Original Contribution

Evaluation of a Real-World In-Hospital Antiplatelet-Switching Strategy Following Coronary Interventions: The SWITCH Study

Antoine E. Soueid, MD, PharmD, JD1;  Ibrahim Kassas, MD2;  Jeffrey Rade, MD1; Nikolaos Kakouros, MBBS, MD(Res), PhD1

Antoine E. Soueid, MD, PharmD, JD1;  Ibrahim Kassas, MD2;  Jeffrey Rade, MD1; Nikolaos Kakouros, MBBS, MD(Res), PhD1

Abstract

Background. Antiplatelet therapy is paramount to reduce the risk of coronary stent thrombosis after percutaneous coronary intervention (PCI). Newer agents are reliable and have a fast onset of action, but have significantly higher cost, leading to compliance concerns. We adopted and evaluated an acute agent-switching strategy, using prasugrel or ticagrelor for rapid and reliable periprocedural antiplatelet action, followed by a switch to generic clopidogrel. Methods. This large, single-center study included all patients who underwent PCI between January 1, 2013 and December 31, 2016. Study endpoints were 30-day mortality and bleeding events. Results. A total of 5007 patients met inclusion criteria. Average age was 63.5 ± 12.5 years. Prior to PCI, 54.8% of patients were preloaded with ticagrelor, 8.5% with prasugrel, and 36.7% with clopidogreI. The majority of patients (93%) loaded with ticagrelor and more than half (58%) of those loaded with prasugrel were subsequently switched prior to hospital discharge to clopidogrel for long-term therapy. Patients pretreated with ticagrelor or prasugrel and switched to clopidogrel had overall lowest bleeding rates (0.9% and 0.8%, respectively). The highest rates of bleeding were noted in patients maintained on ticagrelor or clopidogrel throughout (2.5% and 1.7%, respectively). After accounting for additional periprocedural use of intravenous glycoprotein IIb/IIIa inhibitors, the lowest bleeding rates were observed in patients loaded with ticagrelor and switched to clopidogrel (0.75%), with the highest bleeding observed in patients maintained on ticagrelor throughout. There were no events of acute stent thrombosis. Conclusions. A strategy of using newer, fast-acting, and reliable antiplatelet agents prior to PCI and acutely switching to long-term clopidogrel therapy appears safe and efficacious. Although the superiority of the newer antiplatelet agents for long-term post-PCI dual-antiplatelet therapy in a trial setting is well established, the impact of increased adherence to lower-cost clopidogrel therapy in the real-world setting merits further consideration.

J INVASIVE CARDIOL 2021;33(4):E263-E268.

Key words: DAPT, dual-antiplatelet therapy, platelet adenosine diphosphate P2Y12 receptor inhibitor


Percutaneous coronary intervention (PCI) with stent implantation is a common treatment for patients with symptomatic coronary artery disease despite optimal medical therapy, as well as patients presenting with acute coronary syndromes (ACS). During PCI, antiplatelet and anticoagulant medications are required to prevent equipment thrombus formation as well as acute coronary arterial closure. Subsequent dual-antiplatelet therapy (DAPT) with aspirin and a platelet adenosine diphosphate P2Y12 receptor (P2Y12) inhibitor for 6 months following PCI for stable coronary artery disease and for 12 months after an ACS presentation is currently recommended to reduce the risk of thrombotic events. 

There are several oral P2Y12 inhibitors available on the market. The oldest of these therapies that remains in common clinical use is clopidogrel bisulfate, available since the late 1990s and in generic form in the United States (US) since 2012. Two additional P2Y12 inhibitor agents, prasugrel and ticagrelor, received approval by the US Food and Drug Administration (FDA) in 2009 and 2011, respectively. Both newer agents have a more rapid onset of action and a more pronounced and reliable platelet inhibition effect when compared with clopidogrel, making them a more dependable option to achieve periprocedural platelet inhibition. Subsequent DAPT using prasugrel and ticagrelor has also been shown to reduce rates of myocardial infarction (MI), cardiovascular death, and stroke compared with clopidogrel, leading to a class IIa recommendation by the American College of Cardiology/American Heart Association guidelines and a class I recommendation by the European Society of Cardiology for their use over clopidogrel in patients who are not at high risk of bleeding.1,2

