Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Mar 1.
Published in final edited form as: J Vasc Surg. 2011 Jan 12;53(3):668–675. doi: 10.1016/j.jvs.2010.08.029

Variability and Persistence of Poor Response to Aspirin by Aggregometry and PFA-100 over Twelve-Months in Peripheral Arterial Disease from the Effect of Lipid Modification on Peripheral Arterial Disease after Endovascular Intervention Trial (ELIMIT)

Justin Saunders a, Vijay Nambi b,c, Kay T Kimball d, Salim S Virani b,f, Joel D Morrisett b, Alan B Lumsden c, Christie M Ballantyne b,c, Jing-Fei Dong b,e, for the ELIMIT Investigators
PMCID: PMC3197816  NIHMSID: NIHMS264611  PMID: 21227624

Abstract

Objective

To determine the prevalence of poor response to aspirin (ASA) therapy over 12-month follow-up in patients with lower extremity peripheral arterial disease (PAD), and to compare the classification agreement among different ASA response assays.

Subjects

Patients with PAD on ASA therapy at baseline were included from the ongoing Effect of Lipid Modification on Peripheral Arterial Disease after Endovascular Intervention Trial (ELIMIT), which is a randomized trial testing whether combination treatment with ezetimibe, niacin, and a statin will halt/regress atherosclerosis compared to statin monotherapy.

Main Outcome Measures

Patients who had baseline platelet testing and repeat testing at 6-month or 12-month follow-up were included. ASA responsiveness was tested using three different assays: Optical aggregation with 0.5-mg/mL of arachidonic acid (AA), optical aggregation with 10-μM of adenosine diphosphate (ADP), and platelet function analyzer-100 (PFA-100) testing with collagen/Epinephrine (Epi) loaded cartridges. ASA response was defined as AA aggregation <30%, ADP aggregation <70%, or PFA-100 Epi> 164 sec. Patients who showed response to ASA at baseline were classified as Responders. Poor response to ASA was defined as AA aggregation ≥30%, ADP aggregation ≥70%, or PFA-100 Epi ≤164 sec. Patients who showed poor response (PR) to an assay at baseline, but then were responsive at follow-up visits were classified as Initial PRs. Patients who showed poor response at baseline and all follow-up visits were classified as Persistent PRs. The classification agreement between assays was tested using the kappa statistic.

Results

Of 102 patients randomized in ELIMIT, 80 patients satisfied inclusion criteria. There were no significant baseline demographic differences between Responders, Initial PRs, and Persistent PRs. The prevalence of persistent poor response varied by the assay used; 5% of subjects (4/80) were Persistent PRs by AA aggregation, compared to 27.5% (22/80) of subjects by ADP aggregation, and 9.9% (7/71) of patients by PFA-100 Epi. Regarding the agreement of the assays, only AA aggregation and PFA-100 Epi agreed significantly (K=0.3223; 95% CI 0.15–0.493, P=0.0001), and though statistically significant, the magnitude of this agreement is small. AA aggregation and ADP aggregation did not agree (K=0.1161; 95% CI −0.004–0.236, P=0.029), nor did ADP aggregation and PFA-100 Epi (K=0.0044; 95% CI −.151–0.160, P=0.48).

Conclusions

Between 5–27.5% of PAD patients were Persistent PRs to ASA over 6–12-month follow-up using different platelet assays. Further, these commonly used platelet assays show weak agreement in determining poor response to aspirin.

Introduction

Aspirin (ASA) therapy is recommended for adverse cardiovascular event risk reduction in patients with atherosclerosis including peripheral arterial disease (PAD),1 but many patients still develop atherothrombotic events despite ASA therapy,2, 3 and cardiovascular events remain a major cause for morbidity and mortality in PAD patients. Poor anti-platelet response to ASA may explain some of the residual risk, and has been shown to be associated with increased risk of death, myocardial infarction (MI), or stroke (CVA) in patients with stable coronary artery disease (CAD) and increased post angioplasty occlusion in PAD patients.4–6 The prevalence of poor response to ASA in PAD patients is not well established, and figures from published studies vary widely depending on the assay used.6, 7 Poor ASA response has been associated with hyperactive platelets in CAD patients,8 and we have previously shown that platelets from PAD patients are more hyperactive than the platelets from CAD patients.9 There is poor agreement among commonly used platelet response assays,10, 11 thus it is unclear which assay is most reliable for identifying subjects at risk for future death, MI, or stroke. As there are limited reports on the agreement of the assays in PAD patients over time, we sought to determine the prevalence of persistent poor response to ASA over long-term follow-up in patients with PAD and to assess the agreement between commonly used assays of platelet response to ASA.

Methods

We evaluated the prevalence of poor response to ASA therapy in a PAD cohort from the Effect of Lipid Modification on Peripheral Arterial Disease after Endovascular Intervention Trial (ELIMIT) (clinical trials identifier NCT00687076, grant funding R01 HL075824), which is a randomized, double-blinded, placebo controlled trial approved by the Institutional Review Board at our institutions, that tests whether combination treatment with a statin, ezetimibe, and niacin will halt/regress the progression of atherosclerosis compared to statin monotherapy. The ELIMIT recruited subjects aged 40–80 with a history of symptomatic femoral atherosclerosis and ABI <0.9 or imaging results confirming femoral artery occlusive disease.12 After inclusion in the study, subjects underwent percutaneous transluminal angioplasty (PTA) with or without stenting. The effect of lipid-lowering therapy on thrombosis parameters is a planned secondary outcome measure. For our analysis, we inlcuded subjects on ASA who underwent baseline platelet testing and at least one other follow-up platelet testing (either 6-month and/or 12-month follow-up).

