Aspirin Resistance: An Evaluation of Current Evidence and Measurement Methods

Christopher P. Martin, Pharm.D.; Robert L. Talbert, Pharm.D., FCCP


Pharmacotherapy. 2005;25(7):942-953. 

In This Article

Proposed Mechanisms of Aspirin Resistance

The exact mechanism of aspirin resistance remains elusive, but several theories have been proposed and investigated. Factors that have been studied in the greatest detail are increased platelet sensitivity to collagen, increased COX-2 activity, and the platelet alloantigen 2 (PLA2) polymorphism of platelet glycoprotein IIIa. Other mechanisms that have been proposed are poor patient compliance,[8] poor gastrointestinal aspirin absorption,[9] increased isoprostane activity,[10,11] and a COX-1 polymorphism.[12,13] However, absorption and compliance present a bioavailability issue and should be distinguished from purported issues of pharmacodynamic resistance.

Collagen is a potent platelet agonist and is often the first to initiate the process of platelet aggregation as platelets bind to exposed intimal collagen in an area of vascular injury (Figure 2). In theory, certain individuals may produce platelets that are oversensitive to collagen or other platelet agonists enough to supersede any inhibitory effect conferred by the aspirin. This possibility was explored in a study involving eight healthy Japanese subjects who were stratified into two groups based on bleeding time measurements obtained before and after a single dose of aspirin 324 mg.[14] Three of the eight subjects were considered aspirin resistant as defined by a postdose bleeding time less than two standard deviations above the predose bleeding time. When subjected to platelet aggregometry, a significantly greater concentration of collagen was required to produce 50% of maximal aggregation in aspirin-sensitive platelets (0.91 µg/ml, 95% confidence interval [CI] 0.73-1.14) than in aspirin-resistant platelets (0.48 µg/ml, 95% CI 0.38-0.60). This suggests that aspirin-resistant platelets were more sensitive to collagen than aspirin-sensitive platelets. However, considering the variability of bleeding time as a platelet activity measure and the small sample size, a relationship between platelet collagen sensitivity and aspirin resistance cannot be proven beyond a chance association in this trial.

Platelet aggregation scheme. Numerous agonists activate the platelet; cytoskeletal contraction changes the activated platelet from a smooth discoid to an amorphous shape. Dense granules and a-granules containing platelet agonists are released, further activating and recruiting local circulating platelets. Platelet plug formation occurs by platelet-to-platelet cross-linking with fibrinogen bound to glycoprotein IIb-IIIa on the platelet surface. TXA2 = thromboxane A2; ADP = adenosine 5'-diphosphate.

Altered expression of COX-2, the second cyclooxygenase isoform, has been implicated as a possible source of aspirin-insensitive TXA2 production. Unlike the constitutive expression of COX-1, COX-2 expression is largely inducible. Cyclooxygenase-2 is upregulated 10-20-fold by inflammatory stimuli in most tissues, including monocytes, macrophages, and vascular epithelium, and more recent evidence suggests platelets as well.[7,15] A 90% structural homology is shared between COX-2 and COX-1; however, the 10% structural difference renders COX-2 relatively insensitive to aspirin. A dose of aspirin 650 mg is required to reduce COX-2-mediated eicosanoid production. The effect of aspirin on COX-2 is transient, with a duration of action of 3 hours.

The association of aspirin resistance with COX-2 expression was investigated in 93 patients who underwent coronary artery bypass graft surgery.[16] The patients were asked to discontinue aspirin therapy at least 7 days before the procedure; aspirin 100 mg/day was started after surgery. Western-blot analysis of COX-2 protein expression was conducted, and optical platelet aggregation was measured with the in vitro addition of moderate and high concentrations of aspirin 30 and 100 B5mol, on days 1, 5, and 10 after the procedure.

Compared with baseline, a significant increase in platelet aggregation despite the presence of aspirin was noted on day 5. This increase coincided with a 15-fold increase in COX-2 protein expression. By day 10, platelet aggregation and COX-2 expression returned to baseline levels. These results suggest a relationship between upregulation of COX-2 expression and transient aspirin resistance after coronary artery bypass graft surgery. However, similar studies involving long-term aspirin therapy have not been conducted, and the role of COX-2 in the development of aspirin resistance remains unclear.

