https://orcid.org/0000-0002-9860-3584
Abstract
In patients with prior myocardial infarction (MI) on aspirin, the addition of ticagrelor reduces ischaemic risk but increases bleeding risk. The simultaneous assessment of baseline ischaemic and bleeding risk may assist clinicians in selecting patients who are most likely to have a favourable risk/benefit profile with long-term ticagrelor.
PEGASUS-TIMI 54 randomized 21 162 prior MI patients, 13 956 of which to the approved 60 mg dose or placebo and who had all necessary data. The primary efficacy endpoint was cardiovascular death, MI, or stroke, and the primary safety outcome was TIMI major bleeding; differences in Kaplan–Meier event rates at 3 years are presented. Post-hoc subgroups based on predictors of bleeding and ischaemic risk were merged into a selection algorithm. Patients were divided into four groups: those with a bleeding predictor (n = 2721, 19%) and then those without a bleeding predictor and either 0–1 ischaemic risk factor (IRF; n = 3004, 22%), 2 IRF (n = 4903, 35%), or ≥3 IRF (n = 3328, 24%). In patients at high bleeding risk, ticagrelor increased bleeding [absolute risk difference (ARD) +2.3%, 95% confidence interval (CI) 0.6, 3.9] and did not reduce the primary efficacy endpoint (ARD +0.08%, 95% CI −2.4 to 2.5). In patients at low bleeding risk, the ARDs in the primary efficacy endpoint with ticagrelor were −0.5% (−2.2, 1.3), −1.5% (−3.1, 0.02), and −2.6% (−5.0, −0.24, P = 0.03) in those with ≤1, 2, and 3 risk factors, respectively (P = 0.076 for trend across groups). There were significant trends for greater absolute risk reductions for cardiovascular death (P-trend 0.018), all-cause mortality (P-trend 0.027), and net outcomes (P-trend 0.037) with ticagrelor across these risk groups.
In a post-hoc exploratory analysis of patients with prior MI, long-term ticagrelor therapy appears to be best suited for those with prior MI with multiple IRFs at low bleeding risk.
NCT01225562 ClinicalTrials.gov
Characteristics (left) and distribution of patients (middle) in PEGASUS-TIMI 54 by categories of high bleeding risk, low bleeding risk with 0–1 ischaemic risk factors, low bleeding risk with 2 ischaemic risk factors, and low bleeding risk with ≥ 3 ischaemic risk factors and the absolute risk difference for the composite of cardiovascular death, myocardial infarction or stroke with ticagrelor vs. placebo over 3 years for each group (right). CAD, coronary artery disease; CV, cardiovascular; MI, myocardial infarction.
See the editorial comment for this article ‘Platelet antiaggregation after an acute coronary syndrome: what about the elderly?’, by K. Huber, https://doi.org/10.1093/eurheartj/ehac405.
Introduction
Patients with myocardial infarction (MI) are at heightened risk of ischaemic events even years after the event occurred. Intensive antiplatelet therapy with the combination of aspirin and a P2Y12 inhibitor (dual antiplatelet therapy or DAPT) is initiated at the time of MI to reduce this risk and is indicated for at least a year in patients who are not at elevated bleeding risk.1–3 Long-term use of DAPT has been shown to reduce ischaemic risk in patients with prior MI and PEGASUS-TIMI 54 prospectively demonstrated that long-term ticagrelor at a dose of 60 mg twice daily reduces major adverse cardiovascular events (MACEs) in this population.4 The benefit of this therapeutic approach has also extended to patients with diabetes and coronary disease who have never had MI, especially in those with prior percutaneous coronary revascularization.5,6
Strategies using long-term DAPT to reduce ischaemic risk after MI, however, also increase the risk of bleeding.4,7 Although this bleeding risk did not extend to irreversible harm events such as intracranial haemorrhage (ICH) or fatal bleeding in the trials, major bleeding is an undesirable event and can lead to hospitalization and cessation of beneficial therapies and is associated with worse outcomes including mortality.1,3,8–10 Such observations necessitate personalized application of these strategies rather than generalized use in all patients with prior MI.1–3,11 Although algorithms have been developed to enable selection of patients at lower bleeding risk, this approach is complicated by the observation that many predictors of bleeding are also predictors of ischaemic risk.8,11 In addition, other scoring systems developed in broader populations immediately after percutaneous coronary intervention (PCI) may not apply in long-term patients with prior MI some of whom were treated medically and where patient characteristics and comorbidities are likely more important drivers of risk relative to stent or procedural characteristics.11–13
Therefore, we developed a stepwise patient selection algorithm for long-term treatment with ticagrelor 60 mg twice daily utilizing subgroups previously observed to be at higher bleeding risk8 and ischaemic risk.14–19 We then evaluated bleeding, ischaemic risk, mortality, and net outcomes with ticagrelor 60 mg twice daily vs. placebo by risk group in PEGASUS-TIMI 54.
Methods
Study population
The PEGASUS-TIMI 54 trial described previously, randomized 21 162 patients with a prior history of spontaneous MI occurring 1–3 years prior to enrolment, who had at least one additional atherothrombotic risk factor (age ≥65 years, diabetes mellitus requiring medication, a second prior spontaneous MI, chronic renal dysfunction, or multivessel coronary artery disease) to ticagrelor 60 mg b.i.d., ticagrelor 90 mg b.i.d., or placebo, all on a background of low-dose (75–150 mg) aspirin.4 Patients were followed for a median of 33 months. Exclusion criteria included planned use of a P2Y12 receptor antagonist or anticoagulant therapy, a known bleeding disorder, history of stroke, a central nervous system tumour, gastrointestinal bleeding within the previous 6 months, or major surgery within the previous 30 days. Enrolling sites were requested to indicate if there was any history of bleeding leading to hospitalization. All patients had central laboratory testing for haemoglobin at baseline. Anaemia was defined as a haemoglobin ≤13.5 g/dL for men and ≤12.0 g/dL for women.8 Because this analysis is focused on an algorithm to select patients for long-term ticagrelor use in patients with an MI at least 1 year prior, the population was restricted to the 14 112 randomized to the approved dose of 60 mg twice daily or placebo. A total of 156 patients were excluded from the analyses due to missing ischaemic and/or bleeding risk factors with 13 956 included in the analysis cohort.