Blood Collection

After an overnight (≥8 hours) fast, and at least ≥4 hours since last smoking, patients underwent phlebotomy with either a 19-gauge or 21-gauge needle with syringe to reduce effects of high fluid shear stress on platelet function. Sodium citrate (0.38% final concentration) was used as an anticoagulant. The first 2-mL of blood was discarded to reduce injury-related platelet reactivity. All platelet function assays were performed on the same vial of blood. Whole blood was directly used for assays by the Platelet Function Analyzer (PFA-100) (Dade-Behring, Deerfield, IL, USA). For the optical aggregation studies, platelet-rich plasma (PRP) was collected by centrifuging whole blood at 150 x g for 15 min at 25 °C. The arachidonic acid-induced aggregation was performed with undiluted PRP according to the manufacturer’s instruction. The adenosine diphosphate (ADP)-induced aggregation was performed with PRP normalized at 2.0–2.5 × 108 platelets/mL. The assays were performed within 2 hours of sample collection.

Platelet Function Assays

Optical aggregation tests utilized either of the following agonists: 0.5-mg/mL of arachidonic acid (AA) (BioData Corp., Horsham, PA, USA) or 10μM of ADP (Sigma-Aldrich, St. Louis, MO), and was measured on an optical PAP-4 aggregometer (BioData Corp.). Platelet aggregation was monitored for 10 minutes at 37 °C with constant stirring (1,000 x rpm). For the PFA-100 assays, 0.8-mL of citrated whole blood was loaded into a collagen and epinephrine coated cartridge that has an aperture allowing blood to pass through. The time required to occlude the aperture was measured and reported as closure time (CT, sec).

There are several reasons for selecting these platelet function assays. AA is a precursor of thromboxane A2 that activates platelets through COX-1 dependent pathway, which is targeted by ASA. The AA aggregation assay therefore measures the degree of platelet inhibition due to ASA. The ADP-induced platelet aggregation by activating platelet ADP receptors was also measured for two reasons. First, ADP is released from activated platelets to propagate or enhance platelet aggregation induced by a primary agonist such as AA (secondary aggregation). Second, there are early reports that AA-induced platelet aggregation is mediated by or synergized with ADP.13, 14 The PFA-100 assays platelet activation, aggregation, and thrombus formation in response to shear stress and exposure to collage and epinephrine. Thus, these assays provide perspective on multiple pathways that lead to platelet aggregation.

For AA aggregometry, poor response to ASA was defined ≥30% aggregation. For ADP aggregometry, poor ASA response was defined as ≥70% aggregation. Poor response to ASA was defined as CT ≤164 sec for the epinephrine cartridge. The thresholds we have used are based on analysis of 400 normal controls from our lab (data not shown). For quality control, 10% of samples were duplicated and all equipment underwent daily calibration before sample analyses. Our definition is similar to those used in other studies of PFA-100,15 and ADP aggregometry,16 whereas our definition of aspirin non-response by arachidonic acid aggregometry is more restrictive than that used in the study by Gum et al.16

Statistical Methods

A subject was classified by ASA response into one of three different categories based on the results of each individual platelet assay. Therefore, a single subject could be categorized differently by the different assays. The three categories were ASA “Responders,” “Initial poor responders (PRs),” or “Persistent PRs.” These categories were assessed separately for each test and for the composite of all tests.

Subjects were classified as ASA Responders if baseline platelet aggregation to arachidonic acid was <30%, or aggregation to ADP was <70%, or baseline PFA-100 Epi CT >164 sec. If a subject initially showed responsiveness to an assay, they were classified as a Responder regardless of follow-up testing since the primary aim of our study was to identify the prevalence of Persistent PRs. It was common for subjects to change from responsive to non-responsive (see Results for full description).

Initial PRs had baseline AA aggregometry ≥30%, ADP aggregometry ≥70%, or PFA-100 Epi ≤164 sec. On subsequent follow-up testing, the Initial PRs showed appropriate ASA response (AA aggregometry <30%, ADP aggregometry <70%, or PFA-100 Epi >164sec).

Persistent PRs had baseline poor response (AA aggregometry ≥30%, ADP aggregometry ≥70%, or PFA-100 Epi ≤164sec) and continued to show poor response at all follow-up tests in the same assay.

Baseline demographic and historical data were collected and compared between Responders, Initial PRs, and Persistent PRs for AA aggregometry using the non-parametric Kruskal-Wallis test for continuous data, and χ2 test or Fisher’s exact test for categorical data where appropriate. Repeated follow-up measurements from each assay type were compared to their own baseline measurements with Spearman rank correlation. The classification agreement for poor responders based on different assays calculated both as a raw percentage agreement and with the kappa statistic. For subjects where there was only one assay result, the pair was dropped from the comparison so that the number of results compared was equal.

Results

Of the 102 patients initially randomized in the ELIMIT trial, 80 subjects met qualifications for inclusion in the analysis. Among the 22 patients excluded, 4 patients lacked a baseline sample for platelet testing, and 18 subjects did not have follow-up testing after baseline. In total, all 80 included subjects had platelet testing at baseline and at least one follow-up visit. Optical aggregometry for both AA and ADP was available for 72 patients at 6-month follow-up, and 65 patients at 12-month follow-up. There were fewer PFA-100 results than optical aggregometry results as blood samples sometimes obstructed the PFA system, and hence, could not be evaluated. For PFA-100 Epi testing, there were 71 samples at baseline, 67 samples at 6 month follow-up, and 59 at 12 month follow-up. Subjects without baseline PFA-100 measurement were excluded in analysis of PFA-100 Epi classification and variability, but they were not excluded from analysis of AA- and ADP-induced platelet aggregation.