Although the newer antiplatelet agents carry guideline recommendations in favor over the older clopidogrel, they have a higher associated cost and often higher patient copayments, leading to compliance concerns due to non-affordability. Prior studies suggest a more than 3-fold increase in P2Y12 inhibitor non-compliance after PCI with increased use of the more expensive prasugrel and ticagrelor.3 Early discontinuation of P2Y12 inhibitor therapy following PCI is associated with increased major adverse cardiac event (MACE) rates.4 In view of these concerns, clopidogrel remains the most commonly prescribed P2Y12 inhibitor in the United States following MI.5

Limited expert consensus opinion exists regarding de-escalation from long-term ticagrelor or prasugrel to clopidogrel therapy, but there is a lack of data on acute antiplatelet agent switching in the real-world clinical setting.6 The purpose of this retrospective study was to evaluate the safety and efficacy of using ticagrelor or prasugrel for rapid periprocedural antiplatelet effect and switching antiplatelet agents to clopidogrel acutely following the PCI procedure and prior to hospital discharge. 

Methods

This is a large, retrospective, single-center study of patients undergoing PCI. All patients who underwent a PCI between January 1, 2013 and December 31, 2016 at the University of Massachusetts Memorial Medical Center (UMass) were included in the study. Data were abstracted from the UMass electronic medical record (EMR) systems, and included demographics, diagnoses, comorbid conditions, past medical history, list of medications during PCI, procedural technique, physical examination, pharmacy and emergency room medication history, and laboratory results. Mortality at 30 days was also abstracted from EMR.

An institutional policy on antiplatelet therapy was agreed among local providers prior to the study. Patients taking clopidogrel at the time of the index presentation were continued on this drug. Patients presenting with acute ST-segment elevation MI (STEMI) were loaded with 180 mg ticagrelor at the time of hospital presentation. Patients presenting with non-ST segment elevation MI (NSTEMI) not previously loaded with an antiplatelet agent were loaded with 60 mg of prasugrel after the coronary anatomy was defined by coronary angiography. Patients presenting with unstable angina were loaded with antiplatelet medications prior to coronary angiography at the treating physician’s discretion.

Switching antiplatelet therapy was performed on an individual case basis, according to the agreed-upon algorithm, including consideration of the drug’s affordability to the patient evaluated by establishing the expected out-of-pocket copay and presenting the information regarding relative drug costs to the patient, as well as risk/benefit evaluation depending on the coronary anatomy and PCI performed. Risk evaluation included patient factors (such as recent MI, diabetes mellitus, or smoking), procedural factors (such as length and size of stents, bifurcation stenting), and potential impact of stent thrombosis (stent subtending large myocardial territory, multivessel PCI). The final decision on whether to switch antiplatelet agent was made following discussions between the primary cardiology team, implanting interventional cardiologist, and patient.

In patients where a switching strategy was employed, an antiplatelet loading dose of the new agent was administered. A clopidogrel loading dose of 600 mg was used in patients <75 years old, and a 300 mg dose was used in patients ≤75 years old due to increased risk of bleeding.7,8 The medication switch load was administered at the time the next dose of the established drug was due. All switching occurred prior to patient discharge.

Patients were divided into 5 subgroups based on choice of loading and maintenance antiplatelet agent: ticagrelor to ticagrelor; ticagrelor to clopidogrel; prasugrel to prasugrel; prasugrel to clopidogrel; and clopidogrel to clopidogrel. The incidence of significant bleeding during hospitalization was assessed across all subgroups, accounting for those patients who received glycoprotein (GP) IIb/IIIa inhibitors during or after PCI. Endpoints were 30-day all-cause mortality and bleeding events assessed using definitions set forth by the Blood Academic Research Consortium (BARC).9 Continuous variables are reported as the mean ± standard deviation and compared using Student’s t-test. Categorical variables are reported as absolute values with percentages and compared with Chi-squared analysis.  

Results

A total of 5007 patients met the inclusion criteria; a small number of patients who were prescribed ticlopidine (n = 18) were excluded from the analysis. Patient demographic information are presented in Table 1. The average patient age was 63.5 ± 12.5 years and the majority (73%) of patients were men. Most patients (89%) were white and had high prevalence of traditional risk factors for coronary artery disease. Most patients (94%) presented with an ACS (STEMI, NSTEMI, unstable angina), with only 5% of patients presenting with stable angina or other indications (notably including preoperative evaluation for non-cardiac surgery). 

