Article Text


Aspirin plus clopidogrel for optimal platelet inhibition following off-pump coronary artery bypass surgery: results from the CRYSSA (prevention of Coronary arteRY bypaSS occlusion After off-pump procedures) randomised study
  1. Vito Antonio Mannacio1,
  2. Luigi Di Tommaso1,
  3. Anita Antignan2,
  4. Vincenzo De Amicis1,
  5. Carlo Vosa1
  1. 1Department of Cardiac Surgery, University of Naples Federico II, Naples, Italy
  2. 2Department of Cardiology, Azienda Ospedaliera Santobono-Pausillipon, Naples, Italy
  1. Correspondence to Dr Vito Antonio Mannacio, Via S Domenico 62, 80127 Naples, Italy; vitomannacio2{at}


Objective To determine the individual variability in the response to aspirin and/or clopidogrel and its impact on graft patency after off-pump coronary artery bypass grafting.

Design A single-centre prospective randomised controlled study designed according to the Consolidated Standards of Reporting Trials statement. Randomisation was obtained by a computer-generated algorithm.

Setting University medical school in Italy.

Patients 300 patients who underwent off-pump coronary artery bypass grafting were randomised to receive aspirin (n=150) or aspirin plus clopidogrel (n=150).

Intervention Aspirin 100 mg or aspirin 100 mg plus clopidogrel 75 mg daily was initiated when postoperative chest tube drainage was ≤50 ml/h for 2 h and patients were followed up for 12 months.

Main outcome measures Qualitative and quantitative assessment of platelet function, angiographic evaluation of coronary revascularisation by 64-slice CT and clinical outcome.

Results In the aspirin group, 49 patients (32.6%) were aspirin resistant and, in the aspirin-clopidogrel group, 19 patients (12.6%) were aspirin and clopidogrel resistant. The platelet response to aspirin was similar in all aspirin responders despite the study arm (Aspirin Reaction Units 313.2±44.8 vs 323.6±53.6; p=0.07). The platelet response to clopidogrel was enhanced by aspirin in patients responsive to both aspirin and clopidogrel (synergistic effect) compared with responders to clopidogrel only (P2Y12 Reaction Units 139.9±15.5 vs 179.4±18.5; p<0.001). Combined therapy was associated with a reduced vein graft occlusion rate (7.4% vs 13.1%; p=0.04). Antiplatelet resistance was a predictor of graft occlusion (RR 3.6, 95% CI 2.5 to 6.9; p<0.001). Synergistic aspirin and clopidogrel activity was a strong predictor of vein graft patency (RR 5.1, 95% CI 1.4 to 16.3; p<0.01).

Conclusions Combined clopidogrel and aspirin overcome single drug resistances, are safe for bleeding and improve venous graft patency.

  • Off-pump CABG
  • aspirin
  • clopidogrel
  • drug resistance
  • graft occlusion
  • aorta
  • great vessels and trauma
View Full Text

Statistics from

Platelet inhibition represents a therapeutic mainstay in patients who undergo coronary artery bypass grafting (CABG). It is associated with a reduced risk of death, ischaemic complications and improved graft patency.1–4

Aspirin is the most popular platelet aggregation inhibitor in daily practice, but its clinical efficacy is limited by an incomplete antiplatelet response in 10–60% of patients, depending on the population studied and the specific definition used.5–7 Clopidogrel, a second-generation thienopyridine, is a promising alternative but there is a large interindividual variability of response in 4–40% of patients.8 ,9 Double antiplatelet therapy with aspirin plus clopidogrel has been suggested to optimise platelet inhibition; however, it remains unknown whether clopidogrel covers the resistance to aspirin or vice versa. Moreover, it is unclear whether aspirin alone, as recommended for conventional procedures, is adequate following off-pump CABG which, unlike cardiopulmonary bypass, induces a temporary state of hypercoagulation.10 In order to quantify whether double therapy overcomes antiplatelet resistance and its impact on graft patency and outcome after off-pump CABG, we performed a randomised trial comparing combined aspirin and clopidogrel treatment with conventional aspirin. Final angiographic evaluation of graft patency was obtained with 64-slice multidetector CT.


The CRYSSA trial is a prospective randomised controlled study designed according to the Consolidated Standards of Reporting Trials statement and performed at Federico II University of Naples (Italy). The study was approved by the ethics committee of the Federico II University Medical School of Naples and by the hospital's Institutional Review Board. Each patient gave written informed consent.