Baseline demographic data for the overall study population are reported in Table I, and baseline demographic data comparing ASA Responders, to Initial PRs, and Persistent PRs for each assay is presented in Table II. As there were a small number of women (n=5) in the study, we assessed the baseline demographics with and without women. The findings were the same; thus, we report analysis for the entire 80 subjects. Overall, the baseline characteristics of Responders, Initial PRs, and Persistent PRs were similar. In all three assays, the persistent PRs had higher age, higher hsCRP value, but these differences were not statistically significant secondary to the small number of patients in the persistent PRs subset. The individual assay information is reported below.

Table I.

Baseline demographic information for overall study cohort Continuous variables are indicated with median (range)

Age (yrs) 64.7 (42.8–85.1)
Sex (% male) 93.8%
Race (%)
 • White, not of Hispanic origin 70%
 • Black, not of Hispanic origin 18.8%
 • Hispanic 11.3%
History of Smoking (%) 91.3%
Currently Smoking (%) 45.2%
Past Medical History
 • Hypertension (%) 82.5%
 • Hyperlipidemia (%) 92.5%
 • Coronary Artery Disease (%) 21.3%
 • Stroke (%) 16.3%
 • Diabetes (%) 39.2%
Aspirin Dose (% on 325 mg) 85%
Baseline Medications
 • Clopidogrel (% taking) 26.3%
 • Statin (% taking) 97.5%
Body Mass Index (kg/m2) 29.8 (18.0–53.8)
Systolic Blood Pressure (mm Hg) 138 (92–199)
Diastolic Blood Pressure (mm Hg) 74.5 (48–98)
Total Cholesterol (mg/dL) 166 (106–337)
Triglycerides (mg/dL) 140 (43–1986)
High-density lipoprotein cholesterol (mg/dL) 40 (19–85)
Low-density lipoprotein cholesterol (mg/dL) 92 (49–244)
hs-CRP (mg/L) 2.96 (0.35–68.1)
Open in a new tab

Table 2.

Baseline demographic information by response category among the different assays

AA aggregometry (n=80) Responders (n=65) Initial PR (n=11) Persistent PR (n=4) P
 Age (yrs) 64.7 63.6 67.6 0.31
 Race (% Black, White, Hispanic) 19%, 68%, 13% 9%, 91%, 0% 50%, 50%, 0% 0.23
 ASA dose (% 325 mg) 86% 91% 50% 0.12
 Clopidogrel use (%) 26% 36% 0% 0.37
 NSAID use (%) 18% 9% 0% 0.84
 History of Smoking (%) 92% 82% 100% 0.43
 Current Smoking (%) 45% 56% 25% 0.65
 Hypertension (%) 82% 82% 100% 0.64
 CAD (%) 22% 18% 25% 0.93
 Diabetes (%) 47% 0% 25% 0.01
 Body Mass Index (kg/m2) 29.8 31.7 26.5 0.33
 Hematocrit (%) 40.9% 52.5% 35.8% 0.02
 hsCRP (mg/L) 2.83 2.98 4.27 0.61
ADP aggregometry (n=80) Responders (n=33) Initial PR (n=25) Persistent PR (n=22) P
 Age (yrs) 64.8 62.9 65.7 0.50
 Race (% Black, White, Hispanic) 15%, 79%, 6% 20%, 60%, 20% 23%, 68%, 9% 0.44
 ASA dose (% 325 mg) 82% 92% 82% 0.50
 Clopidogrel use (%) 36% 28% 9% 0.08
 NSAID use (%) 9% 7% 27% 0.20
 History of Smoking (%) 91% 92% 91% 0.99
 Current Smoking (%) 50% 43% 40% 0.77
 Hypertension (%) 82% 84% 82% 0.97
 CAD (%) 18% 28% 18% 0.73
 Diabetes (%) 42% 38% 36% 0.88
 Body Mass Index (kg/m2) 29.2 32.4 31.5 0.18
 Hematocrit (%) 41.6% 44.3% 39.95% 0.63
 hsCRP (mg/L) 2.73 2.93 3.87 0.70
PFA-100 Epi (n=71) Responders (n=44) Initial PR (n=20) Persistent PR (n=7) P
 Age (yrs) 64.6 63.5 66.4 0.55
 Race (% Black, White, Hispanic) 25%, 61%, 14% 15%, 80%, 5% 14%, 72%, 14% 0.64
 ASA dose (% 325 mg) 86% 75% 86% 0.52
 Clopidogrel use (%) 25% 35% 14% 0.52
 NSAID use (%) 20% 20% 0% 0.59
 History of Smoking (%) 91% 95% 86% 0.73
 Current Smoking (%) 43% 58% 33% 0.52
 Hypertension (%) 82% 80% 86% 0.94
 CAD (%) 30% 15% 14% 0.46
 Diabetes (%) 40% 30% 57% 0.44
 Body Mass Index (kg/m2) 29.8 28.9 31.3 0.60
 Hematocrit (%) 41.3% 39.3% 42.3% 0.80
 hsCRP (mg/L) 2.93 2.73 3.43 0.49
Open in a new tab

