Recommendations and Guidelines
Periprocedural antiplatelet therapy: recommendations for standardized reporting in patients on antiplatelet therapy: communication from the SSC of the ISTH
- Manuscript handled by: S. Eichinger
- Final decision: F. R. Rosendaal, 28 April 2013
Correspondence: Alex C. Spyropoulos, Anticoagulation Services and Clinical Thrombosis, North Shore-LIJ Health System at Lenox Hill Hospital, Associate Professor of Medicine, Hofstra North Shore-LIJ School of Medicine, 100 E 77th St, New York, NY 10065, USA.
Tel.: +1 212 434 6776; fax: +1 212 434 6781.
The periprocedural management of patients receiving antiplatelet (AP) therapy such as aspirin (ASA) or P2Y12 receptor antagonists (e.g., clopidogrel, ticagrelor, and prasugrel) for the primary or secondary prevention of cardiovascular disease is an increasingly common clinical problem . Approximately 10–15% of normal platelet function is restored for each day of interruption of AP therapy, and therefore it would take 7–10 days for an entire platelet pool to be restored . Cohort studies involving patients receiving combined anticoagulant and ASA therapy typically stopped ASA 7–10 days before the procedure or surgery [[3-7]. When resuming AP therapy, the maximum antiplatelet effect of ASA can occur within minutes, and with agents like clopidogrel, when administered with a 150- to 600-mg loading dose, maximum platelet inhibition can occur within 12–15 h after administration [8, 9]. Cohort studies involving patients receiving anticoagulant therapy and ASA typically resumed receiving ASA within 24 h for most surgical procedures [6, 7].
Meta-analyses of periprocedural AP therapy assessed > 15 000 patients who were receiving ASA before elective noncardiac surgery or had perioperative continuation of ASA in patients undergoing coronary artery bypass graft surgery [10, 11]. Studies have generally been of moderate or low quality of mostly observational cohort designs or used indirect data from databases involving patients in non-periprocedural settings [12, 13]. There were no uniform definitions of patient- and procedure-related outcomes such as cardiovascular and bleeding risk—including major bleeding definitions—and heterogeneity of time frames for capturing adverse events and duration of follow-up.
The objective of this ISTH communication is to recommend a standardized approach for reporting patient and procedural thromboembolic and bleed risk, AP regimens, key outcome definitions including arterial thromboembolism, major bleeding, and standardized reporting of the duration of follow-up when conducting future studies in periprocedural AP therapy, which include elective major and minor procedures and surgeries.
Patient thromboembolic risk stratification and periprocedural bleeding risk
Several perioperative risk stratification scores or indexes have been validated for the prediction of cardiac risk in major non-cardiac surgery, with the most widely used being the Revised Cardiac Risk Index (RVSI) and the Perioperative Cardiac Risk Calculator [14, 15]. The RVSI has the advantage of simplicity and validation across a wide range of patient populations, with event rates ranging from 0.4% to 0.6% in classes I and II, 6.6% in class III, and 11.0% in class IV . The RVSI includes six variables that are easily obtainable in the perioperative setting: (i) high-risk surgery, (ii) ischemic heart disease, (iii) history of congestive heart failure, (iv) history of cerebrovascular disease, (v) diabetes (with insulin therapy), (vi) and preoperative serum creatinine > 2.0 mg dL−1. The RVSI and other risk scores neither are specific to the periprocedural use of AP therapy nor incorporate the full range of thromboembolic events. Biomarkers such as the use of high-sensitive troponin T that may provide incremental value to the RVSI in perioperative risk stratification are being evaluated .
The thromboembolic risk depends on whether patients are receiving single or dual AP therapy for the primary or secondary prevention of cardiovascular events, as well as the timing of the index event, such as a recent myocardial infarction or recent placement of a coronary stent . We propose a thromboembolic risk stratification scheme as shown in Table 1 to encompass a three-tiered risk scheme for patients receiving dual AP therapy in the periprocedural period, which includes patients requiring dual AP therapy for coronary as well as possibly stroke prevention indications. Last, studies should include the type, site, and number of coronary stents in the overall patient assessment, as these would strongly affect the overall thromboembolic risk estimate.
Table 1. Proposed thromboembolism risk stratification in periprocedural dual antiplatelet studies
|DES within 1 year||DES within 1–2 years||DES > 2 years|
|BMS within 3 months||BMS within 3–12 months||BMS > 1 year|
|MI within 3 months||MI > 3 months||Primary aortic thrombolembolism prevention|
|History of stent thrombosis||CHADS2 2–4||CHADS2 0–1|
|CHADS2 5–6|| || |
There have been several attempts to standardize periprocedural bleeding risk, although none have been validated . We believe that it would be clinically useful to explicitly divide procedures and surgeries into a two-tier risk scheme of (i) high postprocedural bleed risk (including most major surgeries lasting > 45 min and cardiothoracic and orthopedic surgeries) and (ii) low, or non-high, postprocedural bleed risk (including invasive procedures, minor surgery, and same-day office procedures), as suggested for patients receiving anticoagulants during the periprocedural period . Emphasis should also be placed on certain “low-bleed risk” procedures (e.g., pacemaker or defibrillator insertion or minor neurospinal procedures) due to the potential serious consequences of a bleed. We suggest that studies state explicitly the general type of procedure or surgery performed (i.e., cardiac catheterization, joint replacement).