Endpoints
The primary efficacy endpoint was MACE, consisting of the composite of cardiovascular (CV) death, MI, or stroke.4 The primary safety endpoint was TIMI major bleeding defined as any ICH, or clinically overt signs of haemorrhage associated with a reduction in haemoglobin ≥5 g/dL (or, when haemoglobin was not available, a fall in haematocrit ≥15%), or fatal bleeding (a bleeding event that directly led to death within 7 days). TIMI minor bleeding was defined as any clinically overt haemorrhage (including that detected by imaging) that was associated with a fall in haemoglobin of 3 to <5 g/dL (or, when haemoglobin was not available, a fall in haematocrit of 9 to <15%). Haemoglobin measurements were adjusted for any packed red blood cells or whole blood given between baseline and post-transfusion measurement; transfusion of one unit of blood was assumed to result in an increase of 1 g/dL of haemoglobin. Bleeding events leading directly to mortality were classified as fatal bleeding events. All bleeding events, as well as their relationship to mortality, were adjudicated by a Clinical Events Committee blinded to treatment allocation. A net clinical benefit analysis was defined as a composite outcome consisting of CV death, MI, stroke, ICH, or fatal bleeding.
Patient selection algorithm
A post-hoc algorithm conceptually aligned with the risk stratification approach described by the European Society of Cardiology (ESC) guidelines for a second antithrombotic therapy was developed, including first identifying high-bleeding risk patients and then stratifying low-bleeding risk patients into thrombotic risk categories (one risk criteria and two or more risk criteria).2 Previously identified independent predictors of bleeding, but not ischaemic risk, were anaemia at baseline and prior history of spontaneous bleeding requiring hospitalization.8 Factors that were not independent predictors of bleeding or were predictors of both ischaemic risk and bleeding (e.g. age, kidney dysfunction) were excluded as published previously.8 Patients with at least one of these criteria had three-fold higher rates of TIMI major or minor bleeding with ticagrelor 60 mg vs. placebo compared with those with neither feature.8
The initial step in the stratification approach, therefore, was to exclude patients at high bleeding risk, defined by the presence of either or both of the bleeding risk predictors. Patients identified as low bleeding risk were stratified according to ischaemic risk factors. Because PEGASUS-TIMI 54 included medically managed patients (for whom complexity of coronary disease was unknown), an approach based solely on the number of clinical risk factors was utilized. Factors included the pre-specified enrichment criteria for the trial that have also been observed to be associated with greater ischaemic risk (see Supplementary material online, Figure S2) including type 2 diabetes mellitus,14 non-end-stage chronic kidney disease,17 multivessel coronary disease,18 and multiple prior MIs. In addition, more recent MI or P2Y12 inhibition which showed formal heterogeneity for benefit of ticagrelor and was the basis for the EU label19 and polyvascular disease (comorbid peripheral artery disease) which was associated with greater risk and more favourable risk benefit were included.16,20 A summary of risk factors chosen, criteria and rationale for exclusion are shown in Supplementary material online, Table S1. Low-bleeding risk patients were then grouped according to 0–1 ischaemic risk factor, 2 risk ischaemic factors, and ≥3 ischaemic risk factors in general alignment with lower and higher ischaemic risk, respectively.1
Statistics
As per the pre-specified statistical analysis plan, the primary efficacy analysis and net clinical outcome were conducted on an intention-to-treat basis, whereas safety analyses included all patients who underwent randomization and received at least one dose of study drug. Event rates for ticagrelor and placebo were estimated by Kaplan–Meier methods from baseline to 3 years and compared with the log-rank test. This was examined separately by treatment arm over time. Categorical variables were compared using χ2 tests and continuous variable with either a t test or Wilcoxon rank-sum test, as appropriate. The hazard ratio (HR) for given outcome of interest between ticagrelor and placebo was examined using a Cox regression model with treatment as a model term in each risk stratification group defined by bleeding and ischaemic risk profiles described in previous section. The proportional hazards assumption was examined and tested by scaled Schoenfeld residuals. In addition, the absolute risk differences (ARDs) between ticagrelor and placebo groups were calculated by subtracting the cumulative incidence in the placebo group from that of the treatment group. The cumulative incidence was defined by the complement of the Kaplan–Meier survival estimates. The null hypothesis that the gradient of risk differences across the level of subgroup is zero was tested by weighted least square regression, with weights being the inverse of variance.21 A sensitivity analysis was also performed estimating the hypothetical ARDs for the primary endpoint by applying the overall HR to the risk group specific baseline risk. All statistical computations were performed with the use of SAS software, version 9.4 (SAS Institute, Cary, NC, USA).
Results
Overall, of 14 112 patients randomized to ticagrelor 60 mg twice daily or placebo, a total of 2721 (19%) had at least one bleeding risk factor and were categorized as high bleeding risk. Among those without a bleeding predictor, 3004 (22% of the total) had only 0–1 ischaemic risk factor, 4903 (35%) had 2 ischaemic risk factors, and 3328 (24%) had 3 or more ischaemic risk factors. The baseline characteristics of these patients are shown in Table 1. When excluding the 156 patients with missing ischaemic and/or bleeding risk factors there were no consistent statistically significant imbalances in baseline characteristics between the placebo and ticagrelor arms across the three risk groups (see Supplementary material online, Tables S2A–C).