Arachidonic-Acid Aggregometry

By AA aggregometry, 15/80 (18.8%) subjects showed ASA poor-response at baseline. In follow-up testing, 11/80 (13.8%) of those showed response to ASA at 6-month follow-up and were classified as Initial PR (Figure I). Four subjects (5%) remained as Persistent PRs after 12-month after initial testing. Baseline demographic characteristics were generally similar between the three response categories. There were differences in the proportion of subjects with a history of diabetes (P=0.01) with a larger proportion of Responders having diabetes compared the Initial PRs (P=0.006). To measure the repeatability of the tests at different time points, Spearman rank correlations (Table III) were computed between baseline and follow-up values. For AA-induced aggregation, the 6-month follow-up test positively correlated with the baseline measurement (r=0.44, p=0.0001). Similarly, the 12-month follow-up correlation with baseline AA aggregation was also statistically significant (r=0.35, P=0.0039).

Figure I.

Figure I

Open in a new tab

Proportion of Responders (light gray), Initial PRs (darker gray), and Persistent PRs (gray stripe) depicted in bar chart. Numbers listed within each bar chart refer to number of subjects in that category. N=80 total for AA-aggregometry and ADP-aggregometry, N=71 total for PFA-100 Epi.

Table III.

Correlation coefficients, number of subjects, and P-values for statistical significance are reported

Baseline AA aggregometry (agg)
Correlation Coefficient (r) P-value
6 months AA agg (n=72) 0.44 .0001
12 months AA agg (n=65) 0.35 .0039
Baseline ADP aggregometry (agg)
Correlation Coefficient (r) P-value
6 months ADP agg (n=72) 0.49 <.0001
12 months ADP agg (n=65) 0.20 .1063
Baseline PFA-100 Epi
Correlation Coefficient (r) P-value
6 months PFA Epi (n=67) -0.03 .7980
12 months PFA Epi (n=59) 0.16 .2518
Open in a new tab

Adenosine Diphosphate Aggregometry

The ADP-induced platelet aggregation classified the highest number of subjects as Persistent PRs. By this assay, 47/80 subjects (58.8%) were identified as Initial PRs, and 22/80 (27.5%) remained Persistent PRs at follow-up (Figure I). There were no significant baseline demographic differences between the three response groups by ADP aggregation (Table II). The 6-month follow-up test (r=0.49, p<0.0001) showed a moderate but significant correlation with baseline values. On the other hand, 12-month values (r=0.20, P=0.1063) were not significantly correlated with baseline measurements. Although a larger number of subjects were defined as Persistent PRs defined by the ADP-induced platelet aggregation did not meet criteria for persistent non-response by any other assay (Figure 2).

Figure II.

Figure II

Open in a new tab

Venn diagram of the overlap in classification of Persistent PRs by the three assays tested. Numbers within the circles represent numbers of subjects satisfying criteria for persistent PR by the particular assay. Subjects who satisfied 2 definitions of persistent PR are in the overlapping areas of the circle. One subject satisfied all three criteria for Persistent PR and is shown in the overlap of all three circles.

Platelet Function Analyzer-100 Epinephrine

The PFA-100 Epinephrine assay classified 27/71 (38.0%) subjects as Initial PR, and of those, 7/71 (9.9%) remained Persistent PRs (Figure I). Of note, 2 subjects who did not have baseline values reported, had both 6-month and 12-month follow-up values that were responsive, but they were not included in the analysis. There were no significant baseline demographic differences between the response categories defined by the PFA-100 Epi assay (Table II). This assay showed the weakest correlation on repeat testing and found no significant correlation between baseline and 6-month (r= −0.03, P=0.798), or baseline and 12-month (r=0.16, P=0.2518) follow-up values.

Variability and Classification Agreement

Responders

At the baseline visit, 65/80 (81.2%) subjects were classified as Responders by AA aggregation, 33/80 (41.2%) subjects by ADP aggregation, and 44/71 (62.0%) subjects by PFA-100 Epi. Only 15 subjects were classified as a Responder by all three assays at baseline. Of those 15 Responders, only 7 were found to be responsive at baseline and all follow-up time points tested. For the individual assays, there was variability in the response status as well. Of the 65 subjects identified as Responders by AA aggregation at baseline, 59/65 (90.8%) were found to be responsive at all time points tested. For ADP aggregation, 23/33 (69.7%) of the Responders at baseline remained responsive at all time points, and for PFA-100 Epi, 27/44 (61.4%) of subjects were responsive at all time points tested.

Persistent Poor Responders

A total of 26 subjects were identified as Persistent PRs by any of the three assays, but only 6 of those satisfied criteria for two assays, and only 1 subject satisfied criteria for Persistent PR for all three assays. Figure II depicts the overlap in patients identified as Persistent PR by the various assays. Although few patients were persistently poor responders, the majority of subjects (91.25%) had at least a single poor response to one of the assays at one point during the study when the various tests were pooled, indicating that a patient’s platelet responsiveness to ASA may be a dynamic one with a small number of patients who are persistently poor responders, or that patient non-compliance is a significant issue.

The agreement between the assays was assessed through raw percentage agreement and though the calculation of the kappa statistic. Results are summarized in Table IV with kappa statistics for each comparison and p-values for significance listed in each box. The AA aggregation and PFA-100 Epi assays showed the highest raw agreement at 69%, and the highest kappa statistic value (K=0.3223; 95% confidence interval [CI] 0.152–0.493, P=0.0001). However, this is only “fair” agreement according to published scales.17 The comparison between ADP-induced aggregation and PFA-100 Epi had a raw agreement of 36.6% and non-significant kappa statistic (K=0.0044). Similarly, the comparison between ADP- and AA-induced platelet aggregation had a raw agreement of 46.3% and very slight agreement by kappa statistic (K=0.1161).

