Erik L. Grove, Department of Cardiology, Aarhus University Hospital, Skejby, Brendstrupgaardsvej 100, DK-8200 Aarhus, Denmark. Tel.: +45 8949 6229; fax: +45 8949 6009. E-mail: firstname.lastname@example.org
Summary. Background: Previous studies have demonstrated considerable variation in the antiplatelet effect of aspirin. Objectives: To investigate the impact of platelet turnover on the antiplatelet effect of aspirin in patients with stable coronary artery disease (CAD) and to identify determinants of platelet turnover. Methods: Platelet turnover was evaluated by measurements of immature platelets and thrombopoietin in 177 stable CAD patients on aspirin monotherapy, including 85 type 2 diabetics and 92 non-diabetics. Whole blood platelet aggregation was determined using the VerifyNow® Aspirin test and multiple electrode aggregometry (MEA, Multiplate®) induced by arachidonic acid (AA) (1.0 mm), adenosine diphosphate (ADP) (10 μm) and collagen (1.0 μg mL−1). Results: Immature platelet levels significantly correlated with MEA (r = 0.31–0.36, P-values < 0.0001) and the platelet activation marker sP-selectin (r = 0.19, P = 0.014). Contrary to the VerifyNow® test, MEA significantly correlated with variations in platelet count (r = 0.45–0.68, P-values < 0.0001). Among patients with residual platelet reactivity according to AA, there were significantly more diabetics (61% vs. 41%, P = 0.027) and higher levels of sP-selectin (77.7 ± 29 vs. 70.2 ± 25 ng mL−1, P = 0.070) and serum thromboxane B2 (0.81 [0.46; 1.70] vs. 0.56 [0.31; 1.12] ng mL−1, P = 0.034). In a multivariate regression analysis, immature platelet levels were determined by thrombopoietin levels (P < 0.001), smoking (P = 0.020) and type 2 diabetes (P = 0.042). Conclusions: The antiplatelet effect of aspirin was reduced in CAD patients with an increased platelet turnover. Once-daily dosing of aspirin might not suffice to adequately inhibit platelet aggregation in patients with an increased platelet turnover.
Aspirin is the most widely used drug for the prevention of cardiovascular events. However, several studies have reported a considerable inter-individual variation in the platelet response to aspirin. Importantly, high residual platelet reactivity correlates with an increased risk of recurrent cardiovascular events [1,2]. Insufficient platelet inhibition is likely to be explained by multiple mechanisms , including an increased platelet turnover [4–7], yielding a larger population of immature, reticulated platelets. Recent studies have suggested that individuals with an increased platelet turnover may have a reduced response to antiplatelet drugs [4–7].
Circulating platelets are heterogeneous with respect to size and density, and large, dense platelets constitute a highly reactive subpopulation [8–10]. When platelet turnover is high, the proportion of large, reactive platelets released from the bone marrow megakaryocytes increases [10,11]. Newly produced platelets are capable of producing membrane and secretory proteins because of significant amounts of megakaryocyte-derived mRNA. The rate of platelet turnover can be quantified using flow cytometry with fluorescent RNA-staining dyes [12,13].
The response to antiplatelet therapy is probably affected by the ability of young platelets to produce proteins including the final common pathway of platelet aggregation, glycoprotein IIb/IIIa, cyclooxygenase (COX)-2 and alpha-granule proteins, such as fibrinogen and von Willebrand factor [5,14,15]. Furthermore, the increase in platelet turnover indicated by the presence of immature platelets may be important per se, as platelets unaffected by drugs are introduced into the blood stream possibly causing the overall platelet inhibition to be insufficient .
Recently, we have demonstrated that immature platelet levels are increased in patients with acute coronary syndromes . The study showed a trend towards increased levels of immature platelet levels in patients with coronary artery disease (CAD) and type 2 diabetes.
In the present study, we aimed to investigate the relationship between platelet aggregation, platelet activation and platelet turnover in a population of stable CAD patients including a high proportion of diabetics.