Patient population and study design

From December 2006 to October 2009, 300 consecutive patients undergoing isolated CABG for stable coronary artery disease who received at least one saphenous vein graft (SVG) were enrolled. The design of the study is illustrated in figure 1. To avoid any misleading interference, preliminary exclusion criteria were as follows: contrast allergy; additional cardiac or surgical procedures; contraindications for off-pump surgery; emergency operations for acute coronary syndrome or for failed percutaneous coronary intervention occurring within hours of the procedure; platelet count <70x109/l; contraindications to antiplatelet treatments (active bleeding or bleeding diathesis); previous recent treatment (15 days) with any kind of antiplatelet drugs; previous recent treatment (30 days) with steroidal or non-steroidal anti-inflammatory drugs; previous recent treatment (20 days) with lipophilic statins; Moderate renal failure (serum creatinine >2.0 mg/dL)2 high-sensitivity C reactive protein levels of >5 mg/l as a marker of acute or chronic preoperative inflammatory response.11 Diabetic patients were excluded because of increased platelet aggregability and higher prevalence of platelet aspirin resistance.12 Patients with active liver disease or increased liver enzyme levels were excluded as well because clopidogrel, which is a prodrug, needs to be oxidised by the hepatic cytochrome P450 system to its active metabolite.13

Figure 1

Flow of participants through study.

Coronary stenoses were evaluated by the Quantitative Coronary Angiography Data System (Centricity Carddas Xi2, GE Healthcare, Burlington, Vermont, USA). The number of coronary lesions was defined as the total number of stenoses ≥70%. Coronary blood flow was assessed by the thrombolysis in myocardial infarction (TIMI) frame-counting method with a frame counter on the cine viewer. The corrected TIMI frame count was calculated for the longer left anterior descending artery.14 The study cohort was very homogeneous for demographic and clinical features (table 1).

Table 1

Preoperative clinical and demographic features

All patients underwent off-pump CABG through a median sternotomy. Heparin was given at a dose of 300 IU/kg to achieve a target-activated clotting time of >400 s. At the end of the grafting procedure the effect of heparin was reversed by protamine administration (1:1 ratio). The left internal mammary artery was used in all patients and routinely anastomosed to the left anterior descending artery; the right internal mammary artery and/or the radial artery and SVG were additional conduits. Intracoronary shunts (Medtronic Inc, Minneapolis, Minnesota, USA) were routinely employed and the size was registered as indicative of the target artery diameter. Graft flow was measured by means of a transit-time flowmeter (HT 323-CS; Transonic Systems, Ithaca, New York, USA) at the end of the procedure.

The need for conversion to cardiopulmonary bypass due to mechanical and/or electrical instability in the operating theatre was a further exclusion criterion, as well as postoperative mechanical ventilation >36 h, significant bleeding (>500 ml at 1 h, >1000 ml in the first 3 h or >200 ml/h in the first 5 h after surgery) and the need for secondary sternotomy or for intra-aortic balloon pump support.

At this time point, patients who met the preliminary and postoperative selection criteria were randomly assigned to aspirin alone or double antiplatelet treatment by a computer-generated algorithm and antiplatelet therapy was initiated. According to the study design, 150 patients received aspirin 100 mg or 150 aspirin 100 mg plus clopidogrel 75 mg. The doses of aspirin and clopidogrel used were in agreement with current literature.15 ,16 Randomisation was fully blinded without taking account of clinical or demographic features. Antiplatelet drugs were administered when postoperative chest tube drainage was <50 ml/h for two consecutive hours and were continued daily for 12 months. To avoid potential interference of concomitant drugs, all patients received hydrophilic statins (rosuvastatin) independently of the cholesterol level and were free from proton-pump inhibitors (when necessary, ranitidine was used). No additional non-steroidal anti-inflammatory drugs, low molecular weight heparin or antifibrinolytics (eg, aprotinin, tranexamic acid) were used during the first 15 days after surgery.