Table IV.

Kappa statistic for comparisons between assays

Arachidonic Acid Aggregometry
K = 0.1161 (–0.004–0.236), P=0.02946.3% positive ADP Aggregometry
K = 0.3223 (–0.152–0.493), P=0.000169% positive K = 0.0044 ( 0.151 0.160), P=0.48 36.6% positive PFA-100 Epinephrine
Open in a new tab

Discussion

In our analysis of platelet function assays in PAD patients on ASA, there are several major findings: First, 5–27.5% of PAD patients may be classified as Persistent PRs to ASA depending on the type of platelet assay used. Second, though some patients could be defined as poor responders by single testing, they could be found to be responsive through subsequent testing or through using a different platelet assay. This suggests that either platelet responsiveness to ASA is dynamic with several activation pathways subject to change in response to patient’s conditions and treatments, or that compliance with therapy is a major issue in determining ASA responsiveness. Third, the classification agreement and correlation between these three widely used platelet assays is poor in terms of identifying patients as being poor responders to ASA. In total, these data call into question the usefulness of a single test at a single time to determine a patient’s ASA responsiveness, and the utility in using these assays to identify subjects at higher risk for future vascular events.

The prevalence of poor response to ASA in our study population is higher than has been seen in primary prevention populations,18 but is similar to that seen in prior PAD studies19 and stable cardiovascular disease patients.20 However, most studies have used a single measurement of platelet function to determine a subject’s responsiveness. There is limited data on the agreement of different platelet function assays over time. In survivors of acute myocardial infarctions, Andersen and colleagues randomized 202 patients to high- or low-dose aspirin and coumadin combinations and assessed platelet function by PFA-100. They found some stability in the PFA-100 measure over time with only 10% of subjects changing their responsiveness category.21 In contrast, patients with a history of stroke were assessed by Helgason et al., where it was found that 33% of subjects who initially were aspirin responders, lost some of their responsiveness at follow-up platelet testing.3 Similarly, Harrison and colleagues found in stroke patients, that aspirin non-response at 1-year follow-up was variable based on light transmission aggregometry, PFA-100, or the VerifyNow ASA assay.22 These studies did not include large numbers of subjects with lower extremity peripheral arterial disease. To our knowledge, only one other study of PAD patients used repeated platelet function testing during long-term follow-up.19 In their study, Linneman et al reported poor response to ASA occurred in 4.1% of those patients by aggregometry, but it did not persist in any at both baseline and follow-up. While they assessed the correlation within a given assay at baseline and follow-up, the classification agreement between different assays was not described. We have found that there are a small number of patients who have persistently abnormal values over 6–12-month follow-up. Further, only 60–90% (28.8–74.4% of the overall study population) of those originally found to be ASA responders remained so, again emphasizing that ASA response is a changing variable, but the cause for the variation is not fully understood. Nevertheless, the considerable variation does call for evaluating the state of platelet inhibition by ASA by multiple tests at multiple time points.

A strong message from this study is that there are individuals who have persistent poor response to ASA. There are several potential mechanisms beyond non-compliance that may account for poor response to ASA therapy as seen in our PAD patients. First, 26% of subjects were on clopidogrel, and while the difference in clopidogrel use was not significantly different between responders and poor responders, the persistent poor responders had the lowest clopidogrel use. However, a recent study of clopidogrel use in aspirin resistant subjects by Fontana,23 demonstrated no effect of clopidogrel on aspirin resistance or pseudoresistance. Thus, while our data are in limited numbers, they suggest that PAD subjects on dual antiplatelet therapy achieve greater inhibition as measured by platelet function assays. Second, poor response to ASA may result from either the failure to inhibit TXA2 production through the COX-1 pathway, or persistence of platelet aggregation due to COX-1 independent pathways. ASA may fail to completely inhibit COX-1 mediated TXA2 if the platelets are in a hyperactive state24 or are reticulated,8 both of which are associated with PAD.9, 25, 26 We have previously shown that reticulated platelets have incomplete COX-1 inhibition in response to ASA therapy.27 Finally, a persistent poor response to ASA could also be caused by genetic factors, which have been implicated in studies focusing on known SNPs of selected molecules.28–30 However, these studies may not be sufficient to distinguish between genetic and acquired factors associated with ASA response because they often focused on individuals with high risk for cardiovascular diseases. An unbiased genomic study on young subjects without confounding diseases may be needed to identity genetic factors associated with ASA response.

In our current analysis, we have established that these assays measure different aspects of platelet activity, and also that they identify individuals differently. Aspirin responsiveness is often considered as a dichotomous variable,31 however, the true likelihood is that responsiveness is a continuum and reflects the responsiveness of multiple intersecting platelet aggregation pathways. As a result, we would hypothesize that each of these assays offers clinically relevant data, but that the overall picture of responsiveness needs to be established through multiple testing or at separate timepoints. Specifcally in subjects with peripheral arterial disease, the data is even more limited. Based on our data, we can not say that any single test is clinically beneficial in identifying a group of subjects unless it is tested on more than one occasion, or multiple tests are used.