As previously described, a total of 177 patients with stable CAD were recruited from the Western Denmark Heart Registry from November 2007 to April 2008 [17,18]. Among 727 invited patients, 423 did not want to participate and 127 were either not eligible or were excluded according to the criteria given below. In this observational study, patients were eligible for inclusion if they were above 18 years of age and had significant CAD verified by coronary angiography. Our aim was to provide a study population of stable CAD patients with a relatively high-risk profile. The study population included about 50% diabetics and a high prevalence of patients with previous myocardial infarction. All diabetic patients had been diagnosed with type 2 diabetes and were treated with oral antidiabetic agents and/or insulin. In the remaining patients, fasting glucose was measured to exclude undiagnosed diabetes (glucose ≥ 7 mm) or impaired glucose tolerance. Patients were excluded if they had had any ischemic event or revascularization procedure within the previous 12 months, if their platelet count was < 120 × 109 L−1 or if they took warfarin or any drug known to affect platelet function [e.g. clopidogrel, non-steroidal anti-inflammatory drugs (NSAIDs), ticlopidin, dipyridamol]. All patients were taking 75 mg of plain aspirin daily. The study was conducted in agreement with the Helsinki-II-declaration and was approved by The Central Denmark Region Committees on Biomedical Research Ethics (M-20070180). All participants gave written informed consent.
To optimize compliance and uniform pharmacokinetics all patients received a pill-dispensing box with seven tablets of 75-mg non-enteric coated aspirin (Hjerdyl®; Sandoz, Copenhagen, Denmark) in separate compartments for each day of the week. Compliance was further optimized by face-to-face interviews and pill counting and was confirmed by measurements of serum thromboxane B2.
Standardized blood sampling was performed 1 h after aspirin intake. Samples were collected using a large bore needle (19-G) and a minimum of stasis with patients in the supine position after 30 min of rest in sitting position. The first milliliters of blood were collected in tubes without anticoagulant for separation of serum. Samples for analyzes of platelet aggregation and platelet characteristics were collected in tubes containing 3.2% citrate and EDTA, respectively (Terumo, Leuven, Belgium).
Platelet characteristics and hematological parameters
These variables were all measured using a XE-2100 hematology analyzer (Sysmex, Kobe, Japan) allowing flow cytometric detection of immature platelets as previously described [12,16]. Briefly, platelet RNA was stained with fluorescent dyes (polymethine and oxazine) before stained cells were passed through a semiconductor diode laser beam. The resulting fluorescence intensity (RNA content) and forward light scatter (cell volume) were measured, and a preset gate (XE IPF Master software) (Sysmex) discriminated between mature and immature platelets. Absolute immature platelet counts (IPC) were obtained and the immature platelet fraction (IPF) was calculated as the ratio of immature platelets to the total platelet count and is given in percent. The method for platelet count determination has previously been described . Platelet size parameters were derived from the platelet size distribution. Mean platelet volume (MPV) was calculated by dividing the platelet crit by platelet impedance count. Platelet distribution width (PDW), a measure of platelet anisocytosis, is the width of the size distribution curve in femtoliters (fL) at the 20% level of the peak. The platelet large cell ratio (P-LCR) is the number of cells falling above the 12-fL threshold divided by platelet count. Samples from all study participants were measured on the same Sysmex XE-2100 apparatus and were analyzed within 90 min of collection.
Platelet aggregation tests
Platelet aggregation was evaluated according to manufacturer’s instructions using two different point-of-care tests; multiple electrode aggregometry (MEA) and the VerifyNow®(Accumetrics, San Diego, CA, USA) Aspirin test. All analyzes were performed within 2 h of sampling.
Platelet aggregation assessed by MEA was performed using an impedance aggregometer (Multiplate®; Dynabyte, Münich, Germany). Aggregation was induced by final agonist concentrations of arachidonic acid (AA) 1.0 mm (Medinova Scientific, Glostrup, Denmark), ADP 10 μm (Sigma-Aldrich, Broendby, Denmark) and collagen 1.0 μg mL−1 (Collagen Reagent Horm; Nycomed, Linz, Austria). Aggregation was recorded for 6 min and reported as the area under the curve (AUC, Aggregation Units × minutes) .
