Journal of Cardiovascular Translational Research

, Volume 7, Issue 1, pp 82–90

Bridging antiplatelet therapy in patients requiring cardiac and non-cardiac surgery: from bench to bedside


    • Ferrarotto HospitalUniversity of Catania
  • Corrado Tamburino
    • Ferrarotto HospitalUniversity of Catania

DOI: 10.1007/s12265-013-9517-5

Cite this article as:
Capodanno, D. & Tamburino, C. J. of Cardiovasc. Trans. Res. (2014) 7: 82. doi:10.1007/s12265-013-9517-5


Dual antiplatelet therapy (DAPT) is the mainstay of pharmacotherapy after an acute coronary syndrome or percutaneous coronary intervention. While patients requiring interruption of DAPT at the time of cardiac or noncardiac surgery face an increased risk of thrombotic complications, the opportunity of continuing DAPT in the perioperative period should be balanced against the risk of bleeding. Tailoring antiplatelet therapy on patient- and surgery-specific characteristics mandates a clear understanding of pharmacodynamic and clinical data on using antithrombotic agents in the perioperative period. This is also important given the introduction of novel antiplatelet agents that are already adopted in practice (prasugrel, ticagrelor) or will likely be adopted in the near future (cangrelor). This article explores the theoretical background and rationale for bridging patients on antiplatelet drugs to their surgical procedure, and provides insights on how patient and procedural characteristics translate into different considerations for the use of antithrombotic agents in the surgical setting.



Surgery (cardiac or noncardiac) is required within 2 years in about 20 % of patients who have undergone percutaneous coronary intervention (PCI) with stent implantation, being performed early within 6 weeks in 12 % of patients implanted with a bare metal stent (BMS), and within 12 months in 47 % of patients implanted with a drug-eluting stent (DES) [1]. At that time, consistent with clinical practice guidelines, these patients are typically under treatment with aspirin with or without a P2Y12 receptor antagonist (mostly clopidogrel, or prasugrel/ticagrelor as alternatives to clopidogrel in case of a recent acute coronary syndrome [ACS]) [24]. Antiplatelet therapy after stent implantation is recommended to minimize the chance of in-stent thrombotic complications during the vulnerable window when the stent is not fully endothelialized or other ischemic events may arise (particularly in the ACS setting) from non-target lesions located in the same or different vessels. However, while providing ischemic protection, antiplatelet therapy exposes surgical patients to a certain degree of bleeding risk. Therefore, at the time of surgery, physicians (and patients) are faced with the challenge of managing antiplatelet therapy to fine tune the balance between the risk of stent thrombosis or other ischemic events on one side, and the risk of life-threatening bleeding on the other side. In patients treated with DES on dual antiplatelet therapy (DAPT) with aspirin and clopidogrel, surgery has been identified as the second cause of antiplatelet discontinuation within one year, and the leading cause thereafter [5].

Understanding whether or not an antiplatelet drug should be continued (or abandoned) in the setting of a surgical procedure and tailoring drug choices on individual characteristics and varying surgical scenarios is relevant not only to interventional cardiologists, but also in the perspective of general cardiologists, anesthesiologists, and primary care physicians. This article explores the theoretical background and rationale for bridging patients on antiplatelet drugs to their surgical procedure, and provides insights on how thrombotic and bleeding profiles of these patients translate into different considerations for the use of antithrombotic agents in the perioperative period.

Risk of ischemic complications in surgical patients with previous stent implantation

Several alarming reports of ischemic perioperative complications in BMS patients undergoing early surgery after PCI have been published since 2000 [610]. Data on the risk of early surgery after DES placement also comes from small retrospective series [1117]. Importantly, in many of these studies, stent thrombosis has been consistently and significantly associated with DAPT interruption [11, 12, 15, 16, 18]. In general, premature DAPT cessation after PCI is well known to increase the risk of stent thrombosis [19]. In such cases, thrombotic complications occur at a time when metal stent struts and/or polymeric surfaces might be still not covered by tissue, hence being unsafely exposed to direct contact with blood. This phenomenon may have even more risky consequences in a clinical setting like surgery, where blood represents a prothrombotic and inflammatory environment with increased levels of cytokines, neuroendocrine inflammatory mediators, platelet counts/adhesiveness and impaired fibrinolysis [2022].

