Early tirofiban versus heparin for bridging dual antiplatelet therapy in patients undergoing coronary endarterectomy combined with coronary artery bypass grafting: a multicenter randomized controlled trial protocol (the THACE-CABG trial)

Creation of distal surgical anastomoses can be difficult in patients with severe complex or diffuse coronary artery disease (CAD) undergoing coronary artery bypass grafting (CABG). Incomplete revascularization can affect patient survival and quality of life [1]. Coronary endarterectomy (CE) was first introduced in 1957 by Bailey et al. and is now being performed as an adjunct with CABG (CE-CABG) in patients with diffuse CAD to achieve complete revascularization [2]. Although long-term survival is comparable between CE-CABG and CABG alone, early surgical outcomes related to postoperative myocardial infarction (POMI) and mortality are worse in those undergoing CE-CABG [3‐5]. These patients mainly experience graft occlusion owing to thrombosis in endarterectomized regions. During endarterectomy, the vascular intima is carefully dissected and removed, which exposes the subendothelial tissue to blood flow, causing fibrin platelet mural thrombus and activation of the coagulation cascade [6‐8]. Dual antiplatelet therapy (DAPT) and postoperative anticoagulation may play a role in avoiding POMI after CE-CABG.

Current clinical guidelines recommend DAPT as soon as feasible after CABG and continuation for at least 12 months [9]. However, the use of antithrombotic medication within 24 h of CE-CABG is controversial. Reported rates of hemorrhage, POMI, and short-term mortality after CE-CABG vary substantially between studies [10‐12]. Some centers have begun to use a heparin infusion 4 h after surgery in patients undergoing CE if the chest tube output is less than 50 to 100 mL/h. However, the incidence of POMI appears to be relatively high even with early heparin infusion [10, 11].

Tirofiban is a small-molecule peptide that selectively inhibits fibrinogen–platelet GP IIb/IIIa binding and reduces ischemic events in patients with acute coronary syndrome (ACS) undergoing percutaneous coronary intervention [13, 14]. Since perioperative use of GP IIb/IIIa inhibitors does not increase surgical bleeding after CABG [15‐17], we hypothesized that early tirofiban administration before initiation of DAPT reduces the incidence of thrombosis and POMI after CE-CABG but does not increase bleeding incidence or severity.

The primary goal of the THACE-CABG trial is to demonstrate the safety of early tirofiban administration after CE-CABG (phase I). If safety is demonstrated, its efficacy in reducing the incidence of major cardiovascular and cerebrovascular events (MACCEs) will be evaluated in comparison to early heparin administration (phase II).

The THACE-CABG trial is a multicenter prospective randomized controlled trial. Patients will be randomized into one of two parallel groups at a 1:1 allocation ratio: an experimental group (tirofiban) and a control group (heparin). The safety portion of the study is a noninferiority trial with a primary endpoint of cumulative chest tube drainage in the first 24 h after surgery. The efficacy portion of the study is a superiority trial that will evaluate MACCEs in the 30 days after surgery as the primary endpoint. MACCEs include all-cause death, nonfatal MI, nonfatal acute ischemic stroke, and early revascularization.

Beijing Anzhen Hospital, Capital Medical University, Beijing, China.

Beijing Tiantan Hospital, Capital Medical University, Beijing, China.

Peking University Shenzhen Hospital, Shenzhen, Guangdong Province, China.

The First Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang Province, China.

Inclusion criteria

Exclusion criteria

A member of the study team will check eligibility and obtain written informed consent after explaining the study and answering any questions during a visit to the research site for the baseline assessment. No study operations will take place until consent is provided.

Individuals who choose to participate in the study will be asked to give permission for long-term monitoring of their electronic medical records (with no additional patient contact). Permission will be obtained to utilize study data in other ways, such as for pooled analyses of anonymized data, natural history studies, meta-analyses, and health outcomes research.