The VerifyNow® Aspirin system (Accumetrics) is based on turbidimetric optical detection of platelet aggregation in whole blood using AA as agonist . The instrument measures light transmittance, reported as Aspirin Reaction Units.
Platelet activation and thrombopoietin
Platelet activation was determined by analysis of soluble P-selectin (sP-selectin) and thrombopoietin, the main regulator of platelet production from the bone marrow, was measured to investigate determinants of platelet turnover. Whole blood was allowed to clot for 30 min at room temperature, before serum was separated by centrifugation at 1000 × g for 15 min and stored at −80 °C until analysis. Soluble P-selectin and thrombopoietin concentrations were determined by ELISA kits according to the manufacturer’s instructions (R&D Systems Europe, Abingdon UK).
Serum thromboxane B2
Measurements of serum thromboxane B2 were performed according to Patrono et al.  with the modifications that serum was collected from whole blood by centrifugation (10 min at 2600 × g) after 1 h of clotting at 37 °C, and an ELISA was used (Cayman chemicals, Ann Arbor, MI, USA). All samples were measured in duplicate and in two dilutions. Samples with results outside the standard curve were re-assayed with appropriate dilutions.
Continuous variables are expressed as mean ± SD if normally distributed and as median (25th, 75th percentile) if not. Distributions of categorical variables were compared with the χ2-test and presented as absolute counts and percentages. All variables were tested for normality using the D’Agostino–Pearson normality test. Continuous variables were analyzed as appropriate using an unpaired t-tests or the Mann–Whitney U-test, except that the Kruskal–Wallis test was used, when more than two groups were compared. Correlations were calculated using Spearman’s rank correlation coefficient. As no standardized definition of residual platelet reactivity (RPR) exists, we defined RPR as the upper tertile of platelet aggregation. Multiple linear regression analysis was used to identify determinants of immature platelet levels. Two-sided P-values < 0.05 were considered statistically significant. Figures and statistics were performed using stata® version 10.0 (StataCorp, College Station, TX, USA) and GraphPad Prism 5 (GraphPad Software, San Diego, CA, USA).
Clinical characteristics of the study population are shown in Table 1. We included a population of stable CAD patients with a relatively high-risk profile. Thus, only patients without any cardiovascular events or revascularization procedures within the previous year were included, and there was a high prevalence of type 2 diabetics (48%), previous percutaneous coronary intervention (92%), myocardial infarction (66%), coronary artery bypass grafting (23%) and stroke (10%). Pill counting and face-to-face interviews did not reveal any non-compliant patients, and optimal compliance was confirmed by serum thromboxane B2 levels < 7.2 ng mL−1 in all patients .
Table 1. Clinical characteristics of patients with coronary artery disease (n = 177)
Data are expressed as mean ± SD, median (25th, 75th percentile) or n (%). ACE, angiotensin-converting enzyme; AT2, angiotensin 2; CABG, coronary artery bypass grafting; PCI, percutaneous coronary intervention.
66 ± 8
Body mass index, kg m−2
27.5 (25, 31)
Systolic blood pressure, mmHg
144 (129, 158)
Diastolic blood pressure, mmHg
84 (78, 91)
Diabetes mellitus type 2
81 (72, 94)
Aspirin 75 mg
Proton pump inhibitors
Platelet size heterogeneity, activation and aggregation in whole blood
As shown in Table 2, no significant correlations were observed between platelet aggregation and the IPF, whereas absolute IPC correlated with platelet aggregometry measured by MEA (r = 0.31–0.35, all P-values < 0.0001), but not with VerifyNow® results (r = 0.12, P = 0.104). IPC (but not IPF) correlated with the platelet activation marker sP-selectin (r = 0.19, P = 0.014, Fig. 1).
Table 2. Platelet characteristics and whole blood platelet aggregation in 177 patients with coronary artery disease
Multiple electrode aggregometry (Multiplate®)
AA 1.0 mm
Collagen 1.0 μg mL−1
ADP 10 μm
Data are expressed as mean ± SD or median (25th, 75th percentile). AA, arachidonic acid; IPC, immature platelet count; IPF, immature platelet fraction; MPV, mean platelet volume.