Specimen collection and data definition

Venous blood samples for platelet aggregation assessment were drawn in vacutainer 2 ml tubes containing 3.2% sodium citrate and analysed within 4 h. Samples were collected at baseline, at 12 h after surgery before administration of the study drugs; at 24 and 48 h; at 3, 5, 7 and 12 days after first administration of the study drugs; and every 30 days thereafter. The time points were selected on the basis of previous findings suggesting that the first inhibitory response to the standard clopidogrel and aspirin regimen occurs within 24 h and appears to last for 5–30 days.17 Platelet function was assayed by the VerifyNow System (Accumetrics, San Diego, California, USA). This includes the VerifyNow Aspirin Test, which uses arachidonic acid as the agonist to measure the effect of aspirin on platelet reactivity, and the VerifyNow P2Y12 test, which measures direct inhibition of clopidogrel on the platelet P2Y12 receptor. The VerifyNow Aspirin Test results are reported as Aspirin Reaction Units (ARU), which indicate the amount of thromboxane A2-mediated activation of G protein-coupled receptors involved in platelet aggregation (the lower the ARU value, the greater the degree of platelet inhibition). The VerifyNow P2Y12 Test results are reported as P2Y12 Reaction Units (PRU), which indicate the amount of ADP-mediated aggregation specific to the platelet P2Y12 receptor, and the percentage inhibition which is the percentage change from baseline aggregation (the lower the PRU level and the higher the percentage of inhibition, the greater the degree of platelet inhibition). ARU values >550 are consistent with impaired aspirin-induced inhibition of platelet function whereas PRU values >230 and platelet inhibition >30% were considered consistent with impaired clopidogrel-induced inhibition of platelet function.18 ,19 Aspirin or clopidogrel resistance was defined as a persistently impaired drug response after 1 week of treatment.

Angiographic outcome

All patients were scheduled for 64-slice multidetector CT (Aquilion; Toshiba Medical Systems, Rome, Italy) angiographic evaluation at 12 months after surgery. Heart rate >70 beats/min was reduced with intravenous propranolol. Two-dimensional reconstructions (curved multiplanar reformation) of the coronary arteries and grafts were performed on several planes to assess the patency of the grafts and anastomoses. Two experienced readers, blinded to the study drugs, evaluated the CT images by consensus reading. The quality of the anastomosis and conduits was graded according to Fitzgibbon and colleagues.20 For statistical purposes, the small number of SVGs classified as diseased (III-HP-LP grade) were recorded as patent.

End points

The primary end point was assessment of whether the addition of clopidogrel to aspirin increases the number of patients fully responding to platelet inhibition after off-pump CABG. Secondary end points were: (1) the impact on graft occlusion at 12 months after off-pump CABG; (2) the impact on major and minor bleeding (defined according to the CURE trial15); and (3) the incidence of major adverse cardiac and cerebrovascular events (MACCEs), defined as a composite end point including cardiac deaths, any repeat revascularisation (percutaneous coronary interventions or CABG), cerebrovascular accident and documented myocardial infarction. According to the joint ESC/ACCF/AHA/WHF definition, myocardial infarction was diagnosed when three different criteria were fulfilled: (1) indicative ECG changes (ST-T changes, new Q-wave appearance, reduction of R waves >25% in at least two leads); (2) echocardiographic evidence of new akinetic/dyskinetic ventricular wall segments; and (3) biochemical indicators (maximal concentration of troponin I exceeding the 99th percentile of the value for the reference controlled group on two subsequent samples or maximal value exceeding five times the normal upper limits on one occasion).21

Statistical analysis

The primary outcome for sample size calculation was aspirin non-responders. Assuming a postoperative percentage of aspirin non-responders of 40%, we hypothesised that double antiplatelet therapy may reduce the occurrence of an impaired postoperative response to antiplatelet therapy by about 10%.19 ,22 Thus, a total sample of 300 patients (150 in each group) would provide a statistical power of more than 90% to detect a probability of 0.05 at the two-sided α level.

Continuous variables were expressed as means±SD and categorical data as proportions. Comparison between continuous variables in the two groups was made by means of the Student t test for normally distributed features values. The Mann–Whitney U test was used for variables not normally distributed. Categorical variables were analysed with the χ2 test or Fisher exact test, when appropriate. Analysis of risk factors for end points was performed, calculating the RR with a 95% CI. Multiway analysis of variance with correction for serial measurements was performed for evaluation of platelet aggregation data. Variables with p values of <0.05 were considered significant and were analysed in a multivariate logistic regression model to assess the effect of each risk factor on the results. Statistical analysis was performed with SPSS V.13.0 for Windows.


Clinical course

A total of 960 conduits were implanted. The operative data are shown in table 2. At the time of enrolment, all patients were in sinus rhythm with no or minimal inotropic support (dobutamine 3–5 μg/kg/min). Postoperative echocardiographic examinations revealed the same or slightly better left ventricular performance as preoperatively. Enzymatic, ECG and echocardiographic examinations evidenced perioperative myocardial infarctions in four patients (2.5%) in the aspirin group and in five patients (3.3%) in the aspirin-clopidogrel group. All made an uneventful recovery. The first doses of study drugs were administered at a mean of 28±12 h after surgery. The amount of chest tube drainage after study drug administration was similar in both groups: 115±80 ml in the aspirin group vs 125±70 ml in the aspirin-clopidogrel group (p=0.2). No patient developed thrombotic or bleeding complications in the early postoperative period after aspirin or clopidogrel administration.