After CE, the lack of endothelium leads to activation of the coagulation cascade [18‐20]. Therefore, antithrombotic treatment is required [20]; however, no standard protocol exists. Heparin infusion followed by warfarin for several months has been recommended by several authors [21, 22]. Pre- or intraoperative clopidogrel dosing followed by postoperative aspirin and clopidogrel administration is another anticoagulation scheme which has been used [22]. In our hospital, we have used a postoperative heparin infusion followed by DAPT (aspirin and clopidogrel or ticagrelor) for many years. Tirofiban infusion followed by DAPT is another potential regimen which is purely antiplatelet in nature.

Before surgery, patients hospitalized for isolated CABG will receive the following standard antithrombotic regimen: aspirin will be administered up to the day of the procedure; clopidogrel, ticagrelor, and prasugrel will be stopped at least 5, 5, and 7 days before, respectively.

CE-CABG will be performed by experienced surgeons via a median sternotomy utilizing either an on- or off-pump grafting technique. CE will be indicated for patients with: (1) a long, diffusely diseased coronary artery segment or severely calcified coronary artery, or (2) severe long-segment in-stent restenosis with or without major side branch occlusion [23]. CE is occasionally unanticipated prior to CABG surgery. CE-CABG is typically the last option for obtaining full revascularization. Patients who do not undergo CE have a better prognosis than those that do when antithrombotic medication is used according to current recommendations; therefore, they are not required to receive tirofiban therapy and will not be included.

Graft flow and PI will be assessed at the end of the operation before protamine administration using the VeriQ Flowmeter System (MediStim ASA, Horten, Norway), which is based on an ultrasonic transit time flow meter. If the PI is more than 5, the graft must be anastomosed again. If it remains above 5, the surgery is deemed a failure and the patient will be removed from the study.

After the procedure, protamine will be administered to reverse the heparin (100 IU heparin: 1 mg protamine) and the active coagulation time will be corrected to the pre-heparinization value of 5%. Chest tube blood loss will be evaluated 2 h after surgery; if < 150 mL, the patient will be randomized.

Eligible patients will be randomized to either the heparin or tirofiban group in a 1:1 ratio utilizing an interactive online response system. Using the biased-coin minimization approach, a static randomization list will be created and applied taking three strata into consideration: research site, patient age group (< 70 and ≥ 70 years), and gender. Figure 1 depicts the enrollment and randomization procedures.

Two hours after randomization, heparin (100 IU/kg) will be administered intravenously in 5 mL of 0.9% sodium chloride every 4 h in the heparin group. In the tirofiban group, tirofiban (0.05 g/kg/min) will be infused over 22 h using a micropump starting within 30 min of randomization. To maintain treatment blinding, a 5-mL volume of 0.9% sodium chloride will be administered to patients in the tirofiban group, and patients in the heparin group will receive an infusion of 0.9% sodium chloride via micropump. Only the nurses who prepared the fluids will know which are genuine medications. Once heparin or tirofiban stop at 24 h, DAPT will be initiated the next day of surgery and continued daily (ticagrelor 90 mg twice daily plus aspirin 100 mg daily in patients with platelet count ≥ 9 × 10⁹ cells/L; aspirin alone daily in patients with platelet count < 9 × 10⁹ cells/L). For patients who are unable to be extubated within 24 h, heparin or tirofiban will be continued until the patient is extubated, at which point DAPT will start. The timeline of antithrombotic drugs is showed in Fig. 2.

To ensure patient safety, heparin and tirofiban doses will be adjusted according to routine blood and coagulation parameters obtained 4 and 10 h after treatment initiation. If the platelet count is < 90,000/mm³ after a repeat test, tirofiban infusion will be discontinued. If the activated partial thromboplastin time increases to 2.5 to 3 times the normal upper limit value, the heparin or tirofiban dose will be halved; if more than 3 times, they will be discontinued. If the fibrinogen level decreases to < 1.0 g/L, heparin and tirofiban infusions will be discontinued and fresh frozen plasma will be administered.

