Clopidogrel Dosing Scheme in Pediatric Cardiac Patients 0–24 Months Old Using P2Y₁₂ Reaction Unit Monitoring

Abstract

OBJECTIVE

The use of clopidogrel for postprocedural primary thromboprophylaxis in pediatric cardiac patients is becoming more common. This study aimed to explore, using P2Y12 reaction unit (PRU) values, whether the currently recommended static low clopidogrel dose of 0.2 mg/kg/day for patients under 24 months is optimal, or if a gradual dosage escalation would be more appropriate; and to propose a hypothetical dosing scheme for clopidogrel thromboprophylaxis in these patients.

METHODS

Exploratory, retrospective cohort study in cardiac patients 0–24 months old receiving clopidogrel for thromboprophylaxis between 2018 and 2021. Data collected from medical records included patient demographics and diagnoses, clopidogrel dosing and duration, PRU values, concomitant anticoagulant and antiplatelet therapies, adverse events, and outcomes. Exponential and linear regression analyses were employed to model dosage as a function of time using therapeutic PRU values and to identify the best-fit dosing scheme for clopidogrel.

RESULTS

Forty-four cardiac patients on clopidogrel for thromboprophylaxis were included. No statistically significant difference was observed between the predicted dosing from the fitted and the referent regressions, indicating that the observed clopidogrel doses that achieved a therapeutic PRU followed a dosing gradient inconsistent with the static low dose (0.2 mg/kg/day) recommended by the Platelet Inhibition in Children On cLOpidogrel (PICOLO) trial during the first 24 months of age.

CONCLUSIONS

Pediatric cardiac patients may require a gradual escalation of clopidogrel dosing with age, in contrast to the static low dose recommended by the PICOLO trial during the first 24 months of life. An age-based dosing scheme may prove beneficial for this age group. Prospective studies evaluating age-based clopidogrel dosing in this patient population could offer further insight into this relationship.

Introduction

Children younger than 24 months of age with congenital or acquired heart disease face an elevated risk for life-threatening postprocedural thrombotic events, stemming from their underlying pathophysiology or the presence of foreign materials in their cardiovascular system following corrective or palliative procedures.^1,2 Aspirin is commonly employed for thromboprophylaxis in these patients, but it may provide insufficient prevention due to its inhibition of only a single pathway of platelet activation.^1–3 Clopidogrel, a P2Y₁₂ adenosine diphosphate–receptor antagonist, irreversibly inhibits the P2Y₁₂ component of adenosine diphosphate receptors on the platelet surface, thereby reducing platelet aggregation through an alternative pathway.^4–6 The therapeutic effect of clopidogrel can be monitored using P2Y₁₂ reaction unit (PRU) assays.

Dual-antiplatelet therapy with aspirin and clopidogrel has been shown to reduce thrombotic event rates in adults with coronary artery disease, peripheral artery disease, and stroke.^7–9 Moreover, the use of clopidogrel, either as an adjunct to aspirin or as a standalone therapy for pediatric patients with aspirin resistance or intolerance, has significantly increased in recent years.^10,11 In 2008, the Platelet Inhibition in Children On cLOpidogrel (PICOLO) trial recommended a fixed daily clopidogrel dose of 0.2 mg/kg/day for cardiac patients younger than 24 months.¹² Other studies proposed a starting dose of 1 mg/kg/day for children at this age, representing a sudden 5-fold increase from the PICOLO-recommended dosage.¹³

Given the lack of consensus regarding the appropriate clopidogrel dosing for postprocedural primary thromboprophylaxis in cardiac patients younger than 24 months receiving aspirin, the purpose of this study was to explore this knowledge gap by modeling clopidogrel doses based on therapeutic PRU values to determine whether the currently recommended static low dose of 0.2 mg/kg/day during the first 24 months of age is optimal, or if a gradual dose escalation would be more appropriate. Additionally, we aimed to propose a hypothetical age-based dosing scheme for clopidogrel thromboprophylaxis in cardiac patients aged 0–24 months.

Materials and Methods

Study Setting and Design.

This study was conducted in a tertiary-level multidisciplinary intensive care unit and a cardiac step-down unit at Le Bonheur Children’s Hospital. It was designed as a retrospective exploratory cohort study.