Procedural characteristics are presented in Table 2. The cardiac catheterization procedures were performed predominantly by radial access (87%) with adjunctive use of unfractionated heparin (95%) and low utilization of other anticoagulants, such as bivalirudin or low-molecular weight heparin. A total of 719 patients received a GP IIb/IIIa inhibitor during the PCI, at the operator’s discretion.

Most patients (63.3%) were preloaded with a newer agent (54.8% ticagrelor and 8.5% prasugrel), while the remaining patients were on clopidogrel at the time of intervention. The majority of patients loaded with ticagrelor (93%) and most of the patients loaded with prasugrel (58%) were subsequently switched to clopidogrel for long-term therapy.

Inpatient and 30-day outcomes are presented in Table 3. A total of 63 patients had major bleeding (BARC type 3a) that required blood transfusion following PCI during the index hospitalization; 1 additional patient had a major bleeding episode, but was not transfused. There were 9 minor events (BARC type 2), including 7 access-site hematomas, 1 patient with minor access-site bleeding, and 1 patient with mild genitourinary bleeding following PCI. Twenty-two of these patients were on periprocedural GP IIb/IIIa intravenous inhibitor. There were no BARC acute or subacute definite stent thrombosis events.10

There were significant differences in major bleeding, depending on the antiplatelet management strategy (P<.001). Patients pretreated with ticagrelor or prasugrel and switched to clopidogrel had overall lowest bleeding rates (0.9% and 0.8%, respectively). The highest rates of major BARC 3a bleeding were noted in patients maintained on ticagrelor or clopidogrel throughout (2.5% and 1.7%, respectively). There were significant differences in the use of adjunctive intravenous GP IIb/IIIa inhibitors during PCI (P<.001) with patients established on clopidogrel the least likely to be given additional GP IIb/IIIa therapy (11.3%), compared with patients loaded with ticagrelor (16.7%) or prasugrel (12.7%). After accounting for additional periprocedural use of intravenous GP IIb/IIIa inhibitors, the lowest BARC 3a bleeding rates were observed in patients loaded with ticagrelor and switched to clopidogrel (0.75%), with the highest bleeding in patients maintained on ticagrelor throughout. 

Thirty-day all-cause mortality rate for the entire population was 1%, with the highest rates in the patients maintained on clopidogrel throughout the index hospitalization (1.9%), followed by patients loaded with ticagrelor and then switched to clopidogrel for maintenance (0.6%). Overall 30-day all-cause mortality was significantly lower in patients where the switching strategy was employed compared with patients on clopidogrel therapy throughout the hospitalization (P<.001).

Discussion

This is a large, retrospective, single-center study addressing a real-life strategy of using ticagrelor or prasugrel for fast and reliable periprocedural antiplatelet inhibition followed by in-hospital switching of antiplatelet therapy to clopidogrel for long-term DAPT in order to improve prescription affordability following discharge. 

Despite guideline recommendations favoring the use of prasugrel or ticagrelor over clopidogrel, the latter is often more affordable to the patients and remains widely used in the US for long-term DAPT. Nonetheless, clopidogrel provides less-rapid and efficient platelet inhibition than the newer agents, even when a higher loading dose is used.11,12 Furthermore, clopidogrel is predominantly hydrolyzed (up to 85%) by esterases into an inactive derivative and by hepatic cytochrome P450 (CYP) isoenzymes, mainly CYP2C19, to the active drug. Consequently, patients with poorly active CYP2C19 polymorphic alleles have diminished antiplatelet inhibition effect with clopidogrel, even when higher drug doses are used. Such “clopidogrel resistance” is associated with worse MACE outcomes, especially stent thrombosis.13

Prasugrel, once converted to its active metabolite by intestinal hydrolysis and oxidation by CYP450 2B6 and 3A4 isoenzymes, causes rapid and reliable irreversible inhibition of platelet function for the lifetime of the affected platelet. Prasugrel, however, is not recommended for patients ≥75 years old, in patients with prior transient ischemia attack or stroke, or in patients weighing <60 kg due to increased risk of bleeding, as observed in the TRITON-TIMI 38 study.14 