Baseline characteristics by bleeding and ischemic risk categories
| High bleeding risk N = 2721 | Low bleeding risk and ≤ 1 risk factor N = 3004 | Low bleeding risk and 2 risk factors N = 4903 | Low bleeding risk and ≥3 risk factors N = 3328 | P-value | |
|---|---|---|---|---|---|
| Age, years, median (IQR) | 68.0 (62.0, 74.0) | 67.0 (63.0, 72.0) | 62.0 (57.0, 69.0) | 64.0 (58.0, 71.0) | <0.0001 |
| Female sex, n (%) | 571 (21.0) | 760 (25.3) | 1107 (22.6) | 906 (27.2) | <0.0001 |
| Body mass index, kg/m2, median (IQR) | 27.1 (24.2, 30.4) | 27.6 (25.1, 30.2) | 28.0 (25.3, 31.2) | 28.7 (25.7, 32.2) | <0.0001 |
| Caucasian, n (%) | 2153 (79.1) | 2750 (91.5) | 4280 (87.3) | 2879 (86.5) | <0.0001 |
| Hypertension, n (%) | 2168 (79.7) | 2148 (71.5) | 3696 (75.4) | 2804 (84.3) | <0.0001 |
| Hypercholesterolemia, n (%) | 2031 (74.6) | 2225 (74.1) | 3777 (77.0) | 2681 (80.6) | <0.0001 |
| Current smoker, n (%) | 312 (11.5) | 454 (15.1) | 941 (19.2) | 614 (18.5) | <0.0001 |
| History of CHF, n (%) | 616 (22.6) | 560 (18.6) | 863 (17.6) | 759 (22.8) | <0.0001 |
| History of stroke or TIA, n (%) | 57 (2.1) | 46 (1.5) | 55 (1.1) | 64 (1.9) | 0.0033 |
| Prior angina, n (%) | 889 (32.7) | 862 (28.7) | 1435 (29.3) | 1158 (34.8) | <0.0001 |
| Prior CABG, n (%) | 165 (6.1) | 32 (1.1) | 104 (2.1) | 349 (10.5) | <0.0001 |
| History of PCI with stenting, n (%) | 2207 (81.1) | 2271 (75.6) | 4178 (85.2) | 2932 (88.1) | <0.0001 |
| Qualifying MI type, n (%) | 2114 (78.1) | 2182 (72.9) | 4043 (82.7) | 2818 (85.0) | <0.0001 |
| NSTEMI, n (%) | 1393 (51.3) | 1636 (54.5) | 2793 (57.0) | 1662 (50.0) | <0.0001 |
| STEMI, n (%) | 1155 (42.6) | 1158 (38.6) | 1825 (37.3) | 1482 (44.6) | |
| Unknown, n (%) | 166 (6.1) | 209 (7.0) | 279 (5.7) | 180 (5.4) | |
| Aspirin 75–100 mg at baseline, n (%) | 2640 (97.2) | 2918 (97.3) | 4774 (97.5) | 3238 (97.4) | 0.8603 |
| Statin use at baseline, n (%) | 2497 (91.8) | 2760 (91.9) | 4578 (93.4) | 3101 (93.2) | 0.0124 |
| Beta blocker use at baseline, n (%) | 2185 (80.3) | 2380 (79.2) | 4144 (84.5) | 2830 (85.0) | <0.0001 |
| ACE-I or ARB use at baseline, n (%) | 2204 (81.0) | 2315 (77.1) | 3921 (80.0) | 2757 (82.8) | <0.0001 |
| Risk Factors for Algorithm | |||||
| Qualifying MI ≥2 years, n (%) | 1038 (38.2) | 1978 (65.9) | 1446 (29.5) | 901 (27.1) | <0.0001 |
| Time from last dose of P2Y12 inhibitor, days, median (IQR) | 63.0 (1.0, 348.0) | 425.0 (102.0, 618.5) | 64.0 (1.0, 290.0) | 39.0 (1.0, 223.0) | <0.0001 |
| Multivessel disease, n (%) | 1582 (58.2) | 736 (24.5) | 3191 (65.1) | 2802 (84.2) | <0.0001 |
| History of diabetes, n (%) | 1118 (41.1) | 309 (10.3) | 1183 (24.1) | 1904 (57.2) | <0.0001 |
| Peripheral artery disease, n (%) | 209 (7.7) | 19 (0.6) | 76 (1.6) | 461 (13.9) | <0.0001 |
| eGFR <60 mL/min (MDRD), n (%) | 946 (35.0) | 152 (5.1) | 662 (13.5) | 1435 (43.1) | <0.0001 |
| Multiple prior MIs, n (%) | 469 (17.2) | 105 (3.5) | 479 (9.8) | 1280 (38.5) | <0.0001 |
| High bleeding risk N = 2721 | Low bleeding risk and ≤ 1 risk factor N = 3004 | Low bleeding risk and 2 risk factors N = 4903 | Low bleeding risk and ≥3 risk factors N = 3328 | P-value | |
|---|---|---|---|---|---|
| Age, years, median (IQR) | 68.0 (62.0, 74.0) | 67.0 (63.0, 72.0) | 62.0 (57.0, 69.0) | 64.0 (58.0, 71.0) | <0.0001 |
| Female sex, n (%) | 571 (21.0) | 760 (25.3) | 1107 (22.6) | 906 (27.2) | <0.0001 |
| Body mass index, kg/m2, median (IQR) | 27.1 (24.2, 30.4) | 27.6 (25.1, 30.2) | 28.0 (25.3, 31.2) | 28.7 (25.7, 32.2) | <0.0001 |
| Caucasian, n (%) | 2153 (79.1) | 2750 (91.5) | 4280 (87.3) | 2879 (86.5) | <0.0001 |
| Hypertension, n (%) | 2168 (79.7) | 2148 (71.5) | 3696 (75.4) | 2804 (84.3) | <0.0001 |
| Hypercholesterolemia, n (%) | 2031 (74.6) | 2225 (74.1) | 3777 (77.0) | 2681 (80.6) | <0.0001 |
| Current smoker, n (%) | 312 (11.5) | 454 (15.1) | 941 (19.2) | 614 (18.5) | <0.0001 |
| History of CHF, n (%) | 616 (22.6) | 560 (18.6) | 863 (17.6) | 759 (22.8) | <0.0001 |
| History of stroke or TIA, n (%) | 57 (2.1) | 46 (1.5) | 55 (1.1) | 64 (1.9) | 0.0033 |
| Prior angina, n (%) | 889 (32.7) | 862 (28.7) | 1435 (29.3) | 1158 (34.8) | <0.0001 |
| Prior CABG, n (%) | 165 (6.1) | 32 (1.1) | 104 (2.1) | 349 (10.5) | <0.0001 |
| History of PCI with stenting, n (%) | 2207 (81.1) | 2271 (75.6) | 4178 (85.2) | 2932 (88.1) | <0.0001 |
| Qualifying MI type, n (%) | 2114 (78.1) | 2182 (72.9) | 4043 (82.7) | 2818 (85.0) | <0.0001 |