Limitations

Our study has several limitations. First, we did not examine all assays of ASA response including measurement of urine thromboxane B2 (TXB2). Elevated levels of TXB2 in patients who are on ASA therapy are a marker of poor response, and have been associated with increased cardiovascular events.4 Second, the ELIMIT is still ongoing, thus the impact of the blinded lipid treatment assignment on follow-up platelet assay results is not known. While statin therapy reduces platelet aggregation,32–34 all subjects were on statin therapies and this is unlikely to account for our findings. Both ezetimibe35 and niacin36 are associated with improved platelet aggregation, but the effects of blinded treatments cannot yet be ascertained. Of note, a recent study demonstrated minimal changes in aggregation when ezetimibe was added to simvastatin, but the effect of niacin when combined with statin therapy has not yet been evaluated. Third, we do not have information on the type of ASA taken. The use of enteric-coated ASA has been associated with more frequent finding of poor-response to ASA in both healthy people37 and in subjects with stable cardiovascular disease.38 Additionally, we do not have data on proton-pump inhibitor medication use, which has recently been associated with reduced aspirin effect,39 but as the proportion of our population that is classified as persistent poor responders is similar to prior studies, this is unlikely to have exerted a large effect. Last, patient non-compliance may still exert a significant effect on these assays. Although these patients are enrolled in a clinical trial, we did not directly observe the ASA dose timing or type of ASA taken; however, this is a problem with many clinical studies of ASA treatment. We minimized this effect through the use of very restrictive criteria for poor response.

Conclusions

Between 5–27.5% of the symptomatic PAD population is persistently non-responsive to ASA therapy when measured by AA aggregometry, ADP aggregometry, or PFA-100 Epi. Individual platelet function assays do not correlate well, highlighting the difficulty in developing a standard laboratory assessment for platelet response to ASA. The clinical implications of these data suggest potential benefit in using different assay types, or using repeated assays over time to evaluate a patient’s antiplatelet response to aspirin. Further, our data serve as caution to defining ASA response on the basis of a single assay.