Although particularly intense in the early period after DES placement, the combined risk of death, myocardial infarction or stent thrombosis remains clinically relevant up to 2 to 3 years [2325]. Because patients who require early surgery are typically sicker than those who may wait for the completion of DAPT, a role for baseline confounding cannot be excluded. However, the understanding that surgery (early > late) and DAPT discontinuation in stented patients are potential triggers for cardiac events at the time of surgery is endorsed by clinical practice guidelines, which recommend delaying surgery 4 to 6 weeks after placement of a BMS and >12 months after placement of a DES [26, 27]. This concept has been recently challenged by a large retrospective study from Hawn et al. on 41,989 operations occurring in the 24 months after either DES or BMS implantation [28]. Major adverse cardiac events (MACE) within 30 days were associated with emergency surgery and advanced cardiac disease but were not associated with stent type or timing of surgery beyond 6 months after stent implantation [28]. Although the authors concluded that the guideline emphasis on stent type and surgical timing for both DES and BMS should be reevaluated, their findings should be judged with caution because they arise from an observational study with potential for residual confounding, where the surgical population was heterogeneous (e.g., the procedures ranged from minor outpatient to emergent inpatient operations) and clinical decision-making factors that influenced stent selection were largely unavailable or limited to administrative data.

DAPT cessation and ischemic complications: platelet rebound or withdrawal of protection?

Clustering of thrombotic events has been described in the early period after discontinuation of DAPT also in ACS patients without stents, a finding that may suggest the existence of a rebound phenomenon [29, 30]. Proposed mechanisms leading to rebound platelet activity following withdrawal of aspirin or clopidogrel are illustrated in Fig. 1.
Fig. 1

Potential mechanisms leading to rebound platelet activity following withdrawal of antiplatelet drugs. a Increased COX-1 synthesis or activity in response to chronic aspirin therapy predisposes to platelet hyperactivity following aspirin withdrawal. The first panel shows normal platelet activity in the absence of aspirin, where arachidonic acid (AA) is converted to thromboxane A2 (TxA2) via the cyclooxygenase (COX)-1 pathway. Addition of aspirin results in inhibition of COX-1 and abolishment of TxA2 formation. Long-term aspirin treatment may result in overabundance of COX-1 in platelets, yet daily aspirin administration is sufficient to palliate this effect, as depicted in the third panel. Finally, upon aspirin discontinuation, AA can be massively converted to TxA2 via uninhibited COX-1 enzymes, as shown in the last panel. b Increased P2Y12 receptor number or activity, or enhanced coupling with downstream mediators, in response to chronic clopidogrel therapy predisposes to platelet hyperactivity following clopidogrel withdrawal. The first panel shows normal platelet activity in the absence of clopidogrel, where adenosine diphosphate (ADP) stimulation results in activation of both P2Y1 and P2Y12 receptors. Addition of clopidogrel results in inhibition of the P2Y12 receptor, while P2Y1 receptor’s function is unaffected. Long-term clopidogrel therapy may result in either overexpression of P2Y12 receptor or enhanced P2Y12 receptor activity, which is masked by daily administration of clopidogrel, as shown in the third panel. Finally, upon clopidogrel discontinuation, available P2Y12 receptors bind ADP and result in marked activation of GPIIb/IIIa receptors, and increased platelet aggregation, as depicted in the last panel. Reproduced with permission from Lordkipanidzé et al. [27]

In the case of aspirin, the rebound effect has been mainly attributed to increased cyclooxygenase-1 (COX-1) activity [31]. Theoretically, because aspirin acetylates irreversibly the COX-1 enzyme, recovery of platelet function following aspirin administration depends on generation of new platelets from megakaryocytic cells [32]. However, upon discontinuation of aspirin, platelet function recovery begins sooner than what might be expected based solely on platelet pool turnover, resulting in nearly normal hemostasis within 48–72 h, while higher levels of serum thromboxane A2 and its urinary metabolite are identified up to 4 weeks in comparison with basal levels [31].