| NSTEMI, n (%) | 1393 (51.3) | 1636 (54.5) | 2793 (57.0) | 1662 (50.0) | <0.0001 |
| STEMI, n (%) | 1155 (42.6) | 1158 (38.6) | 1825 (37.3) | 1482 (44.6) | |
| Unknown, n (%) | 166 (6.1) | 209 (7.0) | 279 (5.7) | 180 (5.4) | |
| Aspirin 75–100 mg at baseline, n (%) | 2640 (97.2) | 2918 (97.3) | 4774 (97.5) | 3238 (97.4) | 0.8603 |
| Statin use at baseline, n (%) | 2497 (91.8) | 2760 (91.9) | 4578 (93.4) | 3101 (93.2) | 0.0124 |
| Beta blocker use at baseline, n (%) | 2185 (80.3) | 2380 (79.2) | 4144 (84.5) | 2830 (85.0) | <0.0001 |
| ACE-I or ARB use at baseline, n (%) | 2204 (81.0) | 2315 (77.1) | 3921 (80.0) | 2757 (82.8) | <0.0001 |
| Risk Factors for Algorithm | |||||
| Qualifying MI ≥2 years, n (%) | 1038 (38.2) | 1978 (65.9) | 1446 (29.5) | 901 (27.1) | <0.0001 |
| Time from last dose of P2Y12 inhibitor, days, median (IQR) | 63.0 (1.0, 348.0) | 425.0 (102.0, 618.5) | 64.0 (1.0, 290.0) | 39.0 (1.0, 223.0) | <0.0001 |
| Multivessel disease, n (%) | 1582 (58.2) | 736 (24.5) | 3191 (65.1) | 2802 (84.2) | <0.0001 |
| History of diabetes, n (%) | 1118 (41.1) | 309 (10.3) | 1183 (24.1) | 1904 (57.2) | <0.0001 |
| Peripheral artery disease, n (%) | 209 (7.7) | 19 (0.6) | 76 (1.6) | 461 (13.9) | <0.0001 |
| eGFR <60 mL/min (MDRD), n (%) | 946 (35.0) | 152 (5.1) | 662 (13.5) | 1435 (43.1) | <0.0001 |
| Multiple prior MIs, n (%) | 469 (17.2) | 105 (3.5) | 479 (9.8) | 1280 (38.5) | <0.0001 |
ACE-I, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; CABG, coronary artery bypass graft; CHF, congestive heart failure; eGFR, estimated glomerular filtration rate; IQR, interquartile range; MDRD, Modification of Diet in Renal Disease; MI, myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction; PCI, percutaneous coronary intervention; STEMI, ST-elevation myocardial infarction; TIA, transient ischaemic attack.
Baseline characteristics by bleeding and ischemic risk categories
| High bleeding risk N = 2721 | Low bleeding risk and ≤ 1 risk factor N = 3004 | Low bleeding risk and 2 risk factors N = 4903 | Low bleeding risk and ≥3 risk factors N = 3328 | P-value | |
|---|---|---|---|---|---|
| Age, years, median (IQR) | 68.0 (62.0, 74.0) | 67.0 (63.0, 72.0) | 62.0 (57.0, 69.0) | 64.0 (58.0, 71.0) | <0.0001 |
| Female sex, n (%) | 571 (21.0) | 760 (25.3) | 1107 (22.6) | 906 (27.2) | <0.0001 |
| Body mass index, kg/m2, median (IQR) | 27.1 (24.2, 30.4) | 27.6 (25.1, 30.2) | 28.0 (25.3, 31.2) | 28.7 (25.7, 32.2) | <0.0001 |
| Caucasian, n (%) | 2153 (79.1) | 2750 (91.5) | 4280 (87.3) | 2879 (86.5) | <0.0001 |
| Hypertension, n (%) | 2168 (79.7) | 2148 (71.5) | 3696 (75.4) | 2804 (84.3) | <0.0001 |
| Hypercholesterolemia, n (%) | 2031 (74.6) | 2225 (74.1) | 3777 (77.0) | 2681 (80.6) | <0.0001 |
| Current smoker, n (%) | 312 (11.5) | 454 (15.1) | 941 (19.2) | 614 (18.5) | <0.0001 |
| History of CHF, n (%) | 616 (22.6) | 560 (18.6) | 863 (17.6) | 759 (22.8) | <0.0001 |
| History of stroke or TIA, n (%) | 57 (2.1) | 46 (1.5) | 55 (1.1) | 64 (1.9) | 0.0033 |
| Prior angina, n (%) | 889 (32.7) | 862 (28.7) | 1435 (29.3) | 1158 (34.8) | <0.0001 |
| Prior CABG, n (%) | 165 (6.1) | 32 (1.1) | 104 (2.1) | 349 (10.5) | <0.0001 |
| History of PCI with stenting, n (%) | 2207 (81.1) | 2271 (75.6) | 4178 (85.2) | 2932 (88.1) | <0.0001 |
| Qualifying MI type, n (%) | 2114 (78.1) | 2182 (72.9) | 4043 (82.7) | 2818 (85.0) | <0.0001 |
| NSTEMI, n (%) | 1393 (51.3) | 1636 (54.5) | 2793 (57.0) | 1662 (50.0) | <0.0001 |
| STEMI, n (%) | 1155 (42.6) | 1158 (38.6) | 1825 (37.3) | 1482 (44.6) | |
| Unknown, n (%) | 166 (6.1) | 209 (7.0) | 279 (5.7) | 180 (5.4) | |
| Aspirin 75–100 mg at baseline, n (%) | 2640 (97.2) | 2918 (97.3) | 4774 (97.5) | 3238 (97.4) | 0.8603 |
| Statin use at baseline, n (%) | 2497 (91.8) | 2760 (91.9) | 4578 (93.4) | 3101 (93.2) | 0.0124 |
| Beta blocker use at baseline, n (%) | 2185 (80.3) | 2380 (79.2) | 4144 (84.5) | 2830 (85.0) | <0.0001 |
| ACE-I or ARB use at baseline, n (%) | 2204 (81.0) | 2315 (77.1) | 3921 (80.0) | 2757 (82.8) | <0.0001 |
| Risk Factors for Algorithm | |||||
| Qualifying MI ≥2 years, n (%) | 1038 (38.2) | 1978 (65.9) | 1446 (29.5) | 901 (27.1) | <0.0001 |
| Time from last dose of P2Y12 inhibitor, days, median (IQR) | 63.0 (1.0, 348.0) | 425.0 (102.0, 618.5) | 64.0 (1.0, 290.0) | 39.0 (1.0, 223.0) | <0.0001 |
| Multivessel disease, n (%) | 1582 (58.2) | 736 (24.5) | 3191 (65.1) | 2802 (84.2) | <0.0001 |