Participants may leave the experiment at any time for any reason as required by the Declaration of Helsinki. Although there are no particular withdrawal requirements, researchers can withdraw patients as necessary and note the cause. All participants will be followed for clinical outcomes unless permission is expressly revoked.

Pregnant women will be excluded. Menstrual history and human chorionic gonadotropin level will be examined in non-menopausal women. Pregnancy after enrollment will not be regarded as an adverse event or significant adverse event (SAE) in and of itself and will not be considered grounds for study exclusion; however, the patient will be given the opportunity to withdraw if she wishes.

During each study visit, patient safety will be addressed. Participants will receive information regarding who to contact in case of a SAE. Those who experience a SAE will be counseled to withdraw.

Three strategies to improve subject adherence to interventions will be implemented. First, a portion of the individuals’ health care and travel expenditures will be reimbursed. Second, physicians will offer free consultations. Finally, the researchers will communicate with the subjects at least once a week.

Standard secondary CAD prevention measures will be implemented. Oral beta-blockers will be used before and after surgery to maintain heart rate between 60 and 80 beats per minute. To maintain low density lipoprotein concentration below 1.88 mol/L, an oral statin or combination of lipid-lowering medications will be prescribed. All episodes of care will be recorded in an electronic case report form (eCRF).

Patients will be contacted every week in 30 days and once a year after trial to monitor for adverse effects, provide encouragement and motivation, answer questions, and assist with any problems.

The secondary outcomes of the phase I trial are as follows: (1) the incidence of universal definition of perioperative bleeding (UDPB) in adult cardiac surgery class 3 or higher in the first 24 h after surgery [24] and (2) the incidence of Thrombolysis in Myocardial Infarction (TIMI) major bleeding in the first 30 days [25]. Table 1 shows the various classes of the universal definition of perioperative bleeding in adult cardiac surgery. TIMI major bleeding is defined as intracranial hemorrhage, > 5 g/dL decrease in hemoglobin concentration, or ≤ 15% absolute decrease in hematocrit.

The secondary outcomes of the phase II trial are as follows: (1) the incidence rates of the four different events comprising MACCEs and (2) the incidence of postoperative acute kidney injury. Both CABG and tirofiban have been associated with high risk of such injury.

All-cause death is defined as death from any cause. POMI is defined as MI within 48 h of CABG. MI is defined according to the fourth universal definition of MI: cardiac troponin concentration > 10 times the 99th percentile upper reference limit or > 20% increase in cardiac troponin concentration plus new pathologic Q waves on electrocardiography, graft occlusion on angiography, or new abnormal segmental ventricular wall motion on echocardiography. MI more than 48 h after surgery is defined as cardiac troponin concentration above the 99th percentile upper reference limit and at least one of the following: (1) new ischemia or new pathologic Q waves on electrocardiography, (2) imaging of new loss of viable myocardium or new regional wall motion abnormality in a pattern consistent with an ischemic cause, and (3) identification of a coronary thrombus on angiography [26].

All patients with stroke symptoms will be examined to rule out conditions that mimic stroke (e.g., seizure, conversion or somatoform disorder, migraine headache, and hypoglycemia) and then undergo urgent computed tomography and/or magnetic resonance imaging to confirm the diagnosis. Acute ischemic stroke is defined as a new neurologic deficit lasting more than 24 h in conjunction with imaging findings of cerebral infarction. Early revascularization is defined as any percutaneous or surgical revascularization performed for graft failure or ACS resulting from a lesion inside or adjacent to a graft within 30 days. Acute postoperative kidney injury is defined according to the Society of Thoracic Surgeons as the following: (1) serum creatinine ≥ 2 0.0 mg/dL or increase to twice the preoperative baseline concentration or (2) new requirement for dialysis [27].

The participant timeline is shown in Table 2.

We computed a sample size of 266 patients (133 per group) for the phase I trial, assuming a one-sided significance level of 2.5%, test power of 90%, chest tube drainage volume standard deviation of 500 mL, and non-inferiority limit of 200 mL. Thus, the non-inferiority will be demonstrated if the upper boundary of the 97.5% confidence interval for the mean difference is lower than 200 mL. A non-inferiority margin of 200 mL was arrived at by consensus among the investigators based on their clinical judgment and the data available at the time of trial design.