Study Population and Data Collection.

The study included patients with a primary cardiac diagnosis aged 0–24 months who received clopidogrel as adjunctive therapy to aspirin for primary postprocedural thromboprophylaxis and underwent PRU assessments at this single institution across all months between 2018 and 2021. Procedural indications for clopidogrel initiation included high-risk palliative cardiac surgeries (such as systemic-to-pulmonary shunts) and the placement of stents or ventricular assist devices. We excluded patients older than 24 months and patients on extracorporeal membrane oxygenation.

Data collected from the patients’ charts included patient demographics, cardiac diagnoses, relevant laboratory values, clopidogrel dosage, PRU values, indication for clopidogrel initiation, duration of clopidogrel therapy, adjuvant anticoagulant and antiplatelet therapy, concomitant use of CYP2C19-interacting medications, adverse events, length of hospital stay, and survival to hospital discharge. Laboratory parameters evaluated included P2Y₁₂ reaction units using the VerifyNow™ P2Y₁₂ assay, aspirin reaction units (ARU) measured using the VerifyNow™ Aspirin assay, serum creatinine, liver function tests, hemoglobin, platelet count, international normalized ratio, prothrombin time, and activated partial thromboplastin time.

Adverse events included hemorrhage and thrombosis. Hemorrhage was further categorized into gastrointestinal (defined as a positive occult blood test from a gastric and/or stool sample), respiratory (defined as bleeding from an endotracheal tube or nasopharyngeal source), and intracranial (defined as the presence of blood within the cranial vault as determined by imaging). Thrombosis was defined by echocardiogram or cardiac catheterization as evidence of complete or partial flow occlusion within the heart and central vascular structures. Of note, deep vein thrombosis of peripheral vessels was not included because of its high prevalence in our cardiac population, largely from long-term central venous access.

Institutional Approach to Anticoagulation and Antiplatelet Therapy.

Given the high risk of life-threatening complications associated with thrombosis within the cardiovascular structures and/or implanted materials in our pediatric cardiac population, intravenous thrombin inhibitor infusions (unfractionated heparin or bivalirudin) and cangrelor are routinely initiated either post-procedure or during periods of fasting for patients requiring thromboprophylaxis, provided no bleeding complications are identified. At our institution, heparin is preferred as the thrombin inhibitor, while bivalirudin is used for patients with heparin resistance. In patients at higher risk of mortality due to thrombosis within the cardiovascular structures or implanted materials (e.g., neonates with systemic-to-pulmonary artery shunts or stents for ductal-dependent circulation), cangrelor is additionally administered. Upon initiation of enteral feeds, these medications are transitioned to enteral aspirin and clopidogrel. Aspirin is typically initiated first at a dose of 5 to 10 mg/kg/day, a range selected to allow rounding to 20.25, 40.5, or 81 mg to facilitate administration at discharge. Once a therapeutic ARU value confirms platelet inhibition, thrombin inhibitor infusions (and cangrelor, if applicable) are discontinued, and clopidogrel therapy is initiated. Patients are initiated on the recommended starting dose of 0.2mg/kg/day, as established by the PICOLO trial, and subsequently titrated to achieve therapeutic PRU values.¹²

P2Y12 Monitoring Assay.

The VerifyNow™ P2Y₁₂ assay (Accriva Diagnostics, San Diego, CA) was used to assess the degree of platelet inhibition achieved by P2Y₁₂ inhibitors, such as clopidogrel.¹⁴ This assay measures light transmittance through a blood sample, expressing platelet inhibition values as PRUs. An elevated PRU indicates normal platelet function, while a low PRU reflects platelet inhibition. According to the manufacturer’s specifications, the reference PRU range for normal platelet activity is 194 to 418, with values less than 194 indicating adequate platelet inhibition. Conversely, PRU values of 194 or higher, considered as subtherapeutic, indicate insufficient platelet inhibition and an associated increased risk of thrombosis.¹⁴ PRU values may be falsely elevated if hematocrit is < 32%, platelet count is < 119/microliter (μL), or fibrinogen is < 171 mg/deciliter (dL); conversely, PRU values may be falsely low or unreadable when hematocrit exceeds 52%, platelet count is more than 502/μL, or fibrinogen is more than 559 mg/dL.^14–17 At our institution, the timing of PRU monitoring is generally determined in accordance with the manufacturer’s recommendations, at or after seven days after clopidogrel initiation. Additionally, PRU assessments are typically repeated after a dose change, when PRU values exceed 194, in cases of bleeding or thrombosis, in coagulopathic or hypercoagulable states, at the initiation of CYP2C19-interacting medications, or at provider discretion. These follow-up assessments are performed to confirm accuracy or to guide clopidogrel dose adjustments until therapeutic values are achieved.