| History of diabetes, n (%) | 1118 (41.1) | 309 (10.3) | 1183 (24.1) | 1904 (57.2) | <0.0001 |
| Peripheral artery disease, n (%) | 209 (7.7) | 19 (0.6) | 76 (1.6) | 461 (13.9) | <0.0001 |
| eGFR <60 mL/min (MDRD), n (%) | 946 (35.0) | 152 (5.1) | 662 (13.5) | 1435 (43.1) | <0.0001 |
| Multiple prior MIs, n (%) | 469 (17.2) | 105 (3.5) | 479 (9.8) | 1280 (38.5) | <0.0001 |
| High bleeding risk N = 2721 | Low bleeding risk and ≤ 1 risk factor N = 3004 | Low bleeding risk and 2 risk factors N = 4903 | Low bleeding risk and ≥3 risk factors N = 3328 | P-value | |
|---|---|---|---|---|---|
| Age, years, median (IQR) | 68.0 (62.0, 74.0) | 67.0 (63.0, 72.0) | 62.0 (57.0, 69.0) | 64.0 (58.0, 71.0) | <0.0001 |
| Female sex, n (%) | 571 (21.0) | 760 (25.3) | 1107 (22.6) | 906 (27.2) | <0.0001 |
| Body mass index, kg/m2, median (IQR) | 27.1 (24.2, 30.4) | 27.6 (25.1, 30.2) | 28.0 (25.3, 31.2) | 28.7 (25.7, 32.2) | <0.0001 |
| Caucasian, n (%) | 2153 (79.1) | 2750 (91.5) | 4280 (87.3) | 2879 (86.5) | <0.0001 |
| Hypertension, n (%) | 2168 (79.7) | 2148 (71.5) | 3696 (75.4) | 2804 (84.3) | <0.0001 |
| Hypercholesterolemia, n (%) | 2031 (74.6) | 2225 (74.1) | 3777 (77.0) | 2681 (80.6) | <0.0001 |
| Current smoker, n (%) | 312 (11.5) | 454 (15.1) | 941 (19.2) | 614 (18.5) | <0.0001 |
| History of CHF, n (%) | 616 (22.6) | 560 (18.6) | 863 (17.6) | 759 (22.8) | <0.0001 |
| History of stroke or TIA, n (%) | 57 (2.1) | 46 (1.5) | 55 (1.1) | 64 (1.9) | 0.0033 |
| Prior angina, n (%) | 889 (32.7) | 862 (28.7) | 1435 (29.3) | 1158 (34.8) | <0.0001 |
| Prior CABG, n (%) | 165 (6.1) | 32 (1.1) | 104 (2.1) | 349 (10.5) | <0.0001 |
| History of PCI with stenting, n (%) | 2207 (81.1) | 2271 (75.6) | 4178 (85.2) | 2932 (88.1) | <0.0001 |
| Qualifying MI type, n (%) | 2114 (78.1) | 2182 (72.9) | 4043 (82.7) | 2818 (85.0) | <0.0001 |
| NSTEMI, n (%) | 1393 (51.3) | 1636 (54.5) | 2793 (57.0) | 1662 (50.0) | <0.0001 |
| STEMI, n (%) | 1155 (42.6) | 1158 (38.6) | 1825 (37.3) | 1482 (44.6) | |
| Unknown, n (%) | 166 (6.1) | 209 (7.0) | 279 (5.7) | 180 (5.4) | |
| Aspirin 75–100 mg at baseline, n (%) | 2640 (97.2) | 2918 (97.3) | 4774 (97.5) | 3238 (97.4) | 0.8603 |
| Statin use at baseline, n (%) | 2497 (91.8) | 2760 (91.9) | 4578 (93.4) | 3101 (93.2) | 0.0124 |
| Beta blocker use at baseline, n (%) | 2185 (80.3) | 2380 (79.2) | 4144 (84.5) | 2830 (85.0) | <0.0001 |
| ACE-I or ARB use at baseline, n (%) | 2204 (81.0) | 2315 (77.1) | 3921 (80.0) | 2757 (82.8) | <0.0001 |
| Risk Factors for Algorithm | |||||
| Qualifying MI ≥2 years, n (%) | 1038 (38.2) | 1978 (65.9) | 1446 (29.5) | 901 (27.1) | <0.0001 |
| Time from last dose of P2Y12 inhibitor, days, median (IQR) | 63.0 (1.0, 348.0) | 425.0 (102.0, 618.5) | 64.0 (1.0, 290.0) | 39.0 (1.0, 223.0) | <0.0001 |
| Multivessel disease, n (%) | 1582 (58.2) | 736 (24.5) | 3191 (65.1) | 2802 (84.2) | <0.0001 |
| History of diabetes, n (%) | 1118 (41.1) | 309 (10.3) | 1183 (24.1) | 1904 (57.2) | <0.0001 |
| Peripheral artery disease, n (%) | 209 (7.7) | 19 (0.6) | 76 (1.6) | 461 (13.9) | <0.0001 |
| eGFR <60 mL/min (MDRD), n (%) | 946 (35.0) | 152 (5.1) | 662 (13.5) | 1435 (43.1) | <0.0001 |
| Multiple prior MIs, n (%) | 469 (17.2) | 105 (3.5) | 479 (9.8) | 1280 (38.5) | <0.0001 |
ACE-I, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; CABG, coronary artery bypass graft; CHF, congestive heart failure; eGFR, estimated glomerular filtration rate; IQR, interquartile range; MDRD, Modification of Diet in Renal Disease; MI, myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction; PCI, percutaneous coronary intervention; STEMI, ST-elevation myocardial infarction; TIA, transient ischaemic attack.
In the placebo arm, patients at high risk of bleeding had rates of major bleeding at 3 years that were almost double those in the patients categorized as low bleeding risk (1.7 vs. 0.9%, P = 0.040, Supplementary material online, FigureS1). With regard to ischaemic risk, in the placebo arm, the ischaemic risk characteristics were associated with higher rates of the primary endpoint at 3 years (see Supplementary material online, Figure S2). Overall application of the algorithm categorized 19% of the population as high bleeding risk and 81% as low bleeding risk, with those categorized as 22% with 0–1 ischaemic risk factors, 35% with 2 ischaemic risk factors, and 24% with ≥ 3 ischaemic risk factors (Graphical abstract).