In contrast to the thienopyridines, ticagrelor, a cyclopentyltriazol-pyrimidine (CPTP) class molecule, has an active metabolite generated through CYP3A4/5 enzymes but does not require activation to exert its effect. Ticagrelor (and its active metabolite) reversibly bind an allosteric site and alter the conformation of the P2Y12 receptors on platelets achieving platelet inhibition in a drug-concentration dependent manner. The faster reversibility of ticagrelor renders it preferable for patients loaded with antiplatelet medication prior to coronary angiography who may then require coronary artery bypass graft surgery. In view of the more rapid and reliable onset of effect, both ticagrelor and prasugrel are preferable to clopidogrel in achieving platelet inhibition prior to a PCI procedure.

Following favorable randomized clinical trial results, there was rapid use of the newer DAPT alternatives to clopidogrel in some countries. A Swedish nationwide observational study of drug prescriptions for an unselected cohort of ACS patients undergoing PCI, for example, demonstrated a rapid increase in prescriptions of ticagrelor and decline in use of clopidogrel between 2009 and 2013.15 In contrast, a similar analysis of 167,455 registry patients in the US between October 2013 and December 2014 found clopidogrel to remain the most commonly used oral platelet inhibitor prescribed after MI.5 This differential practice pattern may be partly ascribed to the significant geographic region interaction noted in a subanalysis of the PLATO trial, which reported reduced efficacy of ticagrelor vs clopidogrel in North American patients; a subsequent systematic analysis suggested differences in aspirin dosing as the likely reason for the interaction.16

As expected, clinical factors, such as young age, invasive management, and in-hospital reinfarction while on clopidogrel were associated with patients being discharged with a ticagrelor prescription. Conversely, clopidogrel was favored over ticagrelor in patients on home anticoagulation and in patients with atrial fibrillation or with prior/in-hospital cerebrovascular accident, as well as in those receiving transfusions during hospitalization, presumably in an effort to reduce bleeding complications. Nonetheless, white race and private insurance were also independently associated with a higher likelihood of ticagrelor prescription at discharge over clopidogrel (P<.001), suggesting that socioeconomic factors and real (or presumed) drug affordability to the patient are affecting medication prescription decisions. 

Prior registry data suggest that switching antiplatelet therapy within the first 12 months after PCI — most commonly, de-escalation from prasugrel/ticagrelor to clopidogrel — occurs in as many as 20% of patients and is uncommonly associated with MACE or bleeding events.17,18 In the TRANSLATE-ACS study, switching from prasugrel or ticagrelor to clopidogrel prior to hospital discharge was not associated with increased MACE when compared with continuation of the higher-potency agent.19 More recently, the TOPIC randomized trial, which examined de-escalation from prasugrel/ticagrelor to clopidogrel in patients who had been event free after 1 month following ACS, showed reduced bleeding complications and no increase in ischemic events.20 Similarly, the TROPICAL-ACS trial randomized patients to de-escalation from prasugrel to clopidogrel just 1 week after PCI and found the switching strategy to be non-inferior in terms of net clinical outcomes compared with prasugrel for 12 months, although notably, 40% of patients were escalated to prasugrel due to high residual platelet reactivity on clopidogrel.21 There remains a lack of data regarding the overall safety of a routine acute in-hospital de-escalation strategy.

The impact of affordability of a P2Y12 inhibitor on outcomes following ACS was recently evaluated in the ARTEMIS (Affordability and Real-World Antiplatelet Treatment Effectiveness After Myocardial Infarction Study) trial.22 In this prospective, multicenter trial, a total of 11,001 patients were enrolled and underwent 1:1 cluster randomization to usual care or copayment intervention, consisting of a voucher to cover their copay costs for clopidogrel or ticagrelor. Patients on the copay intervention arm were significantly more likely to be prescribed ticagrelor. The copay intervention patients were also significantly more likely to self-report P2Y12 compliance compared with usual care (87.0% vs 83.8%, respectively; P<.01). There was no overall difference in MACE (defined as composite of MI, stroke, or death) between the groups. Notably, however, 28% of patients who received vouchers did not use them at all; these patients were more likely to be of non-white race, with cardiovascular comorbidities and lower levels of education. The subgroup of patients in the intervention arm who used their copay vouchers did have a lower MACE rate (10.0% vs 16.2%; P<.001).