Platelets rely on numerous surface glycoprotein receptors to perform their complex function, and many of these receptors are polymorphic. For instance, the PLA2 polymorphism has been identified in glycoprotein IIIa, which constitutes half the fibrinogen receptor glycoprotein IIb-IIIa ( Table 1 ).[17,18,19] The PLA2 polymorphism is a single nucleotide, cytosine (PLA1)-to-thymine (PLA2) polymorphism occurring at position 1565 in exon 2 of the coding region for glycoprotein IIIa (chromosome 17), which results in a leucine-proline exchange.[20] The genotypic frequencies of the polymorphism in North American Caucasians are PLA1/A1 85%, PLA1/A2 13%, and PLA2/A2 2%. Hyperresponsiveness to agonists has been demonstrated in platelets positive for the PLA2 allele in vitro.[21,22]

The hyperaggregability conferred by the PLA2 allele has been attributed to increased surface expression of glycoprotein IIb-IIIa receptors and an increased affinity for fibrinogen.[22,23] In addition, the presence of the PLA2 allele has been associated with a greater risk of coronary events,[17,18] although this is not universally accepted.[24] The effect of the PLA2 polymorphism on aspirin sensitivity was investigated in 40 healthy men; 25 homozygous for PLA1 and 15 positive for the PLA2 allele.[25] Thrombin generation was used as an in vitro measure of platelet aggregation. After both groups were given aspirin 75 mg/day for 1 week, a significantly greater reduction in thrombin generation was seen, compared with baseline, in the PLA1 group than in the PLA2 group (63% vs 28%, p=0.02).

The same investigators studied aspirin response in two groups of healthy men (54 and 26 in the PLA1 and PLA2 groups, respectively) by comparing bleeding time before and 4 hours after a single dose of aspirin 300 mg.[26] Compared with baseline, a significantly greater increase in bleeding time was found in the PLA1 versus the PLA2 group (116.5 vs 37.5 sec, p=0.0002) after aspirin administration.

One study investigated the variable response of the different PLA genotypes to platelet agonists in vitro with and without aspirin.[22] Platelets were drawn from 56 healthy donors with the following genotypes: PLA1/A1 (20 subjects), PLA1/A2 (20), and PLA2/A2 (16). Platelets from the PLA2/A2 subjects showed significantly greater aggregation (45.3%) in response to an epinephrine agonist (in the absence of aspirin) than either PLA1/A1 (22.6%) or PLA1/A2 (16.4%) subjects (p=0.03). However, response to adenosine 5'-diphosphate (ADP) and thrombin receptor-activating peptide (TRAP) agonists was the same in all three groups. When platelet aggregation was performed in the presence of 5 µM aspirin, the PLA2/A2 group was relatively aspirin resistant and had significantly greater aggregation than PLA1/A1 subjects.

Paradoxically, aggregation was significantly less in the PLA1/A2 group than in the PLA1/A1 group at aspirin concentrations of 2.5 and 5.0 µM, suggesting that PLA heterozygotes are more sensitive to aspirin than homozygotes. By comparison, a plasma concentration of 2.5 µM (450 µg/L) correlates with enteric-coated aspirin 100-160 mg/day; a dose of regular-release aspirin 325 mg produces a maximum concentration of 27 B5mol (4860 µg/L).[27,28]

These studies indicate a possible role of the PLA2 allele in aspirin resistance based on measures of platelet function, particularly in patients homozygous for PLA2/A2. Thus far, only the PLA2 polymorphism has been associated with aspirin resistance; studies of the C807T (glycoprotein IIa) and Kozak (glycoprotein Ibα) polymorphisms have been negative ( Table 1 ).[17,18,19] Further study is needed to fully characterize the effect of the PLA2 polymorphism on the antithrombotic effect of aspirin and, more important, to determine whether this polymorphism can influence clinical outcomes.