Safety with ticagrelor 60 mg vs. placebo by risk group
Whereas the relative increase in bleeding risk with ticagrelor 60 mg twice daily vs. placebo was consistent across the risk groups, due to their higher baseline bleeding risk, there was an approximately two-fold higher ARD in TIMI major bleeding in high-bleeding risk patients vs. those at low bleeding risk (regardless of ischaemic risk factors): +2.3% [95% confidence interval (CI) 0.60–3.9] vs. +1.0% (95% CI 0.5, 1.5) (P = 0.25). The ARDs in bleeding across ischaemic risk groups among patients with low bleeding risk was similar (P-trend for ARD 0.96).
Drug discontinuation rates at 3 years were higher in those at high bleeding risk vs. those at low bleeding risk both in the placebo arm (26.5% in high bleeding risk, 22.7% in low bleeding risk and 0–1 ischaemic risk factors, 20.8% in low bleeding risk and 2 ischaemic risk factors, and 25.9% in low bleeding risk and ≥ 3 ischaemic risk factors) and the ticagrelor arm (33.8, 30.3, 28.9, and 33.7%, respectively).
Efficacy with ticagrelor 60 mg vs. placebo by risk group
The relative reduction in the risk of MACE with ticagrelor was 2% (HR 0.98, 95% CI 0.77–1.26, P = 0.88), 13% (HR 0.87, 95% CI 0.63–1.22, P = 0.42), 19% (HR 0.81, 95% CI 0.64–1.01, P = 0.062), and 23% (HR 0.77, 95% CI 0.62–0.95, P = 0.013; Figure 1A) in patients at high bleeding risk and at low bleeding risk with 0–1, 2, and ≥3 risk factors, respectively. Furthermore, the ARDs in the primary endpoint of MACE at 3 years in those groups were +0.08% (95% CI −2.39, 2.54), −0.47% (95% CI −2.21, 1.26), −1.53% (95% CI −3.1, 0.024), and −2.61% (95% CI −4.99, −0.24) with a P-value for the trend in ARD of 0.076 across the three groups. An alternative approach estimating the hypothetical ARDs for the primary endpoint in the groups by applying the overall HR to the placebo event rate in each group is shown in Supplementary material online, Table S3. In patients at high bleeding risk, 2154 had 2 or more ischaemic risk factors and 549 had 0–1 ischaemic risk factors with a non-significant trend towards an ∼10% reduction in risk in those with multiple risk factors (HR 0.90, 95% CI 0.69–1.18) and no apparent benefit in those with 0–1 risk factors (HR 1.81, 95% CI 0.91–3.61; P-interaction 0.0714).
Outcomes at 3 years with ticagrelor 60 mg twice daily vs. placebo by difference in hazard (left) and absolute risk difference (right) for (A) primary outcome composite of cardiovascular death, myocardial infarction, or stroke, (B) cardiovascular death, (C) all-cause mortality, and (D) net composite outcome of cardiovascular death, myocardial infarction, stroke, intracranial haemorrhage, or fatal bleeding.
Likewise, the effect of ticagrelor on CV death varied across these four patient subgroups, with no benefit (HR 1.24, 95% CI 0.87–1.76, ARD +1.0%, 95% CI −0.8, 2.8) in patients at high bleeding risk, a 19% relative risk reduction (RRR; HR 0.81, 95% CI 0.46–1.42) and −0.2 ARD (95% CI −1.3, +0.9, P = 0.46) in patients at low bleeding risk and 0–1 ischaemic risk factor, a 32% RRR (HR 0.68, 95% CI 0.45–1.03, P = 0.070) and −0.86% ARD (95% CI −1.86, 0.13) in patients at low bleeding risk and 2 ischaemic risk factors, and a 36% RRR (HR 0.64, 95% CI 0.44–0.93, P = 0.0020) and −1.62% ARD (95% CI −3.12, −0.14, P-trend for HR 0.0083, P-trend for ARR 0.018, Figure 1B) in those with ≥ 3 ischaemic risk factors. Similar patterns were seen for the effect of ticagrelor on all-cause mortality (high bleeding risk HR 1.14; 95% CI 0.86 to 1.50, ARD 1.46; 95% CI −0.85 to 3.78; low bleeding risk with 0–1 risk factor HR 0.90; 95% CI 0.59 to 1.38; ARD −0.25; 95% CI −1.62 to 1.13; low bleeding risk with 2 risk factors HR 0.83; 95% CI 0.60 to 1.15; ARD −0.53; 95% CI −1.69 to 0.63; low bleeding risk with 3 risk factors HR 0.71; 95% CI 0.52 to 0.96; ARD −2.01; 95% CI −3.82 to −0.20; P-trend for HR 0.021, P-trend for ARD 0.027, Figure 1C) as well as on coronary heart disease death, MI, and stroke (see Supplementary material online, Figure S3).
Net outcomes with ticagrelor 60 mg vs. placebo by risk group
In terms of the net clinical outcome of CV death, MI, stroke, intracranial bleeding, or fatal bleeding, there was no benefit in patients at high bleeding risk (HR 1.03, 95% CI 0.81–1.31; ARD 0.56, 95% CI −1.95, 3.06) and a gradient of benefit in those at low bleeding risk by whether they had 0–1 (HR 0.93, 95% CI 0.67–1.28; ARD −0.21, 95% CI −1.97, 1.56); 2 (HR 0.82, 95% CI 0.66–1.03; ARD −1.45, 95% CI −3.02, 0.12); or ≥3 ischaemic risk factors (HR 0.77, 95% CI 0.62–0.94; ARD −2.64, 95% CI −5.03, −0.25). There was a consistent pattern for the primary outcome and net outcomes with ticagrelor 90 mg vs. placebo in the three risk groups (see Supplementary material online, Table S4).
Discussion
The current analysis provides novel observations regarding the application of long-term DAPT in patients with prior MI. First, a stepwise approach based on the paradigm espoused by the most recent ESC guidelines for selection of patients at lower bleeding risk and higher thrombotic risk for additional antithrombotic therapy applied in a trial of long-term ticagrelor in patients with prior MI identifies a significant proportion of the population (∼60% of the PEGASUS-TIMI 54 population) that is at low bleeding risk and has two or more thrombotic risk factors. Second, use of such an algorithm identifies a population that derives greater absolute benefit, with less bleeding and resulting favourable findings for CV death and net outcomes.