Study Aims.

The primary aim of this study was to model clopidogrel doses based on therapeutic PRU values to evaluate whether the currently recommended static low dose of 0.2 mg/kg/day for cardiac patients who are also receiving therapeutic aspirin and younger than 24 months is optimal, or if a gradual dose increase over the first 24 months of life, potentially following a linear or exponential pattern, would be more appropriate. The secondary aim was to propose a hypothetical dosing scheme for clopidogrel thromboprophylaxis in cardiac patients aged 0–24 months.

Statistical Analysis.

Data were summarized with medians and IQR for continuous metrics and frequencies and percentages for categorical values. Dosing was modeled as a function of time (age) using both linear and exponential regressions to examine 2 possible dosing patterns. The motivation for exploring both linear and exponential regressions stems from controversy over whether children grow at a linear or exponential rate during the 0- to 24-month period.¹⁸ To statistically determine which model and thus dosing pattern was a better fit to our observed data, the Akaike Information Criterion (AIC) was calculated for both models. The lower AIC indicated the better model fit. The AIC follows the traditional definition of −2 log-likelihood + 2k, where k is the number of estimated parameters in the model, in which the model with the lower AIC is a better fit. The AIC () function in R (R version 4.4.0 R Core Team, 2021) was used for this calculation.

To explore whether a gradual increase in dose, as opposed to the static 0.2 mg/kg/day, would be more appropriate, a hypothetical dosing pattern using both exponential and linear regressions was modelled (beginning at 0.2 mg/kg/day and ending at 1 mg/kg/day) as a function of time (in months) from 0 to 24. For ease of reference, these 2 regressions are denoted as the “referent” regressions. Then, our fitted regressions were compared with the “referent” regressions using the mean squared error. The mean squared error summarizes the difference in the predicted dosage between the 2 regressions, if this value is relatively small, it indicates the model outputs are similar. It is defined as the average of the squared differences of the predicted dosages. All analyses were conducted in the full cohort (0–24 months), but because of the limited number of patients aged 12–24 months, repeat analyses were performed on the subset of patients aged 0–12 months, thus resulting in 2 analyses. All analyses were conducted in R version 4.4.0 (R Core Team, 2021. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria; https://www.R-project.org/) with a significance level of 0.05.

Results

A total of 47 pediatric cardiac patients receiving clopidogrel for postprocedural thromboprophylaxis with PRU values were evaluated. Three patients were excluded because they were older than 24 months, resulting in 44 patients included in the study and 65 therapeutic PRU values. The median age and weight of the cohort were 2.1 months (IQR, 0.47–5.61) and 3.8 kg (IQR, 3.2–5.6), respectively. Sex distribution was equal, with the majority of the patients identified as African American and Caucasian (Table 1). Before transitioning to enteral antiplatelet agents, all patients received either unfractionated heparin (n = 40, 90%) or bivalirudin (n = 4, 10%), and 45.5% additionally received cangrelor. These infusions were discontinued in all patients at the time of clopidogrel initiation. All patients received dual antiplatelet therapy. Aspirin was initiated first, followed clopidogrel once therapeutic ARU values were confirmed. Clopidogrel was administered to all patients as an extemporaneous prepared 5 mg/mL suspension without a loading dose. Twelve patients (27.3%) were concurrently treated with a CYP2C19-interacting medication (proton pump inhibitor, n = 9, and azole antifungal, n = 3). The median therapeutic clopidogrel dose and PRU values for the cohort were 0.22 mg/kg/day (IQR, 0.19–0.34) and 153 (IQR, 117–179), respectively. Survival to hospital discharge accounted for 70.5% of the cohort (Table 1). Supplemental Table S1 details the median laboratory values obtained at the time of clopidogrel administration.

