We will perform a phase 2 randomized, single-center, placebo-controlled, triple-blind, superiority clinical trial with 1:1 allocation to receive either 100 mg of aspirin per day or placebo by mouth for 12 months. Randomization will be stratified based on 2 baseline characteristics: (1) the presence versus absence of type 2 diabetes mellitus and (2) the presence versus absence of platelet hyperreactivity. The study will start following approval of the Institutional Review Board (IRB) and will follow the ethical principles of the Declaration of Helsinki of the World Medical Association. The study design is illustrated in Figure 1.

The trial involves patients on chronic intermittent hemodialysis and will be conducted in a large dialysis center (ie, with ≥2000 patients on RRT).

The study population consists of men or women with ESRD who are at least 18 years of age and have started chronic intermittent hemodialysis in the previous 3 months.

Exclusion criteria include any contraindications to ASA, concurrent treatment with anticoagulants or platelet aggregation inhibitors, and pregnancy or lactation. Patients with life-threatening conditions other than renal or vascular disease will also be excluded from the trial: all types of cancer, liver disease, AIDS, or severe lung disease. ESRD due to glomerulopathy has a different pattern of mortality [20], which precludes its inclusion in this trial. Patients on other modalities of RRT will also be excluded.

The intervention is an enteric-coated oral pill containing 100 mg of ASA, administered once daily, after lunch, over 12 months. The comparator will be a placebo tablet with the same characteristics as the active pills. Both arms will have the same administration schedule and will start treatment the day after randomization (day 0).

At every hemodialysis session, participants will receive training about the importance of taking the drug as prescribed by viewing a 10-minute educational video about medication compliance on a digital tablet.

Study patients will have monthly visits with the study team to check on treatment adherence and answer any questions. During these visits, patient compliance with treatment will be assessed by direct questioning and by counting returned tablets provided on a monthly basis. Good adherence will be defined as taking at least 80% of the prescribed daily dose and attending 100% of the visits. If a patient withdraws from the trial, the study coordinator will contact the participant to find out the reasons for withdrawal.

The primary outcome is the incidence of a composite event including all of the following: cardiac death, nonfatal myocardial infarction, nonfatal stroke, arteriovenous fistula thrombosis, and TIMI major bleeding [19] by 12 months of treatment. The secondary outcomes are effect modification of treatment efficacy by the presence of type 2 diabetes mellitus or platelet hyperreactivity and the rate of TIMI minor bleeding between treatment groups.

According to the TIMI criteria, major bleeding is defined as the presence of any intracranial bleeding (excluding microhemorrhages <10 mm evident only on gradient-echo magnetic resonance imaging), fatal bleeding (bleeding that directly results in death within 7 days), or clinically overt signs of hemorrhage associated with a drop in hemoglobin of ≥5 g/dL or a ≥15% absolute decrease in hematocrit [19]. Minor bleeding is defined as clinically overt with a hemoglobin drop of 3 to 5 g/dL [19].

The profile of platelet reactivity will be determined by multiple electrode aggregometry using a Multiplate analyzer (Dynabyte GmBH). This method was chosen because it currently is the most efficient assay used to assess platelet reactivity in humans [21]. The Multiplate will measure platelet activity (defined as aggregation capability after activation with adenosine diphosphate) as the area under the curve reported in area units multiplied by time (AU×min). Platelet hyperreactivity is defined based on the manufacturer’s recommendations and published data, with a cutoff of 50 AU×min for adenosine diphosphate–induced aggregation [22].

A blinded multidisciplinary team composed of a nephrologist, cardiologist, and nurse will follow participants and review medical records and adjudicate outcomes and end points in the study.

A medical appointment will be scheduled for weeks 0, 13, 26, 39, and 52, where the multidisciplinary team will assess the primary and secondary outcomes and report them to the Data Coordinating Center (DCC). At each visit, a venous blood sample (5 mL) will be collected in order to measure hematocrit, hemoglobin, and platelet hyperreactivity.

After the end of the treatment phase of the trial, participants will be followed for an additional 6 months to assess further adverse events. For that purpose, the multidisciplinary team will contact patients monthly during their hemodialysis sessions. The total length of the study will be 30 months.

The outcome variables are the incidence of cardiac death, nonfatal myocardial infarction, nonfatal stroke, and arteriovenous fistula thrombosis, as well as the presence or absence of bleeding and its severity based on TIMI criteria [19]. The treatment variables are the exposure to the intervention (aspirin 100 mg/d or placebo). The stratification variables are the presence versus absence of type 2 diabetes mellitus and the presence versus absence of platelet hyperreactivity.

The schedule and procedures during patient visits to the participating center (recruitment, hemodialysis sessions, and monthly assessment of outcomes) are summarized in Table 1.

Sample size was calculated based on estimates for the rate of events in the primary composite outcome over 1 year: TIMI major bleeding (placebo 1% vs aspirin 2.5% [15]), cardiovascular events (placebo 3.0% vs aspirin 2.5% [10]), and fistula thrombosis (placebo 19% vs aspirin 8% [23]), for a total 1-year event rate of 23% in the placebo group and 13% in the aspirin group. A sample size of 318 patients (159 per arm) provides 80% power to detect a difference between groups of this magnitude (corresponding to an HR of 0.53) using a 2-sided log-rank test and alpha=.05 (Power Analysis and Sample Size Software version 14, NCSS LLC). The sample size for the trial was increased to 342 patients (171 per arm) to account for estimated attrition of 10%.