Platelet count (109 L−1) 227 (194, 265)
r = 0.45 P < 0.0001
r = 0.45 P < 0.0001
r = 0.68 P < 0.0001
r = 0.14 P = 0.066
MPV (fL) 11.0 ± 0.8
r = 0.17 P = 0.028
r = 0.18 P = 0.017
r = 0.01 P = 0.906
r = 0.02 P = 0.755
IPF (%) 3.2 (2.3, 4.3)
r = 0.11 P = 0.165
r = 0.13 P = 0.088
r = 0.03 P = 0.711
r = 0.09 P = 0.239
IPC (109 L−1) 6.8 (5.2, 9.4)
r = 0.31 P < 0.0001
r = 0.35 P < 0.0001
r = 0.33 P < 0.0001
r = 0.12 P = 0.104
Whole blood platelet aggregometry results obtained using MEA significantly correlated with platelet count (r = 0.45–0.68, all P-values < 0.0001), whereas only a trend was seen with the VerifyNow® Aspirin test (r = 0.14, P = 0.066). Similarly, the VerifyNow® analyzes were not affected by MPV (r = 0.02, P = 0.755), whereas results obtained using MEA correlated with MPV (r = 0.17–0.23, P-values < 0.05), except when aggregometry was induced by ADP (r = 0.01, P = 0.906).
Platelet count and characteristics
As expected, there was a negative correlation between platelet count and MPV (r = −0.30, P < 0.0001) and between platelet count and IPF (r = −0.32, P < 0.0001). Platelet count also correlated with PDW (r = −0.30, P < 0.0001), P-LCR (−0.29, P < 0.001) and IPC (r = 0.13, P = 0.08, Fig. 1). Significant positive correlations were seen between PDW and MPV, IPF and IPC (r = 0.64–0.95, P-values < 0.0001) as well as between P-LCR and MPV, IPF and IPC (r = 0.63–0.99, P-values < 0.0001).
Platelet characteristics and the antiplatelet effect of aspirin
As shown in Table 3, platelet characteristics differed between patients with and without residual platelet reactivity defined as the upper tertile of platelet aggregation. Most strikingly, platelet counts (MEA: P < 0.0001, VerifyNow®: P = 0.055) and IPC (MEA: P < 0.001, VerifyNow®: P = 0.060) were elevated, whereas IPF was not. MPV was slightly increased according to MEA induced by AA (P = 0.038) and collagen (P = 0.055).
Table 3. Platelet characteristics according to residual platelet reactivity (RPR)
AA 1.0 mm
No RPR (n = 58)
+RPR (n = 58)
As no standardized definition of RPR exists, we defined RPR as the upper tertile of platelet aggregation. Data are presented as median (25th, 75th percentile) or mean ± SD for lowest vs. highest tertile of residual platelet aggregation (Mann–Whitney test or unpaired t-test). IPC, immature platelet count; IPF, immature platelet fraction; MPV, mean platelet volume; RPR, residual platelet reactivity.
Platelet count (109 L−1)
10.8 ± 0.9
11.0 ± 0.8
IPC (109 L−1)
Collagen 1.0 μg mL−1
No RPR (n = 58)
+RPR (n = 61)
Platelet count (109 L−1)
10.8 ± 0.8
11.1 ± 0.9
IPC (109 L−1)
ADP 10 μm
No RPR (n = 58)
+RPR (n = 61)
Platelet count (109 L−1)
10.9 ± 0.8
11.0 ± 1.0
IPC (109 L−1)
No RPR (n = 59)
+RPR (n = 58)
Platelet count (109 L−1)
11.0 ± 0.9
11.0 ± 0.9
IPC (109 L−1)
Among patients with residual platelet reactivity according to AA-induced aggregometry, there were significantly more diabetics (61% vs. 41%, P = 0.027) and slightly higher levels of sP-selectin (77.7 ± 29 vs. 70.2 ± 25 ng mL−1, P = 0.070) and serum thromboxane B2 (0.81 [0.46;1.70] vs. 0.56 [0.31;1.12] ng mL−1, P = 0.034).