Periprocedural antiplatelet regimens
We propose that studies make explicit mention of the type, dose, and timing of each preprocedural and postprocedural AP agent used (including ASA, clopidogrel, ticagrelor, and prasugrel), as well as other agents such as ASA combined with dipyridamole and cilostazol. Mention should also be made of whether one antiplatelet agent was switched for another in the periprocedural setting (i.e., clopidogrel for ASA) and whether loading doses of antiplatelet agents were used in the postprocedural setting. Last, mention should be made of the type and dose of other antithrombotic agents in the postprocedural period (i.e., use of heparin for thromboprophylaxis).
Thromboembolic and major bleeding outcome definitions
For arterial thromboembolism, most studies of periprocedural AP therapy used major adverse cardiac and cerebrovascular events (MACCEs), including stroke, myocardial infarction, systemic embolism, stent thrombosis, and cardiovascular death, as the primary outcome. We propose that studies in periprocedural AP therapy explicitly use MACCEs as part of the primary outcome.
Many studies defined “bleeding” using TIMI (Thrombolysis in Myocardial Infarction) or GUSTO (Global Use of Strategies to Open Occluded Coronary Arteries) definitions. These definitions included intracranial hemorrhage, bleeding that causes hemodynamic compromise, bleeding that requires intervention, a ≥ 5 g dL−1 drop in hemoglobin, or a 15% absolute decrease in hematocrit as definitions of major, severe, or life-threatening bleeding or a bleed score that was based on a point accumulation depending on the severity of hemorrhage . Other studies have included definitions that have approximated ISTH definitions of major bleeding, including overt bleeding with a fall in hemoglobin > 2 g dL−1, unexpected transfusion of ≥ 2 units of packed red blood cells, bleeding into a critical organ, fatal bleeding, and bleeding requiring rehospitalization, prolonged hospitalization, or need for repeat surgery . Last, the 2008 European medicines agency guidelines incorporate estimated blood loss calculation and percentage of patients transfused with allogeneic blood as part as part of bleed definitions . Within the definition of critical organ bleeding as a component of major bleeding, intracerebral bleeding (which incurs a mortality rate of > 45–50%) should be explicitly stated. We propose that future studies adopt the Perioperative Ischemic Evaluation (POISE)-2 definitions of serious bleeding, which we believe capture all of the essential elements of periprocedural bleeding associated with AP therapy, with a time frame of 30 days to capture bleed events (P.J. Devereaux, personal communication, 2012). The POISE-2 definitions of serious bleeding include the following.
This includes fatal bleeding or bleeding that leads to:
- Significant hypotension that requires inotrope or vasopressor therapy, or
- Urgent (within 24 h) surgery (other than superficial vascular repair), or
- Intracranial hemorrhage.
This includes bleeding that is not specified under ‘life-threatening’ and satisfies at least one of the following criteria: a postoperative hemoglobin ≤ 7 g dL−1 and a transfusion of ≥ 2 units of red blood cells; a hemoglobin drop of ≥ 5 g dL−1 and a transfusion of ≥ 2 units of red blood cells; a transfusion of ≥ 4 units of red blood cells within a 24-h period; or bleeding that leads to one of the following interventions (i.e., embolization, superficial vascular repair, nasal packing) or is retroperitoneal, intraspinal, or intraocular (confirmed clinically or on imaging).
Recommendations for standardized reporting of periprocedural antiplatelet therapy
We propose the following recommendations for future randomized controlled trials and cohort studies on periprocedural AP therapy:
- We suggest a standardized description of a patient's thromboembolic risk based on the RVSI. For patients receiving dual AP therapy, we propose that future studies incorporate the suggested periprocedural thromboembolic risk stratification (Table 1), including (i) whether AP therapy is indicated for primary or secondary prevention of cardiovascular events and (ii) timing of the index cardiovascular event.
- There should be a detailed description of each patient's elective invasive procedure or surgery, preferably documenting duration based on anesthesia records if available.
- There should be a detailed description of the type and dose of AP therapy, the timing of preprocedural and postprocedural discontinuation and reinitiation, and a description of the use of loading doses.
- The primary thromboembolic outcomes should be MACCEs, including stent thrombosis
- The POISE-2 definitions for life-threatening and major bleeding should be used.
- A 30-day postprocedural follow-up period should be used to report adverse events, which should be linked to a specific patient group and procedure as per bleed risk.
Harmonized reporting would facilitate across-study comparisons of outcomes and would enable study- and patient-level meta-analyses, thereby allowing more robust assessments of the benefits and risks of specific periprocedural AP management strategies. We suggest that the recommendations from this ISTH statement be applied to both randomized controlled trials and prospective cohort studies. Future research in the field should validate the RVSI in patients specifically receiving AP therapy and our proposed thromboembolic risk stratification scheme for patients receiving dual AP therapy to assess an aggressive periprocedural AP approach (e.g., continuing AP therapy, bridging with a heparin or glycoprotein IIb/IIIa inhibitor during AP interruption, or use of platelet transfusion) [23, 24] or a less aggressive one (e.g., discontinuing dual or single AP therapy) for subgroups of patients with low, moderate, and high thromboembolic risk undergoing specific procedures. The POISE-2 definitions for periprocedural life-threatening and major bleeding should be specifically assessed for patients receiving AP therapy. Last, an acceptable upper boundary for the absolute arterial thromboembolic event rate and its relation to major bleeding—possibly as bivariate end points—would be beneficial to design and power future studies in the field of periprocedural AP therapy.
A. C. Spyropoulos, P. Albaladejo, A. Godier, A. Greinacher, G. Hron, J. Levy, C. M. Samama, and J. D Douketis all contributed to the critical writing, revision of the intellectual content, and final approval of the published version of this article.
Disclosure of Conflicts of Interest
The authors state they have no conflict of interest.