In the case of clopidogrel, increased expression of activation and inflammatory biomarkers has been described after days to weeks following the last clopidogrel intake in healthy subjects and diabetic patients compared with their on-clopidogrel period [33, 34]. Although evocative for the presence of a rebound phenomenon after cessation of clopidogrel, similar to aspirin, these studies are limited in their ability to differentiate between return to basal levels versus increased platelet activation, due to lack of baseline measurements before clopidogrel initiation. More recently, randomized pharmacodynamic studies with determination of baseline platelet function profiles, specifically designed to clarify the existence of a rebound phenomenon after cessation of clopidogrel, have excluded this hypothesis based on a wide array of platelet function tests [35, 36]. “Withdrawal of protection” might be therefore a better argument to explain the observed clustering of ischemic events after premature cessation of clopidogrel.

The emphasis on DAPT cessation as the trigger for early increased ischemic events in patients with prior stent(s) implantation has been recently challenged. The patterns of non-adherence to antiplatelet regimens in stented patients (PARIS) registry (N = 5,018) explored whether the clinical circumstances and reason for cessation of DAPT may have an impact on the cardiovascular risk that follows PCI [37]. Interestingly, the underlying context in which antiplatelet treatment was discontinued was explored in more detail than in previous studies, and stratification of patients was performed among those with physician-recommended discontinuation, brief interruption (typically due to surgery), or disruption (e.g., DAPT cessation due to non-compliance or bleeding). Physician-guided discontinuation, the most common modality of DAPT cessation (40.8 % within 2 years), was associated with a substantially lower risk of MACE, possibly due to treatment bias, with physicians appropriately discontinuing DAPT in lower risk patients. Brief interruption, occurring in 10.5 % of patients for a mean of 6 days, was not associated with an increased rate of thrombotic events, while disruption, occurring in 14.4 % of patients, was associated with a substantial increased risk of MACE particularly within the first month. Intriguingly, most adverse events occurred while patients were taking DAPT, and the overall contribution of DAPT cessation on cardiac risk was small.

In the above-mentioned study from Hawn et al. [28], 284 patients with MACE within 30 days who had operations occurring more than 6 weeks and less than 24 months from stent implantation were 1:1 matched with control patients who had operations not followed by a MACE. There was no significant difference in the probability of receiving DAPT prior to surgery (59.9 % vs 55.6 %; P = 0.43) or completely stopping DAPT >5 days (22.9 % vs 25.4 %; P = 0.49) between patients who had a MACE and those who did not. Although the study was underpowered to detect a true association, there was no significant relationship between perioperative antiplatelet cessation and 30-day MACE (odds ratio 0.86, 95 % confidence interval 0.57–1.29).

Overall, these findings are consistent with those of other studies in suggesting that briefly interrupting DAPT in proximity to surgery is not necessarily a harmful practice, especially if aspirin is continued throughout and the thienopyridine restarted as soon as possible after the procedure [38, 39]. Whether this concept can automatically be translated to all stented patients undergoing surgery remains undefined. In fact, investigation on the impact of DAPT interruption in the perioperative period should take into account the complex interplay between individual risk profiles and the varying risk (both ischemic and bleeding) linked to different types of surgery.

Risk of bleeding complication in surgical patients on antiplatelet therapy

Perioperative bleeding may occur as a consequence of the surgical procedure itself, with hazards varying according to the type of intervention, or as a consequence of being on antiplatelet drugs at the time of surgery. In many cases, discriminating between different causes of bleeding (e.g., operative vs drug-induced) is not necessarily obvious [40]. On the other hand, there are surgical procedures that do not carry an increased risk of bleeding per se, but may be associated with bleeding if the patient is on antithrombotic medications in proximity to the procedure. The American College of Chest Physicians guidelines for the perioperative management of antithrombotic therapy have identified a group of surgeries and procedures that appear to be associated with a high risk for bleeding in the context of perioperative anticoagulant and antiplatelet drug use (Table 1) [41]. Cardiac surgery should be regarded as a special category and possibly distinguished from noncardiac surgeries due to incorporation of additional risk factors for bleeding, including heparinization, pump-induced platelet dysfunction, and impaired fibrinolysis [42]. In the following paragraphs, data on the risk of perioperative bleeding with different antiplatelet drugs will be reviewed.
Table 1

Surgeries and procedures associated with an increased bleeding risk during perioperative antithrombotic drug administration (Adapted from Douketis et al. [41])

• Urologic surgery and procedures consisting of transurethral prostate resection, bladder resection, or tumor ablation; nephrectomy; or kidney biopsy in part due to untreated tissue damage (after prostatectomy) and endogenous urokinase release