Despite treatment with the same dose of aspirin, CAD patients with diabetes mellitus type 2 had significantly higher levels of serum thromboxane B2 (1.03 [0.5;2.2] vs. 0.58 [0.35;0.92] ng mL−1, P < 0.0001) and higher levels of residual AA-induced platelet aggregation (159 [91;232] vs. 121 [68;180] aggregation units × min, P = 0.009). Furthermore, sP-selectin (77.8 ± 25 vs. 66.3 ± 28 ng mL−1, P = 0.005) and thrombopoietin levels (36 [19;56] vs. 24 [12;35], P < 0.001) were significantly higher in diabetics.
Determinants of platelet turnover
Independent determinants of platelet turnover were investigated using multivariate regression analyzes including platelet count, age, gender, body mass index, previous myocardial infarction, diabetes, smoking and thrombopoietin levels. Analyzes were performed to identify determinants of the relative (IPF) and absolute (IPC) number of immature platelets. The IPF was affected by platelet count (P < 0.001), whereas this was not the case for the IPC (P = 0.219). Both IPC and IPF were significantly affected by smoking, diabetes and thrombopoietin levels. In the multivariate regression analyzes, IPC and IPF were elevated by 17–21% in diabetics and smokers (P-values < 0.05).
The present study provides novel insight into the relation between platelet turnover and platelet aggregation in aspirin-treated patients with CAD. Previous studies have primarily evaluated the relative fraction of immature platelets, but our study showed that the absolute number of immature platelets (i) has a stronger correlation with platelet aggregation, (ii) is not significantly affected by the total platelet count, (iii) significantly correlates with the platelet activation marker sP-selectin and (iv) is stronger associated with the presence of residual platelet reactivity. Furthermore, we showed that commonly used platelet function tests seem to differ in their dependence on platelet count. Contrary to MEA, the VerifyNow® Aspirin test is not significantly affected by variations in platelet count and other platelet parameters. Importantly, this is the largest study employing measurements of thrombopoietin and immature platelets to investigate platelet production and turnover in a population of aspirin-treated stable CAD patients. In a multivariate regression analysis, smoking, diabetes and thrombopoietin levels were identified as significant determinants of platelet turnover.
Platelet turnover and platelet aggregation in whole blood
No previous studies have evaluated platelet turnover in stable CAD patients on chronic, low-dose aspirin. We specifically aimed to investigate this population in order to strengthen the external validity of our results. We have demonstrated that aspirin-treated CAD patients with a high platelet turnover have a reduced antiplatelet effect of aspirin. Our findings are consistent with recent reports of reduced antiplatelet efficacy in individuals with high levels of immature, reticulated platelets. This has been demonstrated in healthy volunteers after a single-dose administration of aspirin  and in CAD patients on dual antiplatelet therapy with aspirin and clopidogrel [6,7].
There are several possible explanations for these findings. First, the response to antiplatelet therapy is probaly affected by the ability of newly formed RNA-containing platelets to produce glycoprotein IIb/IIIa, COX-2 and alpha-granule proteins, such as fibrinogen and von Willebrand factor [5,14,15]. The fact that newly formed platelets express more COX-2 is likely to be particularly important, as COX-2 is not blocked by low-dose aspirin, thus resulting in an increased thromboxane production [5,15]. Additionally, an increased platelet turnover may be important per se, as platelets unaffected by aspirin are introduced into the blood stream possibly causing the overall platelet inhibition to be insufficient, especially during the last hours of the dosing interval. Owing to the irreversible inhibition of the platelet COX-1 enzyme, once-daily dosing of low-dose aspirin is sufficient in patients with a normal platelet turnover, whereas this may not be the case in individuals with an increased platelet turnover .
A high frequency of residual platelet reactivity according to MEA was found in patients with high platelet counts, large MPV and high levels of immature platelets. Furthermore, residual platelet reactivity was associated with higher levels of sP-selectin and serum thromboxane B2, suggesting a lower aspirin-induced COX-inhibition in these patients. These findings are in accordance with previous studies showing that complete inhibition of COX is necessary to achieve sufficient inhibition of platelet aggregation [4,22].