Patients with prior MI have been described through a number of trials to be at long-term heightened risk of recurrent atherothrombosis. More potent antiplatelet therapy reduces this risk but increases bleeding requiring clinicians to personalize the approach, which acknowledges heterogeneity in the population both for ischaemic and bleeding risk. Established risk scores have shown the ability to discriminate ischaemic risk in broad atherosclerosis populations but have not evaluated bleeding risk with P2Y12 inhibition, limiting utilization for long-term ticagrelor. Scores developed to evaluate net outcomes such as the DAPT score13 have utility at the time of stenting in broad populations but do not necessarily translate to stable post-MI patients where procedural aspects are less relevant to risk relative to clinical characteristics and a significant proportion of patients have not had PCI. More recently, scores like PRECISE-DAPT11 have shown the ability to risk stratify in acute coronary syndrome (ACS) but are focused predominantly on bleeding rather than both bleeding and ischaemic risk and have been derived in ACS or PCI populations. Therefore, there is an unmet need for a patient selection algorithm for long-term ticagrelor in stable patients with a history of MI more than a year ago even if they have not been treated with PCI.
The most recent ESC guidelines for non-ST-elevation ACS describe an algorithm where patients at lower bleeding risk are first identified and then those at lower bleeding risk are categorized into two groups, higher thrombotic risk and lower thrombotic risk (called moderate risk in the guideline).1 The current analysis evaluates an approach conceptually aligned with this paradigm and incorporates the pre-specified risk criteria for the trial along with subgroup findings and the guideline characteristics to approximate this approach. Use of this approach not only identified a lower risk group for bleeding on ticagrelor but also identified a gradient of ischaemic risk within the low-bleeding risk group enabling further personalization. This approach translated to greater absolute benefits for MACE, heterogeneity with lower rates of CV death and greater net benefit in those at the highest ischaemic risk (at least 2 risk factors). In fact, the 59% of the population at low bleeding risk and with either 2 or ≥3 ischaemic risk factors may be the optimal population for the use of long-term ticagrelor 60 mg. Although those at high bleeding risk had ischaemic events, premature discontinuation of study treatment and potentially other background therapies may partially explain the attenuation of benefit. Future studies may help to further elucidate the exact reasons why those at high bleeding risk do not benefit from prolonged DAPT and explore how much any benefit is offset by ischaemic complications occurring in the context of bleeding events. Regardless, these observations support the guideline recommendation that long-term therapy should be considered (Class IIa recommendation) in such patients and provide extensive data for outcomes and net benefit that may be useful to clinicians in communicating risk/benefit to patients.
There are several limitations to this analysis. First, the data are derived from a single data set and have not been validated externally; however, there are few large data sets examining long-term ticagrelor in a randomized trial. Second, although not all factors outlined in the ESC guidelines were available for inclusion, the characteristics identified for both bleeding risk and ischaemic risk have been identified as independent predictors in other data sets and scores.1 Third, there was no effect modification for relative benefit for the primary endpoint by risk category; however, the goal of the analysis was to characterize ARDs that were strongly statistically significant. Fourth, the results are a post-hoc exploratory analysis and therefore should be considered hypothesis generating. Fifth, the ARDs were calculated by using the subgroup specific rates which was felt to be more conservative in highlighting that the high bleeding risk group may derive lesser benefit. An alternative approach to using the observed ARDs and instead applying the overall trial HR to the placebo event rate in each subgroup showed a more consistent absolute difference between groups. Sixth, this study assigned equal weights to ischaemic risk factors included in this study and readers cannot infer a definitive threshold for number of ischaemic risk factors for net benefit; future studies should derive a scoring system using predictive modelling methods. Finally, PEGASUS specifically excluded patients at highest risk of bleeding (history of previous intracranial bleed at any time, gastrointestinal (GI) bleed within the past 6 months, or major surgery within 30 days), and, like most randomized trials, enrolled a selected population that may not be entirely representative of post-ACS patients encountered in routine clinical care.22 Further validation in ‘real-world’ cohorts, including those with history of gastrointestinal bleeding, will lead to understanding of performance in a broader population of patients.
Conclusions
In conclusion, in stable patients with prior MI, a stepwise patient selection approach evaluating bleeding and ischaemic risk based on post-hoc subgroups may be useful in identifying patients who may derive greater net benefit with long-term ticagrelor. Specifically, those at lower bleeding risk with at least two ischaemic risk factors appeared to have the most favourable risk–benefit profile. This simple patient selection approach may be useful for clinicians identifying patients who may benefit from a strategy of long-term ticagrelor after MI.
Supplementary material
Supplementary material is available at European Heart Journal online.
Funding
This study was supported by a grant to Brigham and Women’s Hospital from AstraZeneca.
Data availability
Data for the analyses cannot be shared by those interested in collaboration are encouraged to contact the corresponding author.