• Pacemaker or implantable cardioverter–defibrillator device implantation in which separation of infraclavicular fascial layers and lack of suturing of unopposed tissues within the device pocket may predispose to hematoma development

• Colonic polyp resection, typically of large (i.e., 1–2 cm long) sessile polyps, in which bleeding may occur at the transected stalk following hemostatic plug release

• Surgery and procedures in highly vascular organs, such as the kidney, liver, and spleen

• Bowel resection in which bleeding may occur at the bowel anastomosis site

• Major surgery with extensive tissue injury (e.g., cancer surgery, joint arthroplasty, reconstructive plastic surgery)

• Cardiac, intracranial, or spinal surgery, especially as small pericardial, intracerebral, or epidural bleeds can have serious clinical consequences


Observational studies that explored the role of aspirin therapy in the perioperative period are limited in their conclusions, because aspirin therapy is per se a marker of increased comorbidities. Therefore, patients on aspirin may bleed not only because of the drug effect, but also because they have a higher prevalence of risk factors for bleeding. A randomized study of aspirin use in noncardiac surgery has been stopped prematurely leading to underpowered conclusions [43]. Keeping in mind these limitations, the trial did not show significant differences in major bleeding between aspirin and placebo (2.0 % vs 0 %, P = 0.24). A meta-analysis of 49,590 patients undergoing noncardiac surgery estimated that the incidence of bleeding complications while on aspirin therapy span from 0 % (e.g., in case of skin lesion excision or cataract surgery) and 75 % (e.g., in case of transrectal biopsy) [44]. Aspirin was not found to increase the severity of bleeding, but acted only quantitatively, with the possible exception of intracranial surgery and transurethral prostatectomy. Another meta-analysis of patients undergoing coronary artery bypass grafting (CABG) showed that aspirin carries an increased risk of postoperative bleeding, but only for doses ≥325 mg/day [45]. On the other side, aspirin use has been positively linked to graft patency at 1-year after CABG [46, 47].


In the Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE) trial, a study of DAPT with clopidogrel versus aspirin (plus placebo) after an ACS, 16.5 % of patients underwent CABG after randomization [48]. In this subset, there was a 27 % relative risk increase of protocol-defined major bleeding with clopidogrel over placebo (relative risk, 1.27; 95 % CI, 0.96 to 1.69; P = 0.095) [49]. However, the risk of bleeding varied accordingly to the time from discontinuation of clopidogrel to surgery: patients who stopped the drug for >5 days before surgery had no increase in bleeding with clopidogrel (clopidogrel 4.4 % vs placebo 5.3 %), while those who continued clopidogrel within 5 days from CABG experienced a 53 % relative increase in major bleeding (clopidogrel 9.6 % vs placebo 6.3 %). Several other studies of patients undergoing noncardiac surgery reported on the increase rates of bleeding with perioperative continuation of clopidogrel [7, 9, 50]. CABG patients on DAPT within 5 and 2 days before surgery, respectively, were found to experience higher blood loss and reoperation for bleeding [51], a finding that seems consistent regardless of whether surgery is performed on- or off-pump [52, 53]. In a systematic review of post-hoc analyses from 3 prospective randomized studies and 17 observational studies, the risk of postoperative death, reoperations for bleeding, blood loss and need for transfusion of blood products was heightened in analyses restricted to observational studies, but not in those restricted to randomized studies [54]. Another meta-analysis concluded that the risk of death with clopidogrel use at the time of CABG depends on the clinical context, suggesting that ACS patients requiring urgent CABG should proceed with surgery without delay for a clopidogrel-free period [55, 56].


In the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel–Thrombolysis In Myocardial Infarction 38 (TRITON-TIMI 38) trial, patients with ACS referred to an invasive strategy were randomized to prasugrel, a third-generation thienopyridine, or clopidogrel [57]. Only 368 (2.7 %) patients underwent CABG during the study period, in which the rate of TIMI major bleeding was greater with prasugrel than with clopidogrel (13.4 % vs 3.2 %, P < 0.001) [57]. This difference was even more pronounced when the time from the last dose of the study drug was ≤3 days prior to CABG (26.7 % vs 5.0 %, P < 0.001), but a remarkable difference was also noted when the drug was discontinued within 4 to 7 days (11.3 % vs 3.4 %, P < 0.001). Coherently, a significantly higher mean 12-h chest tube blood loss was observed with prasugrel compared with clopidogrel, while there were no differences in red blood cell transfusion [57]. In keeping with these results, the package insert recommendation of prasugrel mandate for discontinuation of prasugrel at least 7 days before cardiac or noncardiac surgery [58, 59]. Two pharmacodynamic studies have recently corroborated this recommendation [60, 61].