Table 1.

Patient Characteristics, Anticoagulant and Antiplatelet Therapy and Adverse Events

Patient Characteristics	N = 44*
Age group, n (%) 0–28 days old	24 (55)
29 days to 12 mo old	18 (40)
12–24 mo old	2 (5)
Age (mo), median (IQR)	2.1 (0.47–5.61)
Weight (kilograms), median (IQR)	3.8 (3.2–5.6)
Male, n (%)	22 (50)
Race/Ethnicity, n (%) African American	21 (47.7)
Caucasian	17 (38.6)
Other/Multiracial	4 (9.1)
Hispanic	2 (4.5)
Cardiac diagnoses, n (%) Complex single ventricle	12 (27)
Coronary aneurism	1 (2)
Dilated cardiomyopathy on VAD	2 (5)
Hypoplastic left heart syndrome	17 (39)
LPA stenosis	1 (2)
Pulmonary atresia ± IVS	4 (9)
Tricuspid atresia	1 (2)
Tetralogy of Fallot ± PA	6 (14)
Indication for clopidogrel initiation, n (%) Cavo-pulmonary connection and stent	1 (2)
Cavo-pulmonary connection and VAD	2 (5)
S-P shunt	27 (61)
S-P shunt and stent	2 (5)
Stent (ductal, LPA, RVOT)	9 (20)
VAD	2 (5)
Presence of coronary aneurism	1 (2)
Length of hospital stay (days), median (IQR)	34 (24.5–82)
Survival to hospital discharge, n (%)	31 (70.5)
Anticoagulant and Antiplatelet Therapy	N = 44*
Transition from anticoagulant and antiplatelet infusions, n (%)
• Thrombin inhibitor	44 (100)
• Unfractionated heparin	40 (90)
• Bivalirudin	4 (10)
• Cangrelor	20 (45.5)
Concomitant aspirin use, n (%)	44 (100)
Duration of clopidogrel use (days) to death or hospital discharge, median (IQR)	17 (9–31)
Concomitant CYP2C19 metabolism medications, n (%) Proton pump inhibitor (omeprazole)	9 (20.5)
Azole antifungal (fluconazole)	3 (6.8)
Adverse Events	N = 44*
Hemorrhage, n (%) Gastrointestinal	4 (9)
Respiratory	1 (2.2)
Intracranial (worsening of preexisting bleed)	1 (2.2)
Thrombosis, n (%) Shunt occlusion	2 (4.5)
LPA occlusion	1 (2.2)

IVS, intact ventricular septum; LPA, left pulmonary artery; PA, pulmonary atresia; RVOT, right ventricular outflow tract; S-P, systemic-to-pulmonary; VAD, ventricular assist device

^* Median (IQR); frequency (%).

Hemorrhage was observed in 6 patients (13.6%), with a median clopidogrel dose of 0.22 mg/kg/day (IQR, 0.21–0.29), a median PRU of 141 (IQR, 115–168), and a median ARU of 464 (IQR, 416–483). Among these cases, all experienced minor and transient bleeding complications, except for 1 neonate who had worsening of a preexisting intraventricular hemorrhage. For safety, clopidogrel therapy was temporarily discontinued in these patients. Thrombosis occurred in 3 patients (6.8%), with a median clopidogrel dose of 0.20 mg/kg/day (IQR, 0.18–0.21), a median PRU of 164 (IQR, 147–183), and a median ARU of 480 (IQR, 448–509). These included shunt occlusions (n = 2) and a left pulmonary artery occlusion (n = 1); all 3 underwent a catheter-based intervention. Coagulation laboratory values and platelet counts were within normal reference ranges for all patients with complications. Table 1 outlines patient characteristics, anticoagulant and antiplatelet therapy usage, and adverse events.