Table 2

Operative data for all patients

Events during follow-up are shown in table 3. Antiplatelet therapy was discontinued in patients with major bleeding (gastrointestinal or urinary) and these patients were not further evaluated. Despite the apparent 50% reduction of MACCEs resulting from double antiplatelet therapy, the difference did not achieve statistical significance in our population of patients (RR 0.2, 95% CI 0.7 to 5.9; p=0.1).

Table 3

Clinical and angiographic results

Platelet function testing

Baseline platelet function was normal in both groups. Twelve hours after surgery, before the study drugs were given, it remained essentially unchanged compared with preoperative values. The postoperative platelet counts were slightly reduced from baseline in both study groups (from 210x109/l to 190x109/l).

Figure 2 shows the time-related antiplatelet response for each study arm. First platelet inhibitory activity appeared between 48 and 72 h after first study drug administration, while maximal inhibition occurred between 3 and 5 days. After 5 days of treatment, when the maximal inhibition of platelet aggregation occurred, 49 patients (32.6%) in the aspirin group were aspirin resistant, 48 (32%) in the aspirin-clopidogrel group were aspirin resistant 33 patients (22%) were clopidogrel resistant. Only 19 patients (12.6%) in the aspirin-clopidogrel group were resistant to both aspirin and clopidogrel (figure 3), and this reduced to 16 (10.6%) after 30 days of treatment. All patients defined as aspirin or clopidogrel responders within 7 and 30 days after commencement of treatment remained so at all time points during follow-up. Hence, aspirin and clopidogrel resistance was not a transient phenomenon in our population of patients who underwent off-pump CABG. When divided into groups according to aspirin and/or clopidogrel response, no baseline clinical or demographic characteristic was identified as a predictor of resistance to platelet inhibitors.

Figure 2

Time-related platelet inhibition reported as number of non-responsive patients at all time points of follow-up for the two study groups.

Figure 3

Schematic of aspirin and aspirin-clopidogrel resistance for each study group. -- and + indicate resistance or response, respectively, to aspirin or clopidogrel.

After the maximum inhibitory response to clopidogrel and aspirin occurred, mean ARU and PRU values as well as the percentage of inhibition remained essentially similar at all time points of follow-up without significant differences within (p=0.7) or between (p=0.09) groups. Mean ARU values were similar in all aspirin-responsive (313.2±44.8 vs 323.6±53.6; p=0.07) and all aspirin-resistant patients (611.2±36.3 vs 605.8±37.1; p=0.3), regardless of the study arm. The response to clopidogrel showed a different pattern. In patients who were responsive to both clopidogrel and aspirin there was an additional inhibitory effect on platelet aggregation (synergistic effect) compared with responders to clopidogrel but not to aspirin (mean PRU and percentage of inhibition 139.9±15.5 and 20±10%, respectively, in responders to both aspirin and clopidogrel compared with 179.4±18.5 (p<0.001) and 35±15% (p<0.001) in responders to clopidogrel only).

Angiographic results

Six grafts were not visualised with diagnostic image quality; 922 grafts from 288 patients were evaluable for the purpose of the study (table 3). Significant univariate predictors of graft occlusion were age >65 years, previous MI, preoperative TIMI frame count >30, use of SVG, ≥3 coronary artery disease, target coronary artery <1.5 mm in diameter, resistance to antiplatelet therapy. Multivariate analysis of SVG occlusion confirmed that preoperative TIMI frame count >30 (RR 2.2, 95% CI 1.2 to 3.2, p<0.001), target coronary artery <1.5 mm in diameter (RR 2.1, 95% CI 1.1 to 2.9, p<0.001) and antiplatelet resistance (RR 4.4, 95% CI 3.8 to 7.2, p<0.001) were independent predictors of graft occlusion. Interestingly, the synergistic effect of aspirin and clopidogrel, when it occurred in the aspirin-clopidogrel group, was a strong predictor of vein graft patency (RR 5.1; 95% CI 1.4 to 16.3, p<0.01).