In the Platelet Inhibition and Patient Outcomes (PLATO) study, a trial comparing the first-in-class cyclopentyltriazolopyrimidine ticagrelor versus clopidogrel in ACS patients treated with an invasive or non-invasive strategy, 1,899 (10.2 %) patients underwent CABG at any time after randomization, and 6.8 % had CABG within 7 days after discontinuation of study treatment [62]. Ticagrelor is a reversible inhibitor of the platelet P2Y12 receptor, with faster offset of action than clopidogrel, comparable residual platelet inhibition at 24 h and 3 days after the last dose and return to baseline platelet reactivity at day 5 similar to clopidogrel on day 7 [63]. In patients from the PLATO trial undergoing CABG within 7 days after the last study drug intake, TIMI major and minor bleeding occurred similarly between patients randomized to ticagrelor and clopidogrel, with also no differences in terms of various bleeding definition rates, including fatal and life-threatening CABG-related bleeding. Such lack of difference was noted irrespective of time from last intake of study drug before CABG, including discontinuation 1 day before surgery. There was a substantial reduction in total mortality with ticagrelor, which in a further analysis has been linked to fewer deaths from cardiovascular, bleeding, and infection complications [64]. The US package insert recommendations of ticagrelor mandate for discontinuation of the drug at least 5 days prior to cardiac or noncardiac surgery, while discontinuation for at least 7 days is recommended in Europe [65, 66].

Is there any benefit with antiplatelet therapy in the perioperative CABG period?

While current guidelines recommend withholding clopidogrel and ticagrelor for at least 5 days, and prasugrel for at least 7 days before elective CABG [67], this may theoretically come at the price of ischemic events while waiting for surgery [68, 69].

In sub-analyses from the TRITON-TIMI 38 and PLATO trials, residual antiplatelet drug activity has been shown to be associated with a significant reduction of mortality in ACS patients treated with prasugrel (adjusted odds ratio 0.26, 95 % confidence interval [CI] 0.08–0.85; P = 0.025) and ticagrelor (hazard ratio 0.49, 95 % CI 0.32–0.77; P < 0.01) compared with clopidogrel [29, 62, 64]. This benefit was mostly concentrated during the first 30 days, suggesting a role for antiplatelet therapy in improving surgery-related perioperative outcomes. Notably, a similar potential benefit of antiplatelet therapy in the CABG scenario has been previously suggested in studies of clopidogrel and abciximab [49, 70]. Overall, this evidence suggests that platelet P2Y12 inhibition does ultimately have a benefit in patients who undergo CABG, including a significant impact on mortality, despite the observed increase in bleeding and need for blood transfusion.

Bridging therapy

The rationale for bridging antiplatelet therapy is that of minimizing the risk of perioperative bleeding while maintaining some degree of ischemic protection in the waiting period before surgery and in the early postoperative phase, when ongoing bleeding concerns prevent many surgeons from restarting DAPT immediately. Bridging agents should ideally have the capacity of effectively inhibiting platelets but this property should be accompanied by rapid onset/offset of action. Heparins (e.g., unfractionated heparin or low molecular weight heparin) are not optimal bridging agents, because arterial thrombosis is a process more dependent from platelets than the coagulation cascade [71]. In addition, unfractionated heparin makes platelets more prone to activation by various agonists, including adenosine diphosphate, and binds to the glycoprotein IIb/IIIa receptor on platelet, resulting in a potentially harmful prothrombotic effect [72].

Small-molecule glycoprotein IIb/IIIa inhibitors (e.g., tirofiban, eptifibatide) hold some of the characteristics of an ideal bridging agent, including rapid activity, potent antiplatelet effects and relatively short duration of action. A small study has shown the feasibility and safety of a bridging strategy with tirofiban in DES patients undergoing noncardiac surgery [73]. However, another study of 67 patients yielded contrasting results [74]. In light of these mixed results and the small numbers of these studies, the role of glycoprotein IIb/IIIa inhibitors for bridging therapy requires more investigation.

