification is shown in online Table 2. After adjusting confounding factors, the risk of all-cause mortality at 2 years was significantly lower in overweight patients (HR 0.75, 95% CI 0.60-0.93, p = 0.010) or obese patients (HR 0.74, 95% CI 0.57-0.95, p = 0.020) than normal weight patients. The correlation between the risks for the primary or secondary endpoint and BMI as a continuous variable showed reverse J-shape curves, as shown in Fig. 2 and online Table 3. The ROC analysis demonstrated that 25.4 kg/m 2 was the optimal cutoff value of BMI for predicting the primary endpoint.

Impact of BMI on antiplatelet strategy
The comparison of 2-year outcomes between the experimental and reference arms are shown in Fig. 3. At the 2-year follow-up, there was no statistically significant treatment effect on the primary endpoint of all-cause mortality or new Q-wave MI between the experimental and reference arm in patients with a BMI < 27 kg/m 2 (4.9% vs. 4  Blue curve with light blue area indicates adjusted hazard ratio with 95% CI for composite of all-cause mortality and new Q-wave MI at 2-year according to BMI with reference of 27 kg/m 2 . Red curve with light red area indicates adjusted hazard ratio with 95% CI for BARC type 3 or 5 bleeding according to BMI with reference of 27 kg/m 2 . The number of knots for the cubic spline curves were three in each model. Adjusted covariates for all-cause mortality or new Q-wave MI are age (years), sex, clinical presentation (ACS or CCS), diabetes mellitus, hypertension, hypercholesteremia, PVD, COPD, renal impairment, previous MI, previous PCI, and previous CABG. Adjusted covariates for BARC type 3 or 5 bleeding are age (years), sex, clinical presentation (ACS or CCS), diabetes mellitus, previous bleeding, renal impairment, anemia according to WHO classification, and radial access in the index procedure. BMI body mass index, ACS acute coronary syndromes, CCS chronic coronary syndromes, HR hazard ratio, CI confidence interval, MI myocardial infarction, BARC Bleeding Academic Research Consortium, PVD peripheral vascular disease, COPD chronic obstructive pulmonary disease, PCI percutaneous coronary intervention, CABG coronary artery bypass graft, WHO World Health Organization effect related to the experimental strategy with ticagrelor monotherapy with regard to other clinical outcomes at 2 years in each BMI group (Fig. 3).

Comparison of clinical outcomes between the two antiplatelet strategies in each BMI group stratified according to their clinical presentation
Clinical outcomes stratified according to clinical  Table 3. In patients with BMI < 27 kg/m 2 and ACS, the observed lower rates of events with the experimental treatment were mainly driven by the lower incidence of all-cause mortality, BARC 3 or 5 bleeding, or BARC 2 bleeding during the first year after index PCI; a landmark analysis after 1 year did not show any treatment effect in the second year ( Fig. 4 and Suppl. Fig. 1).

Impact of BMI on antiplatelet strategy in the setting of CCS
In the setting of CCS, there was no difference between the reference and the experimental arm regardless of BMI group in terms of the primary endpoint (