The findings of the ARTEMIS trial are in support of the hypothesis that lower-cost medications may improve patient medication adherence. A copay voucher strategy, however, as explored in ARTEMIS, proved overall unable to improve patient outcomes, which was likely due to the cumbersome nature of this intervention, as patients may not remember or choose not to use medication vouchers.23 The use of prasugrel or ticagrelor in the acute setting to achieve rapid platelet inhibition, followed by a lower-cost medication for long-term therapy where appropriate, as proposed in the present study, may have a significant impact on patient compliance and real-world outcomes, outside a drug trial setting where ticagrelor and prasugrel have more favorable outcomes.

Study limitations. Our study has several limitations. First, it is a single-center, retrospective study with limited 30-day outcome data. Nonetheless, the impact of the acute use of ticagrelor or prasugrel to achieve rapid periprocedural antiplatelet effects would not be expected to affect outcomes beyond this period. The relative balance between increased medication compliance for clopidogrel and the superior outcomes of using ticagrelor or prasugrel in a trial setting remains unresolved. Second, the rationale for selection and switching from the various P2Y12 inhibitors was at the discretion of treating physicians and not recorded. Third, data were retrospectively abstracted from the charts and hence, dependent on accurate abstraction. 

Conclusion

The present study is a large, single-center study that evaluated a real-life practical strategy of using a newer, faster acting, reliable, and potent P2Y12 inhibitor (predominantly ticagrelor) for rapid periprocedural platelet inhibition for PCI followed by switching therapy to clopidogrel prior to discharge to enhance long-term therapy affordability in an effort to improve compliance. The approach appears safe from a bleeding, reinfarction, stent thrombosis, and 30-day mortality perspective. The overall balance between the proven pharmacological superiority of the newer antiplatelet agents in a trial setting and the impact of increased adherence to lower-cost clopidogrel therapy in the real-world setting merits further consideration.

References

1. Roffi M, Patrono C, Collet JP, et al. 2015 ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation: task force for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation of the European Society of Cardiology (ESC). Eur Heart J. 2016;37:267-315.

2. Levine GN, Bates ER, Bittl JA, et al. 2016 ACC/AHA guideline focused update on duration of dual antiplatelet therapy in patients with coronary artery disease: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines. J Am Coll Cardiol. 2016;68:1082-1115.

3. Dayoub EJ, Seigerman M, Tuteja S, et al. Trends in platelet adenosine diphosphate P2Y12 receptor inhibitor use and adherence among antiplatelet-naive patients after percutaneous coronary intervention, 2008-2016. JAMA Intern Med. 2018;178:943.

4. Mehran R, Baber U, Steg PG, et al. Cessation of dual antiplatelet treatment and cardiac events after percutaneous coronary intervention (PARIS): 2 year results from a prospective observational study. Lancet. 2013;382:1714-1722.

5. Basra SS, Wang TY, Simon DN, et al. Ticagrelor use in acute myocardial infarction: insights from the National Cardiovascular Data Registry. J Am Heart Assoc. 2018;7:e008125.

6. Angiolillo DJ, Rollini F, Storey RF, et al. International expert consensus on switching platelet P2Y12 receptor-inhibiting therapies. Circulation. 2017;136:1955-1975.

7. Cuisset T, Quilici J, Grosdidier C, et al. Comparison of platelet reactivity and clopidogrel response in patients ≤ 75 years versus > 75 years undergoing percutaneous coronary intervention for non-ST-segment elevation acute coronary syndrome. Am J Cardiol. 2011;108:1411-1416. 

8. de Matos Soeiro A, Casale G, Albanez Albuquerque de Medeiros Lopes MA, et al. Is there safety in the use of clopidogrel loading dose in patients over 75 years of age with acute coronary syndrome? Int J Cardiovasc Sci. 2019;32:449-456. 

9. Mehran R, Rao SV, Bhatt DL, et al. Standardized bleeding definitions for cardiovascular clinical trials: a consensus report from the Bleeding Academic Research Consortium. Circulation. 2011;123:2736-2747.

10. Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in coronary stent trials. Circulation. 2007;115:2344-2351.