Based on data from the Anzhen Hospital database from 2020 to 2021, we estimate a 30-day MACCE incidence of approximately 40% in the control group (heparin). Using an absolute risk reduction of 25% as the anticipated intervention effect of tirofiban, acceptable alpha of 5%, and acceptable beta of 20%, 712 participants are required in each group. To allow for loss to follow-up, 730 participants will be recruited; therefore, after recruitment of the phase I trial participants, an additional 464 participants (232 per group) are required to satisfy the requirements for phase II.

The average number of annual CE-CABG procedures over the past three years at Anzhen hospital was approximately 600. Combined with the other two centers, achieving the required number of participants is feasible.

All eligible subjects will be allocated by clinical staff to either heparin group or tirofiban group in a 1:1 ratio at 2 h after CE-CABG using an interactive web response system (IWRS). A static randomization list is generated and implemented within the system which accounted for three strata: study site, patient age group (< 70, ≥ 70 years) and gender group (male, female) using the biased-coin minimization method. Each center will have a designated user of the interactive system. This user will generate a random sequence for patients who meet criteria for enrollment and communicate the allocation to the research nurse.

The random sequence will be immediately sent to a designated nurse who prepare the fluids and medications. For the tirofiban group, a 50-mL volume of tirofiban fluid and a 5-mL volume of 0.9% sodium chloride (labeled as heparin) will be prepared. For the heparin group, a heparin fluid and a 50-mL 0.9% sodium chloride (labeled as tirofiban) will be prepared.

Doctors and nurses working near the bed will have no idea which fluid is genuine medication. These fluids will be infused according their label names. Thus, all subjects, site staff, sponsor staff, assessors, and data analysts will be blinded to randomization outcomes.

All subjects, site staff, sponsor staff, staff (with exceptions as indicated below), assessors, and data analysts will be blinded to randomization outcomes until the database is unlocked. Exceptions will include (1) nurses who prepare the fluids and medications (heparin, tirofiban, or just 0.9% sodium chloride), (2) computer programmers involved in randomization and drug management processes, and (3) the biostatistician creating reports for the Data and Safety Monitoring Board (DSMB).

The DSMB will periodically examine all safety and outcomes data. Appropriate information regarding adverse occurrences will be systematically gathered and presented to regulatory authorities.

Emergency unblinding will be performed only when necessary for subject safety. Only those individuals who are required to know treatment allocation will be given this information in such an event. If medically appropriate, all subjects will resume the study treatment after recovery and continue to do so until the end of the study.

Data will be collected using a study-specific eCRF (Digital Health China Technologies Co., LTD, Beijing, China) and entered onto a paper case report form. Throughout the trial, data management personnel will validate the data. Data fields will be checked and any missing data or inconsistencies will be corrected before downloading into the study database. All identifying information will be hidden to protect patient confidentiality. All study personnel will have 24-h access to the study coordinating center. A data monitoring committee will monitor data accuracy and completeness, evaluate adverse events, and make the final decision to terminate the trial. When all data management and statistical data validation procedures have been completed, the database will be unlocked to perform the final analysis.

There are no further plans for intervention engagement tactics beyond those that have already been described.

The DSMB will consist of a chair, co-chair, and members with recognized expertise in clinical trials, cardiovascular disease, and biostatistics who are not involved in the routine conduct of the THACE-CABG trial. The DSMB will be responsible for two main tasks. First, it will regularly assess the safety and effectiveness of the drugs under investigation by reviewing periodic updates provided during the course of the study. Second, it will provide recommendations to the trial’s chair and vice chair regarding the appropriate course of action, which may involve continuing, modifying, or terminating the study.