Discussion
In the context of a neutral trial, all presented findings should be viewed strictly as hypothesis generating. Nevertheless, for the first time to our knowledge, we have observed a differential effect of ticagrelor monotherapy, when compared with ticagrelor and aspirin, in relation to baseline BMI in patients with ACS-a subgroup who between 31 and 365 days after randomization were assigned to receive either ticagrelor alone, or in combination with aspirin by the GLOBAL LEADERS trial protocol [14].
In the present study, the potential beneficial effect of the experimental strategy was only observed in patients with ACS who had a BMI < 27 kg/m 2 , and was not seen in those with higher BMIs. Platelet hyper-reactivity and activation plays a central role in the progression of atherothrombosis and is the result of interactions of many adaptive responses to obesity: insulin resistance, inflammation, oxidative stress, and endothelial dysfunction [2,26].
Although a plausible pharmacodynamic explanation still needs to be determined, it can be explained by some hypothesis. Patients with high BMI and ACS are more likely to have a prothrombotic state, partly linked to dysglycemia and proinflammatory effects of metabolic syndrome. In the PLATO study, the beneficial effect of potent antiplatelet regimen with ticagrelor was mainly observed when the patient's body weight was higher than the median value for their sex (p interaction = 0.04) [27] In addition, the substudy of the PLATO trial showed that impaired fibrinolysis was an independent predictor of cardiovascular death and was more common in patients with diabetes mellitus and/or higher BMI [28]. In those situations, strong agonist stimulation such as via platelet thrombin receptors as well as via collagen-mediated thromboxane A2 release could overwhelm the effects of potent platelet P2Y12 inhibition.
Furthermore, among obese patients, cyclo-oxygenase (COX) inhibition, which is achieved exclusively by aspirin, may play a more vital role than in non-obese patients. It has been demonstrated that excess adipose tissue is associated with an increased platelet turnover, leading to unacetylated COX-1 and COX-2 in newly formed platelets with subsequent excessive thromboxane formation [29,30]. This is further exacerbated by extra-platelet sources of thromboxane in obese patients driven by inflammatory triggers and enhanced lipid peroxidation, resulting in activation of platelets by a mechanism bypassing COX-1 acetylation or through limiting COX-isozyme acetylation by aspirin [29,30]. Consequently, ticagrelor monotherapy may provide insufficient antithrombotic effect compared to ticagrelor plus aspirin in obese patients with prothrombotic states [31,32]. In other words, it is possible that the balance of inhibition of platelet thromboxane A 2 (TXA 2 ) release vs inhibition of prostacyclin formation with standard DAPT regimens is more favorable in obese patients than in non-obese patients [33]. More than a decade ago, and before the availability of prasugrel and ticagrelor, high BMI was associated with stent thrombosis in the all-comers LEADERS trial, leading to calls for the dose of clopidogrel to be weight adjusted [34].
On the other hand, in patients with ACS and a BMI < 27 kg/m 2 , the potentially favorable results of ticagrelor monotherapy compared to DAPT during the first year require some cautious interpretation. Previously, Leadbeater, et al. and Kirkby, et al. demonstrated that sufficient inhibition of the TXA 2 pathway can be achieved with the sole use of a strong P2Y 12 inhibitor such as prasugrel or ticagrelor without aspirin [35]; however, these findings were not seen consistently [36,37], although, this may have been due to the heterogeneity of the studied populations. Whereas the possibility of a play of chance remains, our results might suggest that in non-obese patients with higher responsiveness to P2Y 12 inhibitors [38,39], sufficient inhibition of TXA 2 pathway could be achieved by ticagrelor monotherapy, and adding aspirin could be associated with higher risks of ischemic and bleeding events than in obese patients [40]. In summary, the BMI-adjusted antiplatelet strategy with or without aspirin may be effective in ACS patients undergoing PCI, and the aspirin-free strategy with a potent P2Y12 inhibitor could be beneficial for those with a relatively low BMI.
In patients with CCS, the experimental strategy resulted in no significant difference in any clinical outcomes, but did lead to numerically higher rates of major bleeding in patients irrespective of their BMI group. Although Orme et al. reported that lower platelet activity achieved with ticagrelor, compared with clopidogrel, also occurred in patients with CCS [41], our results might suggest that the anti-ischemic effect of potent P2Y 12 inhibitors may not be required in low ischemic-risk settings such as patients with CCS.
Finally, in our cohort, and consistent with previous studies, we observed the "obesity paradox" with the reverse J-shape association between adverse events and BMI as a continuous variable [6,42,43]. In addition, normal weight patients had a higher risk of all-cause mortality  (Table 2). Given the fact that most patients with a BMI < 27 kg/m 2 in this study could be categorized as "normal weight" in the WHO classification (Fig. 2), our results may encourage the efficacy of the novel P2Y12 inhibitor monotherapy for those high-risk "normal weight" patients.