References
Author notes
Conflict of interest: The TIMI Study Group has received significant research grant support from Abbott, Amgen, Aralez, AstraZeneca, Bayer HealthCare Pharmaceuticals, Inc., BRAHMS, Daiichi Sankyo, Eisai, GlaxoSmithKline, Intarcia, Janssen, MedImmune, Merck, Novartis, Pfizer, Poxel, Quark Pharmaceuticals, Roche, Takeda, The Medicines Company, Zora Biosciences. M.P.B. reports grant support to CPC clinical research from Amgen, AstraZeneca, Bayer, Janssen, Merck, Novo Nordisk, Pfizer, Sanofi, Wraser. M.D. reports personal fees for lectures and trial steering committee participation from AstraZeneca and Novartis, personal fees for lectures from Boehringer Ingelheim, Sanofi and Bayer. Advisory Board: Amgen, Novo Nordisk, AstraZeneca, Novartis. Personal fees for serving on the Data Monitoring Committee: for the DAPA-MI trial, funded by AstraZeneca. R.F.S. reports institutional research grants/support from AstraZeneca, Cytosorbents, GlyCardial Diagnostics, and Thromboserin; and personal fees from Alnylam, Amgen, AstraZeneca, Bayer, Bristol Myers Squibb/Pfizer, Chiesi, CSL Behring, Cytosorbents, GlyCardial Diagnostics, Hengrui, Idorsia, Intas Pharmaceuticals, Medscape, Novartis, PhaseBio, Sanofi Aventis, and Thromboserin. D.L.B discloses the following relationships—Advisory Board: Bayer, Boehringer Ingelheim, Cardax, CellProthera, Cereno Scientific, Elsevier Practice Update Cardiology, Janssen, Level Ex, Medscape Cardiology, Merck, MyoKardia, NirvaMed, Novo Nordisk, PhaseBio, PLx Pharma, Regado Biosciences, and Stasys; Board of Directors: Bristol Myers Squibb (stock), Boston VA Research Institute, DRS.LINQ (stock options), Society of Cardiovascular Patient Care, TobeSoft; Chair: Inaugural Chair, American Heart Association Quality Oversight Committee; Data Monitoring Committees: Acesion Pharma, Assistance Publique-Hôpitaux de Paris, Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute, for the PORTICO trial, funded by St Jude Medical, now Abbott), Boston Scientific (Chair, PEITHO trial), Cleveland Clinic (including for the ExCEED trial, funded by Edwards), Contego Medical (Chair, PERFORMANCE 2), Duke Clinical Research Institute, Mayo Clinic, Mount Sinai School of Medicine (for the ENVISAGE trial, funded by Daiichi Sankyo; for the ABILITY-DM trial, funded by Concept Medical), Novartis, Population Health Research Institute; Rutgers University (for the NIH-funded MINT Trial); Honoraria: American College of Cardiology (Senior Associate Editor, Clinical Trials and News, ACC.org; Chair, ACC Accreditation Oversight Committee), Arnold and Porter law firm (work related to Sanofi/Bristol Myers Squibb clopidogrel litigation), Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute; RE-DUAL PCI clinical trial steering committee funded by Boehringer Ingelheim; AEGIS-II executive committee funded by CSL Behring), Belvoir Publications (Editor in Chief, Harvard Heart Letter), Canadian Medical and Surgical Knowledge Translation Research Group (clinical trial steering committees), Cowen and Company, Duke Clinical Research Institute (clinical trial steering committees, including for the PRONOUNCE trial, funded by Ferring Pharmaceuticals), HMP Global (Editor in Chief, Journal of Invasive Cardiology), Journal of the American College of Cardiology (Guest Editor; Associate Editor), K2P (Co-Chair, interdisciplinary curriculum), Level Ex, Medtelligence/ReachMD (CME steering committees), MJH Life Sciences, Piper Sandler, Population Health Research Institute (for the COMPASS operations committee, publications committee, steering committee, and USA national co-leader, funded by Bayer), Slack Publications (Chief Medical Editor, Cardiology Today’s Intervention), Society of Cardiovascular Patient Care (Secretary/Treasurer), WebMD (CME steering committees), Wiley (steering committee); Other: Clinical Cardiology (Deputy Editor), NCDR-ACTION Registry Steering Committee (Chair), VA CART Research and Publications Committee (Chair); Research Funding: Abbott, Afimmune, Aker Biomarine, Amarin, Amgen, AstraZeneca, Bayer, Beren, Boehringer Ingelheim, Bristol Myers Squibb, Cardax, CellProthera, Cereno Scientific, Chiesi, CSL Behring, Eisai, Ethicon, Faraday Pharmaceuticals, Ferring Pharmaceuticals, Forest Laboratories, Fractyl, Garmin, HLS Therapeutics, Idorsia, Ironwood, Ischemix, Janssen, Javelin, Lexicon, Lilly, Medtronic, Merck, Moderna, MyoKardia, NirvaMed, Novartis, Novo Nordisk, Owkin, Pfizer, PhaseBio, PLx Pharma, Recardio, Regeneron, Reid Hoffman Foundation, Roche, Sanofi, Stasys, Synaptic, The Medicines Company, 89Bio; Royalties: Elsevier (Editor, Braunwald’s Heart Disease); Site Co-Investigator: Abbott, Biotronik, Boston Scientific, CSI, St Jude Medical (now Abbott), Philips, Svelte; Trustee: American College of Cardiology; Unfunded Research: FlowCo, Takeda. P.G.S. reports research grants from Amarin, Bayer, Merck, Sanofi, and Servier, speaking or consulting fees from Amarin, Amgen, AstraZeneca, Bayer/Janssen, Boehringer Ingelheim, Bristol Myers Squibb, Idorsia, Lilly, Merck, Novartis, Novo Nordisk, Pfizer, Regeneron, Sanofi, and Servier. M.C. reports grants and personal fees from AstraZeneca, during the conduct of the study; personal fees from Merck, personal fees from Janssen, personal fees from Maquet, personal fees from malpractice attorneys, grants from Janssen, grants from Edwards, personal fees from Merck, personal fees from BMS/Pfizer, personal fees from Janssen, personal fees from BI, personal fees from Lilly, outside the submitted work. K.I. is the member of the TIMI Study Group which has received institutional research grant support through Brigham and Women’s Hospital from Abbott, Amgen, Aralez, AstraZeneca, Bayer HealthCare Pharmaceuticals, Inc., BRAHMS, Daiichi Sankyo, Eisai, GlaxoSmithKline, Intarcia, Janssen, MedImmune, Merck, Novartis, Pfizer, Poxel, Quark Pharmaceuticals, Roche, Takeda, The Medicines Company, and Zora Biosciences. P.J. is an employee of AstraZeneca. E.B. reports grant support through Brigham and Women’s Hospital from AstraZeneca, Daiichi Sankyo, Merck, and Novartis, Consultancies with Amgen, Cardurion, MyoKardia, Novo Nordisk, and Verve. M.S.S. received research grant support through Brigham and Women’s Hospital from Abbott, Amgen, Anthos Therapeutics, AstraZeneca, Bayer, Daiichi Sankyo, Eisai, Intarcia, Ionis, Medicines Company, MedImmune, Merck, Novartis, Pfizer, and Quark Pharmaceuticals. Consulting for Althera, Amgen, Anthos Therapeutics, AstraZeneca, Beren Therapeutics, Bristol Myers Squibb, DalCor, Dr Reddy’s Laboratories, Fibrogen, Intarcia, Merck, Moderna, Novo Nordisk, and Silence Therapeutics. Additionally, Dr Sabatine is a member of the TIMI Study Group, which has also received institutional research grant support through Brigham and Women’s Hospital from: ARCA Biopharma, Inc., Janssen Research and Development, LLC, Siemens Healthcare Diagnostics, Inc., Softcell Medical Limited, Regeneron, Roche, and Zora Biosciences.