The estimated hemodialysis admission rate at a large dialysis center is 50 patients per month. Therefore, we expect to include 8 to 10 patients every week, completing our recruitment goal in 9 months. We propose a recruitment period of 12 months in case enrollment is slower than expected.

Eligible patients will be identified using medical records, clinician invitation letters, and internal flyers posted in patient areas. The study team will approach eligible patients to invite them to participate in the trial. All aspects of the study will be explained and all participants who agree to participate must provide written informed consent. One venous blood sample will be collected prior to randomization in order to assess platelet reactivity profile, a stratification factor for randomization.

Patients will be allocated to 1 of the 2 study groups based on a computed-generated blocked randomization, with stratification by 2 factors (type 2 diabetes mellitus and platelet hyperreactivity). For that purpose, randomly permuted blocks of sizes 4, 6, and 8 will be used in order to maintain the integrity of the randomization and blinding [24]. The software used to generate the sequence is available at randomization.com, and the study pharmacists will coordinate the randomization, treatment assignment, and delivery of study medication [25].

The research pharmacists will determine the study allocation and randomization and will be the only individuals to know the identity of the drugs delivered.

The trial is triple-blinded to the treatment allocation: participants, study clinicians, and staff, as well as data analysts, will not know participant treatment assignments. Blinding will be assured by the use of a placebo comparator, which will be identical in look and taste to the active drug and will last until the data are analyzed.

A questionnaire will be used to assess the effectiveness of the participant’s blinding. This will be performed according to the methodology proposed by Rees and collaborators [26]. Every 2 weeks, participants will be asked to complete a survey about the intervention they think have received (active intervention, placebo, or unknown) and their level of certainty on a Likert scale. The accuracy of their answers will be evaluated using the Howard index, and differences in beliefs between 2 successive questionnaires will be assessed using the Fisher exact test. Participants’ beliefs about their treatment assignment are considered consistent if all responses are sequentially identical over time, except one change of opinion that may be explained by a lack of blinding [26].

In exceptional circumstances, unblinding may happen if knowledge of the actual treatment is essential for further management of the patient. In case of severe adverse events, investigators will discuss unblinding within 24 hours with a medical advisor from the Data and Safety Monitoring Board who is not involved with the trial. Unblinding will take place by the research pharmacy immediately after the decision is made.

We intend to minimize missing data by having monthly visits with patients. An adherence check will be done by counting returned tablets as previously stated.

Regarding data collection and storage, the study team will be trained to oversee all key aspects of the protocol: (1) methodology, forms, and tools that must be used for the collection, entry, monitoring, and editing of data; (2) appropriated methods to communicate among investigators and between investigators and participants; and (3) importance of reporting data as close to real time as possible during the course of the study.

For quality control, patient records (source documents) will be stored at the site, and the original data will be shared with the DCC. It is the responsibility of the investigator to keep, maintain, and provide the documents audited by IRB, sponsor, National Institutes of Health, US Food and Drug Administration, or other local regulatory agencies when necessary.

After the collection of study data, patient identification will be encoded, and only the investigator will have access to this information, in accordance with the Good Clinical Practices and the Declaration of Helsinki regarding confidentiality [27].

The collected data will be entered electronically in a Research Electronic Data Capture management system. This database is a cloud-based system, and it will have a backup in a hard disk in the DCC. The dataset will be encrypted in order to guarantee data safety and confidentiality.

Statistical analysis will be performed including all randomized patients according to their assigned treatment group (ie, intention-to-treat). The software used will be Stata 14 (StataCorp LLC). All testing will be 2-sided with statistical significance defined as P<.05. The primary composite outcome (first time-to-event of thrombotic events and TIMI major bleeding) will be analyzed with Kaplan-Meier curves and a log-rank test to detect difference between the groups. Cox proportional hazards regression may be used to adjust for relevant covariates, if appropriate (eg, for any unbalanced baseline characteristics that may occur by chance). Similarly, secondary outcomes will be analyzed with Kaplan-Meier curves and log-rank tests for all events. For the secondary objectives, interaction terms will be included in the Cox models above to test for interaction between treatment status and (1) the presence of type 2 diabetes and (2) platelet hyperreactivity. The proportion of patients with TIMI minor bleeding in each group will be compared with a chi-square test. If necessary, multiple imputation will be used for missing data.

A Data and Safety Monitoring Board consisting of an independent nephrologist and cardiologist (adverse events may happen mainly in those fields) and statistician is planned to oversee the trial. Based on federal regulations, an ethicist may be included. According to our inclusion and exclusion criteria, no vulnerable population is targeted.

Prior to recruitment of study subjects, the full study protocol will be submitted to the local research ethics committee and IRB for evaluation and approval.

The trial will be registered at ClinicalTrials.gov.