The results of the CRYSSA study indicate that (1) the antiplatelet effect of aspirin is independent of additional clopidogrel administration; (2) the antiplatelet effect of aspirin is independent of clopidogrel resistance; (3) the antiplatelet effect of clopidogrel is independent of aspirin resistance; (4) aspirin plus clopidogrel have an additive effect which extends the efficacy of treatment to a larger number of patients; (5) aspirin in combination with clopidogrel in patients responsive to both drugs displays a synergistic inhibitory effect which potentiates the antithrombotic efficacy of clopidogrel; and (6) combined antiplatelet therapy improves SVG patency and does not increase bleeding complications compared with aspirin alone.

Clinical trials and experimental studies have shown that vein graft failure during the first year after CABG is triggered by two consecutive phases: a phase of thrombotic occlusion during the first postoperative month and a late phase of intimal hyperplasia.23 ,24 During the first year after surgery, 10–20% of venous grafts occlude. Reflecting this graft attrition, angina recurs in up to 20% of patients.24 Thus, current guidelines recommend antiplatelet therapy with aspirin starting early after surgery in order to improve graft patency and clinical outcome.25 The use of clopidogrel coupled with aspirin has been proposed for high-risk patients and current guidelines recommend double antiplatelet therapy only in patients with non-ST elevation acute coronary syndrome up to 9–12 months after the CABG procedure.26 Despite these indications, clopidogrel is frequently prescribed by surgeons after CABG in the belief that it will prevent graft occlusion and improve clinical outcomes. Although prospective randomised controlled trial data have been lacking in the field,27 a few observational studies in the cardiac surgery literature have suggested that clopidogrel decreases vascular death, myocardial infarction and stroke after coronary surgery. In a case series of patients undergoing off-pump CABG, Ibrahim et al reported that additional clopidogrel improved SVG patency early after surgery.28 In their observational study, Gurbuz et al noted that clopidogrel was associated with a trend towards less symptom recurrence and fewer adverse cardiac events.29 Most recently, Gao et al reported the results of a randomised study in which aspirin plus clopidogrel were more effective in venous graft patency than aspirin alone in the short-term after CABG.30

Our study expands on this topic. None of these mainly clinical studies evaluated qualitative and quantitative platelet reactivity after off-pump surgery and its impact on graft patency and MACCEs. To that purpose and to avoid all possible risk of bias, we followed strict enrolment criteria and treated all patients with the same doses of aspirin and/or clopidogrel for the same period of time. Our results showed that aspirin or clopidogrel resistance is a frequent phenomenon in patients undergoing off-pump CABG. The combined use of aspirin and clopidogrel, however, reduces the overall variability in the individual response. In this study, clopidogrel combined with aspirin showed either an additive effect in terms of the number of patients with significant inhibition of platelet aggregability or a synergistic effect in terms of degree of platelet inhibition.

Nearly complete angiographic follow-up provided the necessary clinical support to the CRYSSA study. The addition of clopidogrel to aspirin reduced SVG occlusion. This favourable effect was mainly evident in cases of grafts placed in vessels <1.5 mm in diameter and with a higher TIMI frame count as a consequence of platelet-dependent thrombotic occlusion which occurred more frequently in the conduits placed on the smaller vessels with poor run-off. Notably, our results failed to detect any significant effect of combined antiplatelet therapy on variable internal mammary artery (IMA) graft occlusion which, because it occurs only rarely, needs a larger sample for statistical power and is probably influenced by factors other than SVG.

A potential concern of combination antiplatelet therapy is safety, particularly bleeding. In the present study, the incidence of major or minor bleeding complications was low regardless of whether aspirin was given alone or with clopidogrel. This should not imply that clopidogrel combined with aspirin after off-pump CABG is without risk, but it does suggest that these risks can be minimised when administered according to a defined protocol. In our opinion, the daily dose of clopidogrel 75 mg plus aspirin 100 mg was, perhaps, the best compromise between antithrombotic effectiveness and safety in our patients.

In conclusion, our results indicate that the combination of aspirin and clopidogrel, acting through different pathways, is more effective on platelet aggregation than aspirin alone, has beneficial effects on vein graft patency and, when administered according to a defined protocol, is safe for bleeding.


The authors are grateful to Dr GiovanBattista Mannacio from Department of Mathematics (Statistics Section), Imperial College London for assistance with statistical analysis.


View Abstract


  • Competing interests None.

  • Ethics approval The study was approved by the ethics committee of the Federico II University Medical School of Naples and the hospital's Institutional Review Board. The presented data are anonymised and there is no risk of identification.

  • Patient consent Written consent was obtained from each patient.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.