To determine whether a gradient dosing scheme would be superior to current recommendations for achieving therapeutic PRU, and to identify the optimal age-based dosing gradient, our observed data (clopidogrel doses that achieved therapeutic PRU) were fit with linear and exponential regressions. Exponential regression provided a better fit across all analyses, as indicated by a lower (more negative) AIC (Figures 1 and 2). Subsequently, the regressions generated from our observed data were compared with the reference regressions described above, which modeled gradient dosing (0.2–1 mg/kg/day) as a function of age (0–24 months) following an exponential pattern. The linear and exponential regressions, along with their prediction equations, are shown in Supplemental Figures S1 and S2. All mean square errors were small, all less than 0.002. This finding suggests that a dose-escalation strategy during the first 24 months of life may be more accurate than maintaining a static low dose of 0.2 mg/kg/day, as originally recommended in the PICOLO trial.

Figure 1.

Exponential and linear trend of clopidogrel doses with therapeutic P2Y12 reaction unit (PRU) values for patients 0–24 months.

Figure 2.

Exponential and linear trend of clopidogrel doses with therapeutic P2Y12 reaction unit (PRU) values for patients 0–12 months

Had the data been in agreement with PICOLO recommendations, a better fit with linear models and a slope (regression coefficient for age) approaching zero would have been observed. The data demonstrate that exponential modeling results in a better fit, and that the slopes of the fitted linear regressions were statistically different from zero, as indicated in Table 2, further supporting a gradient in the dose required to achieve therapeutic PRUs. Using regression analysis, a hypothetical clopidogrel dosing scheme for thromboprophylaxis in pediatric cardiac patients aged 0–24 months is proposed (Table 3). This regimen involves a gradual, age-based dose escalation, adjusted every 3 months to facilitate dosing refinements. When retrospectively comparing this proposed dosing scheme with the clopidogrel doses at which therapeutic PRU values were achieved, concordance was observed in 80% of the cases. Of note, 1 patient who experienced a thrombotic event and another who experienced a bleeding complication were receiving clopidogrel doses lower than those recommended by our age-based dosing scheme, despite having therapeutic PRU values. All other thrombotic and hemorrhagic complications occurred in patients whose dosing was concordant with the scheme.

Table 2.

Linear Regression Coefficients for Modeling Observed Clopidogrel Dosage as a Function of Age

Age Group	Coefficient* (Standard Error)	p value
0–24 mo	0.0164 (0.0044)	0.0005
0–12 mo	0.0196 (0.0060)	0.0018

^* Regression coefficient for age. The slopes (coefficients) for each fitted linear regressions are statistically different from zero (p < 0.05), supporting that a gradient dose increase is required to achieve therapeutic P2Y12 reaction unit values in these age groups.

Table 3.

Proposed Scheme for Clopidogrel Dosing in Pediatric Cardiac Patients Aged 0–24 Months Grouped in 3-Month Periods using the Prediction Models

Age, mo	Proposed Dose Scheme, mg/kg/day
0 to ≤3	0.2–0.3
>3 to ≤6	0.3–0.4
>6 to ≤9	0.4–0.5
>9 to ≤12	0.5–0.6
>12 to ≤15	0.6–0.7
>15 to ≤18	0.7–0.8
>18 to ≤21	0.8–0.9
>21 to ≤24	0.9–1

Discussion

Our study characterizes clopidogrel dosing based on PRU values in cardiac patients aged 0–24 months. We identified 2 novel insights that contribute to the existing literature. First, the patient-driven PRU observed data revealed a dosing gradient that is inconsistent with the static low dose recommended by the PICOLO trial during the first 24 months of age. Second, we proposed a gradual, age-based clopidogrel dose escalation every 3 months, which demonstrated dose concordance with the clopidogrel dose at which therapeutic PRU values were achieved in 80% of cases.

Children younger than 24 months of age with congenital or acquired heart disease, particularly those who have undergone procedures, are considered at elevated risk for thrombotic events, which constitute a significant source of morbidity and mortality in this population.^1,2 Antiplatelet therapy is routinely employed as an effective form of enteral thromboprophylaxis in these patients.^1,2,19 Dual-antiplatelet therapy with aspirin and clopidogrel has been associated with lower rates of thrombotic events compared with aspirin monotherapy in adults with coronary and peripheral artery disease as well as stroke.^8,9 The use of clopidogrel as an adjunctive therapy to aspirin, or as a standalone treatment for patients exhibiting aspirin resistance or intolerance, is increasing in the pediatric population.^10,11 Consequently, appropriate dosing and effective monitoring of clopidogrel therapy are of paramount importance.