Cangrelor is an adenosine triphosphate analogue that acts as a reversible inhibitor of the P2Y12 receptor and, differently from clopidogrel, prasugrel, and ticagrelor, is administered intravenously with a rapid onset and offset of effect. In addition, the offset is more rapid and the binding to the P2Y12 receptor more specific with cangrelor compared with glycoprotein IIb/IIIa inhibitors. Overall, these characteristics make cangrelor a highly desirable agent for bridging therapy. The phase II randomized, double-blind Maintenance of Platelet inhibition with cangrelor after discontinuation of thienopyridines in patients undergoing surgery (BRIDGE) trial has recently tested cangrelor as a bridging agent for patients undergoing CABG [75]. A total of 210 PCI or ACS patients on a thienopyridine in combination with aspirin were randomized to receive either a tailored dose of cangrelor to achieve “thienopyridine-like” platelet inhibition (0.75 μg/kg/min) or placebo for at least 48 h, with the study drug discontinued 1 to 6 h before CABG. After the dose-finding stage 1 of the study, the primary efficacy endpoint of stage 2 was the proportion of patients with platelet reactivity <240 PRU according to the VerifyNow P2Y12 assay (a threshold known from literature to increase the risk of thrombotic complications) for all samples assessed during study drug infusion prior to surgery. Compared to placebo, cangrelor consistently achieved and maintained platelet inhibition at PRU values <240 (98.8 % vs 19.0 %; relative risk [RR], 5.2 [95 % CI, 3.3–8.1] P < 0.001). In addition, bridging with cangrelor did not increase major bleeding prior to surgery, although minor bleedings were more commonly documented [75]. Following discontinuation of the study drug infusion prior to surgical incision, PRU levels (P = 0.21) and the percentage of patients with PRU < 240 (P = 0.31) were similar between the study arms, a finding that highlights the fast offset of cangrelor. The results of the BRIDGE trial support cangrelor as a valuable bridging agent in the setting of CABG. Whether these findings are also adaptable to different noncardiac surgeries is hypothetical but unproven. Cangrelor is pending approval from regulatory agencies, prompted by the positive results of a recent large-scale PCI trial and a large meta-analysis summarizing the clinical trial experience with cangrelor [76, 77].

Regardless of the bridging strategy adopted, guidelines recommended that oral P2Y12 inhibiting therapy should be restarted as soon as possible after surgery in patients who presented with an ACS [2, 3]. In contrast, restarting P2Y12 inhibiting therapy in non-ACS patients is at the discretion of the physicians after CABG, if the coronary segment at risk of thrombotic complication has been bypassed. Ideally, low-dose aspirin (<100 mg/day) should be administered continuously throughout the perioperative period, with the exception of surgeries at high risk of bleeding where this may be contraindicated (e.g., neurosurgical).


Clinical practice guidelines support preventive strategies aimed at reducing the risk of discontinuing DAPT in patients who are known or possible candidates to surgery, including avoidance of unnecessary revascularization and/or placement of a BMS whenever possible. In contrast, guidelines highlight only few patient- and procedure-specific recommendations to help physicians customizing antiplatelet strategies on a case-by-case basis (Table 2). The introduction of short-acting antiplatelet agents with rapid offset of action, such as cangrelor, has the potential to address the unmet clinical need of providing ongoing ischemic protection while minimizing the risk of bleeding in patients on pre-operative DAPT.
Table 2

Limitations of current guidelines of perioperative antiplatelet management

• Do not stratify bleeding risk for individual interventions

• Do not stratify the risk of thrombosis in relation to the angiographic and clinical characteristics

• Do not offer a general strategy to be applied to different types of interventions in relation to the risk of thrombosis and bleeding, but rather refer to an assessment of the risk/benefit ratio in the individual patient

• Do not provide precise operational guidelines on the management of patients at high thrombotic risk candidate for surgery not be postponed

• In case of necessity for withdrawal of antiplatelet therapy, do not specify the mode of recovery of antiplatelet therapy


Davide Capodanno has received payments as an individual for consulting fee or honorarium from Eli Lilly, Daiichi Sankyo, The Medicines Company, AstraZeneca.

Copyright information

© Springer Science+Business Media New York 2013