The DSMB’s primary responsibility is study oversight. Subject risks and benefits will be determined after an unblinded evaluation of aggregate data. The DSMB charter and meeting schedule will be finalized during the first meeting. A biostatistician, who will not be blinded for report preparation, will attend meetings to provide support. Conference attendees will be able to hear updates and ask questions of the trial’s chair and/or primary investigators during the open section of the conference.

Only the lead investigator and necessary staff will have access to collected data. To protect patient confidentiality, each subject’s laboratory samples, completed forms, reports, and other data will be identified by a special participant ID number.

Not applicable: no biological specimens will be collected.

Both intention-to-treat and per-protocol analyses will be conducted. The primary safety outcome will be compared between groups using a one-sided non-inferiority test. The primary efficacy outcome will be evaluated using the chi-square test or Fisher’s exact test.

Interim assessments and safety and efficacy monitoring will be conducted by the DSMB, which will review unblinded event rates. An outside statistician will perform interim data analyses independently for the DSMB.

A formal interim analysis is planned after the expected number of accumulated primary safety outcome events accrue (phase I). If the interim analyses show clear and consistent safety in both treatment arms, the DSMB may recommend commencement of the phase II trial. If the conditional probability of rejecting the non-inferiority hypothesis appears to be unacceptably high, the DSMB may consider postponing or ceasing the phase II trial.

A subgroup analysis of primary outcomes will be performed to compare CE in the left anterior descending coronary artery with CE in other coronary arteries. Subgroup analyses according to age, gender, diabetes, left ventricular ejection fraction, type of CE (open or closed), and graft type (vein or arterial) will also be performed.

Patient-reported outcomes and an analysis of cost-effectiveness will be reported at the end of the study. Projected changes in economic costs and health outcomes from broad use of tirofiban will be quantified. Market prices or shadow prices will be used to value costs. By counting and assessing the resources used, the costs of delivering the intervention will be determined.

The analyses will not include participants who were randomly assigned but later determined to be ineligible. Additional types of non-compliance that can be identified from the research data, such as visits outside of predetermined times, will be evaluated and classified as significant or minor. Similarly, protocol violations will be rated as significant or minor. After participants with significant procedure violations are eliminated, sensitivity analyses will be conducted. Missing data will not be imputed for these analyses.

The corresponding author will provide trial datasets and algorithms upon reasonable request.

The trial steering committee comprises:

The Heart Health Research Center will serve as the data monitoring committee and receive all trial data. Data will be securely transmitted to the center for data entry and verification in accordance with their normal operating procedures.

Complete information regarding all SAEs (nature of the event, start and end dates, severity, link to the trial and/or trial protocols, and outcome) shall be noted in the medical record and eCRF within 24 h of occurrence. Such incidents will be followed up until satisfactorily resolved and stabilized. With reference to the study protocol and Reference Safety Information, the site principal investigator will utilize medical judgment to assess seriousness, causation, severity, and expectedness. The site team and main investigator will determine the relatedness and expectedness. According to the study protocol, the sponsor will validate data collection and SAEs. The sponsor will report safety information to the chief investigator or a delegate for ongoing risk/benefit assessment and collaborate with the chief investigator to submit an annual safety report to the research ethics committee until the end of the study, when an end of study form will be submitted.

The trial steering committee will assess safety in line with a predefined charter, reviewing recruiting and the overall status of the trial on a regular basis; it will also interact with the sponsor regarding safety problems. A relevant unexpected SAE will be reported to the sponsor, who will then notify the ethics committee. The study data center will offer an eCRF for central data collection of adverse events and SAEs which will be submitted to the trial steering committee every 6 months until the study is completed.

An audit of all entries into the eCRF and number of participants will be conducted annually.

Any protocol amendments will be authorized by the local ethics committee before being implemented and updated in the ChiCTR.org registration. All investigators and study participants will be informed.

Deviations from the protocol will be thoroughly documented using a report form.

The findings of the study will be communicated largely through patient timeline and public forums, conference presentations, and scientific articles. The final manuscript will be reviewed and approved by all writers.

The study began recruitment on September 1, 2022, and is expected to end in late 2026. Current protocol: version 20,220,620.