Limitations
The present study needs to be interpreted in light of the following limitations. First, the present study consists of two prespecified subgroup analyses of a randomized controlled study with multiple testing (BMI and clinical presentation). Because in the GLOBAL LEADERS trial two different P2Y 12 inhibitors are used in the reference group depending on the clinical presentation of ACS (ticagrelor) or CCS (clopidogrel), multiple analyses according to the clinical presentation have to be performed to evaluate specifically the treatment effect strictly. However, the results could be a play of chance and they should be considered as hypothesis generating. Second, BMI data were only available at the time of randomization. BMI can change depending on weight gain or loss during follow-up [44]. Third, in past trials reporting the "obesity paradox", the current threshold of BMI (27 kg/m 2 ) prespecified in the design paper and based on a recent publication [16] was not widely used and was higher than the optimal cutoff value of 25.4 kg/m 2 for stratifying with the risk of the primary endpoint in this study. In addition, the WHO classification is somewhat different. Indeed, the WHO classification classified patients into four or six categories, resulting in lower and uneven statistical power among these groups. Our threshold was close to the median value of 27.68 kg/m 2 in the current study, which allows uniform statistical power in each group. Fourth, in this trial all endpoints were site reported without a clinical adjudication committee for serious adverse events due to limited financial resources. However, the GLASSY study [45], which is a prespecified ancillary study of the GLOBAL LEADERS trial with event adjudication by an independent clinical event committee, confirmed the consistent results with those of site reported.

Conclusion
There was no overall treatment effect of experimental ticagrelor monotherapy versus standard DAPT strategy between the groups with high or low baseline BMI. However, a beneficial treatment effect on ischemic events (primary endpoints of all-cause mortality or new Q-wave MI) without trade-off in bleeding (BARC type 3 or 5 bleeding) of the experimental treatment with ticagrelor monotherapy was observed in patients presenting with ACS with BMI < 27 kg/m 2 , which was not seen in patients with BMI ≥ 27 kg/m 2 . Our results suggest the potential benefit of a novel antiplatelet monotherapy regimen in targeting non-obese ACS patients.
Funding GLOBAL LEADERS study was sponsored by the European Clinical Research. Institute, which received funding from Biosensors International, AstraZeneca and the Medicines Company. The study funders had no role in trial design; data collection, analysis or interpretation; or writing of the report. Piek reports personal fees and non-financial support from Philips/Volcano, outside the submitted work. Dr. Hamm reports personal fees from AstraZeneca, outside the submitted work. Dr. Steg reports grants and personal fees from Bayer/ Janssen, grants and personal fees from Merck, grants and personal fees from Sanofi, grants and personal fees from Amarin, personal fees from Amgen, personal fees from Bristol Myers Squibb, personal fees from Boehringer-Ingelheim, personal fees from Pfizer, personal fees from Novartis, personal fees from Regeneron, personal fees from Lilly, personal fees from AstraZeneca, and grants and personal fees from Servier, outside the submitted work. Dr. Jüni reports research grants to the institution from Astra Zeneca, Biotronik, Biosensors International, Eli Lilly and The Medicines Company, and serves as unpaid member of the steering group of trials funded by Astra Zeneca, Biotronik, Biosensors, St. Jude Medical and The Medicines Company. Dr. Storey reports personal fees from Bayer, personal fees from Bristol-Myers Squibb/ Pfizer, grants and personal fees from AstraZeneca, personal fees from Novartis, personal fees from Idorsia, grants and personal fees from Thromboserin, personal fees from Haemonetics, personal fees from Amgen, grants and personal fees from Glycardial Diagnostics, personal fees from Portola, and personal fees from Medscape, outside the submitted work. Dr. Valgimigli reports personal fees from Astra Zeneca, grants and personal fees from Terumo, personal fees from Alvimedica/CID, personal fees from Abbott Vascular, personal fees from Daiichi Sankyo, personal fees from Opsens, personal fees from Bayer, personal fees from CoreFLOW, personal fees from IDORSIA PHAR-MACEUTICALS LTD, personal fees from Universität Basel | Dept. Klinische Forschung, personal fees from Vifor, personal fees from Bristol Myers Squib SA, and personal fees from iVascular, outside the submitted work. Dr. de Windecker received research and educational grants to the institution from Amgen, Abbott, Boston Scientific, Biotronik, Bayer, BMS, CSL Behring, Medtronic, Edwards Lifesciences, and Polares and Sinomed, outside the submitted work. Dr. Vranckx received personal fees from Astra Zeneca, personal fees from Bayer Health Care, personal fees from Daiichi Sankio, personal fees from Terumo, and personal fees from CLS Behring, outside the submitted work. Dr. Serruys reports personal fees from Biosensors, personal fees from Medtronic, personal fees from Micel Technologies, personal fees from Sinomedical Sciences Technology, personal fees from Philips/ Volcano, personal fees from Xeltis, and personal fees from HeartFlow, outside the submitted work. All other authors declare no competing interests.

Compliance with ethical standards
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.