Acknowledgments

We acknowledge the efforts of the ELIMIT Investigators for their contributions to this manuscript in both development and review. Eric Y. Yang, MD; Anirudh Kumar, BS; Angela Bergeron, BS; Husam Athanemh, MD; Peter Lin, MD; Changyi Chen, MD; and Carlos Bechara, MD.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Smith SJ, Allen J, Blair S, Bonow R, Brass L, Fonarow G, Grundy S, Hiratzka L, Jones D, Krumholz H, Mosca L, Pasternak R, Pearson T, Pfeffer M, Taubert K. AHA/ACC guidelines for secondary prevention for patients with coronary and other atherosclerotic vascular disease: 2006 update: endorsed by the National Heart, Lung, and Blood Institute. Circulation. 2006;113(19):2363–2372. doi: 10.1161/CIRCULATIONAHA.106.174516. [DOI] [PubMed] [Google Scholar]
  • 2.Vejar M, Fragasso G, Hackett D, Lipkin D, Maseri A, Born G, Ciabattoni G, Patrono C. Dissociation of platelet activation and spontaneous myocardial ischemia in unstable angina. Thromb Haemost. 1990;63(2):163–168. [PubMed] [Google Scholar]
  • 3.Helgason C, Bolin K, Hoff J, Winkler S, Mangat A, Tortorice K, Brace L. Development of aspirin resistance in persons with previous ischemic stroke. Stroke. 1994;25(12):2331–2336. doi: 10.1161/01.str.25.12.2331. [DOI] [PubMed] [Google Scholar]
  • 4.Eikelboom JW, Hirsh J, Weitz JI, Johnston M, Yi Q, Yusuf S. Aspirin-resistant thromboxane biosynthesis and the risk of myocardial infarction, stroke, or cardiovascular death in patients at high risk for cardiovascular events. Circulation. 2002;105(14):1650–1655. doi: 10.1161/01.cir.0000013777.21160.07. [DOI] [PubMed] [Google Scholar]
  • 5.Gum P, Kottke-Marchant K, Welsh P, White J, Topol E. A prospective, blinded determination of the natural history of aspirin resistance among stable patients with cardiovascular disease. J Am Coll Cardiol. 2003;41(6):961–965. doi: 10.1016/s0735-1097(02)03014-0. [DOI] [PubMed] [Google Scholar]
  • 6.Mueller MR, Salat A, Stangl P, Murabito M, Pulaki S, Boehm D, Koppensteiner R, Ergun E, Mittlboeck M, Schreiner W, Losert U, Wolner E. Variable platelet response to low-dose ASA and the risk of limb deterioration in patients submitted to peripheral arterial angioplasty. Thromb Haemost. 1997;78(3):1003–1007. [PubMed] [Google Scholar]
  • 7.Ziegler S, Maca T, Alt E, Speiser W, Schneider B, Minar E. Monitoring of antiplatelet therapy with the PFA-100 in peripheral angioplasty patients. Platelets. 2002;13(8):493–497. doi: 10.1080/0953710021000057866. [DOI] [PubMed] [Google Scholar]
  • 8.Guthikonda S, Alviar C, Vaduganathan M, Arikan M, Tellez A, DeLao T, Granada J, Dong J, Kleiman N, Lev E. Role of reticulated platelets and platelet size heterogeneity on platelet activity after dual antiplatelet therapy with aspirin and clopidogrel in patients with stable coronary artery disease. J Am Coll Cardiol. 2008;52(9):743–749. doi: 10.1016/j.jacc.2008.05.031. [DOI] [PubMed] [Google Scholar]
  • 9.Nambi V, Kimball K, Bray P, Bergeron A, Johnson S, Morrisett J, Chen C, Lin P, Lumsden A, Ballantyne C, Dong J. Differences in responses of platelets to fluid shear stress in patients with peripheral artery disease (PAD) and coronary artery disease (CAD) Platelets. 2009;20(3):199–205. doi: 10.1080/09537100902780643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.McGlasson D, Fritsma G. Comparison of four laboratory methods to assess aspirin sensitivity. Blood Coagul Fibrinolysis. 2008;19(2):120–123. doi: 10.1097/MBC.0b013e3282f40dc2. [DOI] [PubMed] [Google Scholar]
  • 11.Faraday N, Becker D, Yanek L, Herrera-Galeano J, Segal J, Moy T, Bray P, Becker L. Relation between atherosclerosis risk factors and aspirin resistance in a primary prevention population. Am J Cardiol. 2006;98(6):774–779. doi: 10.1016/j.amjcard.2006.04.015. [DOI] [PubMed] [Google Scholar]
  • 12.Lumsden A, Rice T, Chen C, Zhou W, Lin P, Bray P, Morrisett J, Nambi V, Ballantyne C. Peripheral arterial occlusive disease: magnetic resonance imaging and the role of aggressive medical management. World J Surg. 2007;31(4):695–704. doi: 10.1007/s00268-006-0732-y. [DOI] [PubMed] [Google Scholar]
  • 13.Salzman E. Interrelation of prostaglandin endoperoxide (prostaglandin G2) and cyclic 3′,5′-adenosine monophosphate in human blood platelets. Biochim Biophys Acta. 1977;499(1):48–60. doi: 10.1016/0304-4165(77)90227-6. [DOI] [PubMed] [Google Scholar]
  • 14.Rao A, Koike K, Willis J, Daniel J, Beckett C, Hassel B, Day H, Smith J, Holmsen H. Platelet secretion defect associated with impaired liberation of arachidonic acid and normal myosin light chain phosphorylation. Blood. 1984;64(4):914–921. [PubMed] [Google Scholar]
  • 15.Crescente M, Di Castelnuovo A, Iacoviello L, Vermylen J, Cerletti C, de Gaetano G. Response variability to aspirin as assessed by the platelet function analyzer (PFA)-100. A systematic review. Thromb Haemost. 2008;99(1):14–26. doi: 10.1160/TH07-08-0530. [DOI] [PubMed] [Google Scholar]
  • 16.Gum P, Kottke-Marchant K, Poggio E, Gurm H, Welsh P, Brooks L, Sapp S, Topol E. Profile and prevalence of aspirin resistance in patients with cardiovascular disease. Am J Cardiol. 2001;88(3):230–235. doi: 10.1016/s0002-9149(01)01631-9. [DOI] [PubMed] [Google Scholar]
  • 17.Landis J, Koch G. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159–174. [PubMed] [Google Scholar]
  • 18.Faraday N, Becker DM, Yanek LR, Herrera-Galeano JE, Segal JB, Moy TF, Bray PF, Becker LC. Relation between atherosclerosis risk factors and aspirin resistance in a primary prevention population. Am J Cardiol. 2006;98(6):774–779. doi: 10.1016/j.amjcard.2006.04.015. [DOI] [PubMed] [Google Scholar]
  • 19.Linnemann B, Prochnow S, Mani H, Schwonberg J, Lindhoff-Last E. Variability of non-response to aspirin in patients with peripheral arterial occlusive disease during long-term follow-up. Ann Hematol. 2009;88(10):979–988. doi: 10.1007/s00277-009-0708-8. [DOI] [PubMed] [Google Scholar]