11. Wiviott SD, Trenk D, Frelinger AL, et al. Prasugrel compared with high loading- and maintenance-dose clopidogrel in patients with planned percutaneous coronary intervention: the prasugrel in comparison to clopidogrel for inhibition of platelet activation and aggregation thrombolysis in myocardial infarction 44 trial. 2007;116:2923-2932.

12. Gurbel PA, Bliden KP, Butler K, et al. Randomized double-blind assessment of the ONSET and OFFSET of the antiplatelet effects of ticagrelor versus clopidogrel in patients with stable coronary artery disease. Circulation. 2009;120:2577-2585.

13. Mega JL, Simon T, Collet J-P, et al. Reduced-function CYP2C19 genotype and risk of adverse clinical outcomes among patients treated with clopidogrel predominantly for PCI. JAMA. 2010;304:1821.

14. Antman EM, Wiviott SD, Murphy SA, et al. Early and late benefits of prasugrel in patients with acute coronary syndromes undergoing percutaneous coronary intervention. J Am Coll Cardiol. 2008;51:2028-2033.

15. Angerås O, Hasvold P, Thuresson M, Deleskog A, Öbraun O. Treatment pattern of contemporary dual antiplatelet therapies after acute coronary syndrome: a Swedish nationwide population-based cohort study. Scand Cardiovasc J. 2016;50:99-107.

16. Mahaffey KW, Wojdyla DM, Carroll K, et al. Ticagrelor compared with clopidogrel by geographic region in the platelet Inhibition and patient outcomes (PLATO) trial. Circulation. 2011;124:544-554.

17. Zettler ME, Peterson ED, McCoy LA, et al. Switching of adenosine diphosphate receptor inhibitor after hospital discharge among myocardial infarction patients: insights from the treatment with adenosine diphosphate receptor inhibitors: longitudinal assessment of treatment patterns and events after acute coronary syndrome (TRANSLATE-ACS) observational study. Am Heart J. 2017;183:62-68.

18. Schiele F, Puymirat E, Bonello L, et al. Switching between thienopyridines in patients with acute myocardial infarction and quality of care. 2016;3:e000384.

19. Bagai A, Peterson ED, Honeycutt E, et al. In-hospital switching between adenosine diphosphate receptor inhibitors in patients with acute myocardial infarction treated with percutaneous coronary intervention: Insights into contemporary practice from the TRANSLATE-ACS study. Eur Heart J Acute Cardiovasc Care. 2015;4:499-508.

20. Cuisset T, Deharo P, Quilici J, et al. Benefit of switching dual antiplatelet therapy after acute coronary syndrome: the TOPIC (timing of platelet inhibition after acute coronary syndrome) randomized study. Eur Heart J. 2017;38:3070-3078.

21. Sibbing D, Aradi D, Jacobshagen C, et al. Guided de-escalation of antiplatelet treatment in patients with acute coronary syndrome undergoing percutaneous coronary intervention (TROPICAL-ACS): a randomised, open-label, multicentre trial. Lancet. 2017;390:1747-1757.

22. Wang TY, Kaltenbach LA, Cannon CP, et al. Effect of medication co-payment vouchers on P2Y12 inhibitor use and major adverse cardiovascular events among patients with myocardial infarction. JAMA. 2019;321:44.

23. Mathews R, Peterson ED, Honeycutt E, et al. Early medication nonadherence after acute myocardial infarction: insights into actionable opportunities from the treatment with ADP receptor inhibitors: longitudinal assessment of treatment patterns and events after acute coronary syndrome (TRANSLATE-ACS) study. Circ Cardiovasc Qual Outcomes. 2015;8:347-356.


From the 1University of Massachusetts Medical School, Worcester, Massachusetts; and 2Advocate Christ Medical Center, University of Illinois, Chicago, Illinois.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript accepted July 17, 2020.

Address for correspondence: Nikolaos Kakouros, MBBS, MD(Res), PhD, UMass Memorial Medical Center, Cardiovascular Medicine, 55 Lake Avenue North, Worcester, MA 01655. Email: nikolaos.kakouros@umassmemorial.org

/sites/invasivecardiology.com/files/articles/images/E263-E268%20Soueid%20JIC%202021%20April%20wm.pdf