The PICOLO study was a prospective, multicenter, randomized, placebo-controlled trial that evaluated the pharmacodynamic effects of clopidogrel in children aged 0–24 months with a systemic-to-pulmonary artery shunt or stent. Of note, 79% of patients were concurrently receiving aspirin. Platelet aggregation was assessed using light-transmission aggregometry. The study found that a clopidogrel dose of 0.2 mg/kg/day in this population achieved a level of platelet inhibition comparable to that observed with standard adult dosing (75 mg/day). Based on these findings, the investigators recommended a static daily clopidogrel dose of 0.2 mg/kg/day for cardiac patients under 24 months with no specified increase.¹² In their retrospective review, Filkenstein et al¹³ reported their experience with clopidogrel therapy for primary and secondary prevention of thrombotic phenomena in children with complex heart disease after interventional cardiac catheterization, with aspirin coadministered in 80% of patients. They recommended a starting dose of 1 mg/kg/day, reflecting an acute 5-fold increase from the PICOLO dosage at this age. This dose discrepancy suggests that individualized dosing regimens, tailored to age and weight, may be necessary to ensure therapeutic efficacy and safety in patients younger than 24 months.

In our study, the clopidogrel doses from patient observations that achieved therapeutic PRUs revealed a dosing gradient that contrasts with the fixed low dose recommended by the PICOLO trial for patients under 24 months of age. Furthermore, analogous to the dosing strategies employed for other medications, we propose that a structured clopidogrel dosing scheme, guided by therapeutic PRU monitoring in pediatric cardiac patients under 24 months old receiving concurrent ASA, may prove advantageous.²⁰ Our proposed regimen was mathematically designed and adjusted to simplify the calculation of dose escalation with age. Although it is consistent with the majority of our patient-driven clopidogrel data, it does not achieve absolute precision. Therefore, additional studies are required to evaluate the efficacy of our scheme, as well as any newly developed clopidogrel dosing regimens. The ontogeny of CYP2C19 in neonates and infants should also be considered.^21,22 Genotypic evaluation for CYP2C19 poor metabolizers was not available at our institution; however, future studies should incorporate developmental pharmacogenomics to support more individualized antiplatelet therapy. ^21,22 We propose that our dosing scheme serve as a foundational reference for clopidogrel dosing calculations, incorporating a comprehensive assessment of patient characteristics, procedural considerations, adverse events, pharmacokinetic and pharmacodynamic profiles, and PRU values to tailor individualized treatment regimens.

Our study has several limitations. It is a retrospective database review based on a single-center experience with a small sample size, which may limit the generalizability of the findings. Additionally, the exploratory nature of the study and the inability to account for unmeasured baseline patient characteristics, procedural outcomes, and age-related changes pose potential confounding factors. Although complications were limited, the retrospective design and available data prevent definitive attribution to dosing levels or drug-drug interactions. As expected, owing to the natural progression of certain cardiac diseases and the subsequent reduction in the need for intensive thromboprophylaxis with age, our study included a limited cohort of patients aged 12–24 months, in alignment with the population studied in the PICOLO trial. Consequently, the model estimates based on observed data exhibited poorer fit for patients aged 12–24 months, and the dosing scheme values for this age group are mostly derived from regression-based predictions. We recommend that future studies with a larger number of patients in the 12- to 24-month age group be conducted to better characterize and validate dosing patterns in this population. Last, pharmacogenomic monitoring is not currently conducted at our institution; therefore, the analysis was limited to collecting data on concomitant CYP2C19 substrate medications, despite clopidogrel’s interactions with cytochrome P450 enzymes.

Conclusions

Our study suggests that cardiac patients may require a gradual escalation of clopidogrel dosing with age, in contrast to the static low dose recommended by the PICOLO trial during the first 24 months of life. An age-based dosing scheme, such as the one proposed, may prove beneficial for this age group. Prospective studies evaluating age-related clopidogrel dosing in this patient population may further optimize therapy.

ABBREVIATIONS

AIC

akaike information criterion;

ARU

aspirin reaction units;

PICOLO

Platelet Inhibition in Children On cLOpidogrel;

PRU

P2Y12 reaction units