When residual platelet reactivity was defined according to the VerifyNow® Aspirin test, no significant association was observed with platelet parameters, indicating that this assay is less dependent on platelet parameters compared with MEA. These results extend our recent findings of a very high reproducibility of the VerifyNow® assay compared with MEA and other platelet function tests . Moreover, a very recent study on 20 healthy volunteers reported that results obtained using MEA are significantly affected by variations in platelet count below the normal range . Our study extends these findings to CAD patients with normal platelet counts.
Platelet size heterogeneity
Large platelets are known to be more reactive than small platelets [8–10]. In this population of aspirin-treated patients with CAD, MEA induced by collagen or AA significantly correlated with MPV, whereas test results obtained with the VerifyNow® Aspirin assay were not affected by MPV. The finding of a significantly higher MPV in patients with residual platelet reactivity according to AA-induced MEA might be explained by an increased hemostatic potential of large platelets. These results are consistent with findings of a lower inhibitory effect of aspirin on large compared with small platelets . Furthermore, even after indexing for surface area, large platelets have been shown to contain more dense granules and express more fibrinogen, von Willebrand factor and sP-selectin in response to AA compared with small platelets . An increased MPV has been reported in patients with myocardial infarction and as a risk factor of recurrent myocardial infarction in patients with CAD [26,27].
In our population of CAD patients, highly significant correlations between platelet distribution width, platelet large cell ratio and platelet parameters were observed. Although PDW and P-LCR have been reported to be elevated in patients with acute coronary syndromes , further studies are needed to investigate the clinical importance of these platelet volume indices. The present study was not designed to identify determinants of platelet count, but a significant negative correlation was seen between MPV and platelet count, thus supporting the well-known phenomenon of a relatively stable platelet mass (platelet count × MPV).
Diabetes mellitus and determinants of platelet turnover
Cardiovascular disease is a major cause of death in diabetics , who may have an insufficient cardiovascular protection from aspirin . We observed a higher residual platelet reactivity in aspirin-treated type 2 diabetics with CAD compared with CAD patients without diabetes, thus confirming previous findings by DiChiara et al. . The present study suggests that this may be explained by increased platelet activation as indicated by significantly higher levels of sP-selectin and less effective inhibition of COX-1 as indicated by significantly higher levels of serum thromboxane B2. Furthermore, platelet turnover might be accelerated in diabetics , and this may result in an increased number of non-aspirinated platelets, especially during the last part of the 24-h dosing interval.
Using multivariate regression analysis, we identified type 2 diabetes, smoking and thrombopoietin levels as three independent determinants of platelet turnover, thus extending our findings in patients with acute coronary syndromes . The accelerated platelet turnover in diabetics is in accordance with a previous study showing that CAD patients with diabetes have increased megakaryocyte ploidy, levels of inflammatory markers and MPV .
In a population of stable CAD patients with a large proportion of diabetics, we carefully optimized compliance, dose and timing between aspirin ingestion and blood sampling. However, when interpreting the results, some limitations should be considered. Off-aspirin measurements were considered unethical in this population and were not performed. Although statistically significant, many of the correlations observed were relatively weak. Platelet turnover was evaluated by measurements of platelet parameters and thrombopoietin levels, but might have been even better explored by including measurements of, for example, interleukin-6.
The present study supports the hypothesis that an increased platelet turnover may render aspirin less effective in patients with CAD. Accordingly, inter-individual differences in immature platelet levels may partly explain previous findings of a variable response to antiplatelet therapy. In particular, these data suggest that the reduced cardiovascular protection from aspirin observed in diabetics might be at least partly explained by an increased platelet turnover. Furthermore, we have shown that commonly used platelet function tests differ in their dependence on platelet parameters such as platelet count. We conclude that the antiplatelet effect of aspirin is reduced in CAD patients with an increased platelet turnover. Once-daily dosing of aspirin might not be sufficient to adequately inhibit platelet aggregation in patients with an increased platelet turnover.
This work was supported by the Danish Research Agency (grant number 2101-05-0052).
Disclosure of Conflict of Interests
The authors state that they have no conflict of interest.