  • 20.Gum PA, Kottke-Marchant K, Welsh PA, White J, Topol EJ. A prospective, blinded determination of the natural history of aspirin resistance among stable patients with cardiovascular disease. J Am Coll Cardiol. 2003;41(6):961–965. doi: 10.1016/s0735-1097(02)03014-0. [DOI] [PubMed] [Google Scholar]
  • 21.Andersen K, Hurlen M, Arnesen H, Seljeflot I. Aspirin non-responsiveness as measured by PFA-100 in patients with coronary artery disease. Thromb Res. 2002;108(1):37–42. doi: 10.1016/s0049-3848(02)00405-x. [DOI] [PubMed] [Google Scholar]
  • 22.Harrison P, Segal H, Silver L, Syed A, Cuthbertson F, Rothwell P. Lack of reproducibility of assessment of aspirin responsiveness by optical aggregometry and two platelet function tests. Platelets. 2008;19(2):119–124. doi: 10.1080/09537100701771736. [DOI] [PubMed] [Google Scholar]
  • 23.Fontana P, Nolli S, Reber G, de Moerloose P. Biological effects of aspirin and clopidogrel in a randomized cross-over study in 96 healthy volunteers. J Thromb Haemost. 2006;4(4):813–819. doi: 10.1111/j.1538-7836.2006.01867.x. [DOI] [PubMed] [Google Scholar]
  • 24.Guthikonda S, Mangalpally K, Vaduganathan M, Patel R, Delao T, Bergeron AL, Dong JF, Lev EI, Kleiman NS. Increased platelet sensitivity among individuals with aspirin resistance - platelet aggregation to submaximal concentration of arachidonic acid predicts response to antiplatelet therapy. Thromb Haemost. 2008;100(1):83–89. doi: 10.1160/TH07-10-0590. [DOI] [PubMed] [Google Scholar]
  • 25.Robless P, Okonko D, Lintott P, Mansfield A, Mikhailidis D, Stansby G. Increased platelet aggregation and activation in peripheral arterial disease. Eur J Vasc Endovasc Surg. 2003;25(1):16–22. doi: 10.1053/ejvs.2002.1794. [DOI] [PubMed] [Google Scholar]
  • 26.Esposito C, Popescu W, Rinder H, Schwartz J, Smith B, Rinder C. Increased leukocyte-platelet adhesion in patients with graft occlusion after peripheral vascular surgery. Thromb Haemost. 2003;90(6):1128–1134. doi: 10.1160/TH03-04-0226. [DOI] [PubMed] [Google Scholar]
  • 27.Guthikonda S, Lev E, Patel R, DeLao T, Bergeron A, Dong J, Kleiman N. Reticulated platelets and uninhibited COX-1 and COX-2 decrease the antiplatelet effects of aspirin. J Thromb Haemost. 2007;5(3):490–496. doi: 10.1111/j.1538-7836.2007.02387.x. [DOI] [PubMed] [Google Scholar]
  • 28.Lepäntalo A, Mikkelsson J, Reséndiz J, Viiri L, Backman J, Kankuri E, Karhunen P, Lassila R. Polymorphisms of COX-1 and GPVI associate with the antiplatelet effect of aspirin in coronary artery disease patients. Thromb Haemost. 2006;95(2):253–259. doi: 10.1160/TH05-07-0516. [DOI] [PubMed] [Google Scholar]
  • 29.Jefferson B, Foster J, McCarthy J, Ginsburg G, Parker A, Kottke-Marchant K, Topol E. Aspirin resistance and a single gene. Am J Cardiol. 2005;95(6):805–808. doi: 10.1016/j.amjcard.2004.11.045. [DOI] [PubMed] [Google Scholar]
  • 30.Fontana P, Dupont A, Gandrille S, Bachelot-Loza C, Reny J, Aiach M, Gaussem P. Adenosine diphosphate-induced platelet aggregation is associated with P2Y12 gene sequence variations in healthy subjects. Circulation. 2003;108(8):989–995. doi: 10.1161/01.CIR.0000085073.69189.88. [DOI] [PubMed] [Google Scholar]
  • 31.Bhatt D. Aspirin resistance: more than just a laboratory curiosity. J Am Coll Cardiol. 2004;43(6):1127–1129. doi: 10.1016/j.jacc.2004.01.022. [DOI] [PubMed] [Google Scholar]
  • 32.Santos M, Fuset M, Ruano M, Moscardó A, Valles J. Effect of atorvastatin on platelet thromboxane A(2) synthesis in aspirin-treated patients with acute myocardial infarction. Am J Cardiol. 2009;104(12):1618–1623. doi: 10.1016/j.amjcard.2009.07.039. [DOI] [PubMed] [Google Scholar]
  • 33.Tirnaksiz E, Pamukcu B, Oflaz H, Nisanci Y. Effect of high dose statin therapy on platelet function; statins reduce aspirin-resistant platelet aggregation in patients with coronary heart disease. J Thromb Thrombolysis. 2009;27(1):24–28. doi: 10.1007/s11239-007-0154-1. [DOI] [PubMed] [Google Scholar]
  • 34.Puccetti L, Pasqui A, Pastorelli M, Bova G, Cercignani M, Palazzuoli A, Angori P, Auteri A, Bruni F. Time-dependent effect of statins on platelet function in hypercholesterolaemia. Eur J Clin Invest. 2002;32(12):901–908. doi: 10.1046/j.1365-2362.2002.01086.x. [DOI] [PubMed] [Google Scholar]
  • 35.Hussein O, Minasian L, Itzkovich Y, Shestatski K, Solomon L, Zidan J. Ezetimibe’s effect on platelet aggregation and LDL tendency to peroxidation in hypercholesterolaemia as monotherapy or in addition to simvastatin. Br J Clin Pharmacol. 2008;65(5):637–645. doi: 10.1111/j.1365-2125.2007.03080.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Serebruany V, Malinin A, Aradi D, Kuliczkowski W, Norgard N, Boden W. The in vitro effects of niacin on platelet biomarkers in human volunteers. Thromb Haemost. 2010;104(2) doi: 10.1160/TH10-01-0015. [DOI] [PubMed] [Google Scholar]
  • 37.Cox D, Maree A, Dooley M, Conroy R, Byrne M, Fitzgerald D. Effect of enteric coating on antiplatelet activity of low-dose aspirin in healthy volunteers. Stroke. 2006;37(8):2153–2158. doi: 10.1161/01.STR.0000231683.43347.ec. [DOI] [PubMed] [Google Scholar]
  • 38.Maree A, Curtin R, Dooley M, Conroy R, Crean P, Cox D, Fitzgerald D. Platelet response to low-dose enteric-coated aspirin in patients with stable cardiovascular disease. J Am Coll Cardiol. 2005;46(7):1258–1263. doi: 10.1016/j.jacc.2005.06.058. [DOI] [PubMed] [Google Scholar]
  • 39.Würtz M, Grove E, Kristensen S, Hvas A. The antiplatelet effect of aspirin is reduced by proton pump inhibitors in patients with coronary artery disease. Heart. 2010;96(5):368–371. doi: 10.1136/hrt.2009.181107. [DOI] [PubMed] [Google Scholar]

RESOURCES