Mechanism of Variability in the Response to Low Dose Aspirin

Treatment failure due to variability in the pharmacologic response to drugs has more commonly been attributed to non-genetic factors than to prototypic pharmacogenetic mechanisms.1 Treatment failure on low-dose aspirin taken to prevent arterial thrombotic events, including myocardial infarction and stroke, is no exception. A closer look at aspirin’s human pharmacology reveals many potential sources of variability that are rooted in platelet biology, from aspirin insensitive mechanisms of activation to accelerated turnover. Although platelet inhibition with low dose aspirin is effective in reducing cardiovascular events in patients at elevated risk, myocardial infarction and stroke can still occur on treatment. However, genetic variation in metabolism and target genes alone are insufficient to explain the breadth of variability in the response to aspirin.2 Aspirin inhibits platelets by depressing their ability to form thromboxane (Tx) A2, a lipid mediator derived from arachidonic acid. TxA2 is one of several agonists that accelerate clot formation upon initial platelet activation. Adhesion of platelets to injured vascular surfaces—primarily through interactions with subendothelial collagen and von Willebrand factor—trigger the release of mediators that act as feed forward signals to augment the recruitment of additional platelets. The TxA2 biosynthetic response pathway is a major such amplification signal in the growing clot. Aspirin treatment failure may result from a more dominant role of pathways other than TxA2,3 such as the adenosine diphosphate (ADP) pathway, which is targeted by the purinergic receptor P2Y12 inhibitors clopidogrel, prasugrel, and ticagrelor. This is the rationale for combined inhibition of both pathways in patients with increased risk of vascular occlusion. Upon platelet activation, cytosolic phospholipase A2 translocates to the membrane where it cleaves arachidonic acid from membrane glycerophospholipids. Free arachidonic acid is sequentially metabolized by the membrane bound enzymes cyclooxygenase (COX)-1—aspirin’s drug target—and Tx synthase. COX-1 forms an unstable endoperoxide product, PGH2, in a two-step redox reaction, which Tx synthase isomerizes to TxA2. In contrast to other platelet mediators, which are stored in platelet vesicles, TxA2 is released as it is produced. Consistent with rapid biosynthesis as a feed-forward signal, the capacity of platelets to form TxA2 far exceeds the concentrations needed for platelet activation which is reflected by a steep dose response curve. COX-1 activity needs to be inhibited almost completely (i.e., > 95%) to suppress TxA2 below levels that effectively amplify platelet activation.4 This can make reliable therapeutic inhibition throughout the entire dosing interval challenging. For example, increased formation of peroxides and glycation products in patients with diabetes, which are thought to render COX-1 somewhat less susceptible to inhibition by aspirin, may lead to considerable variability in the response to aspirin.5, 6 The fluid dynamic properties of red blood cells push circulating platelets toward the vessel wall, an effect known as platelet margination. This reduces their velocity and facilitates recruitment to the site of injury. TxA2 released from the growing clot contributes to the activation of platelets in the vicinity of the injured vessel wall. The platelet TxA2 receptor signals predominantly through Gq and G12/G13. The G12/G13-mediated signaling facilitates platelet shape change and Gq engages protein kinase C-dependent pathways, which participate in platelet aggregation. Conditions associated with more reactive platelets may result in variability in the response to aspirin presumably through interactions with these intracellular signaling pathways.7 TxA2 acts as an autacoid local signal. It is rapidly diluted in the streaming blood and hydrolyzed to the stable and biologically inactive compound TxB2. The TxB2 is eliminated from the circulation through liver metabolism, resulting in urinary excretion of metabolites including the two most abundant compounds, 11-dehydro TxB2 and 2, 3-dinor TxB2. Patients with atherothrombotic events have elevated urinary Tx metabolite levels which are suppressed by aspirin. Aspirin is also hydrolyzed in the circulation. Its short half-life (~ 15–20 minutes in plasma) is due to the activity of esterases, which hydrolyze aspirin’s active acetyl moiety to form salicylic acid. The hydrolytic effectiveness of these esterases varies widely—primarily from non-genetic variation—affecting aspirin bioavailability in blood and consequently platelet aggregation.8 Enzymatic function of COX-1 requires the formation of protein dimers. One of the COX-1 monomers contains a heme group, which acts as an electron source in catalyzing the redox reaction that forms PGH2. The other monomer regulates the reaction kinetics allosterically. Aspirin acetylates serine-529 in the substrate binding channel of the catalytically active COX-1 monomer. This creates a hydrophilic residue, which interferes with the positioning of arachidonic acid in the catalytic center, inactivating the enzyme irreversibly. Nonsteroidal anti-inflammatory drugs, which are reversible COX inhibitors, can directly or allosterically interact with aspirin’s acetylation of serine-529 in COX-1 and thus cause variability in platelet inhibition.9 As platelets lack a nucleus, they have only minimal capacity to form new proteins from cytosolic RNA. Thus, platelets exposed to aspirin are permanently inhibited and, despite aspirin’s short half-life, COX-1 acetylation and the consequent inhibition of platelet aggregation remains close to maximal for about 2 days after a single analgesic dose (i.e., 325 mg and above).9 Low doses of aspirin that are insufficient to acetylate the entire platelet COX-1 pool at once, eventually result in complete platelet inhibition if they are given repetitively as long as the fraction of platelets that is inactivated by each dose exceeds the release of new platelets during the dosing interval. Unsurprisingly, extending the dosing interval, for example, by skipping more than an occasional dose, can result in a reduction in platelet inhibition, but so can accelerated release of platelets from the bone marrow. In this issue of Clinical Pharmacology & Therapeutics, Tosetto et al.10 investigate the variability in the pharmacological response to low-dose aspirin in patients with essential thrombocythemia (ET). ET is a myeloproliferative neoplasm characterized by excessive, clonal platelet production. The majority of patients with ET are prescribed low-dose aspirin to reduce the incidence of platelet-mediated thrombotic and microvascular ischemic manifestations. Patients at elevated risk for thrombotic events typically also receive cytoreductive therapy, most often hydroxyurea or anagrelide, to lower platelet counts. Treatment with low-dose aspirin is complicated by the accelerated turnover of platelets in patients with ET. Thus, many patients show evidence of premature recovery from irreversible COX-1 inhibition by aspirin during the 24-hour dosing interval, because of the release of new platelets with intact COX-1. These patients are likely to be undertreated with a single dose of aspirin per day. Tosetto et al. conducted a prespecified substudy of the Aspirin Regimens in Essential Thrombocythemia (ARES) trial to identify factors associated with accelerated recovery from TxA2 suppression leading to insufficient platelet inhibition by aspirin. ARES is a dose-finding study in which patients with ET were randomized to a 2-week period of aspirin 100 mg once, twice, or 3-times daily. The capacity of whole blood to generate serum TxB2 (sTxB2), the stable hydrolysis product of TxA2, during clotting under standardized conditions was measured as an index of residual COX-1 enzymatic activity on aspirin. A total of 218 evaluable patients were included in the analysis. Only about a quarter of patients on daily 100 mg aspirin were considered to have adequately inhibited platelets as indicated by their minimal residual sTxB2 concentrations. These patients had lower platelet counts and were more likely to receive cytoreductive therapy than those who presented with insufficiently suppressed TxB2 values. Interestingly, failure to suppress sTxB2 was not only associated with an increase in platelet mass, but was also characterized by platelets with higher enzymatic activity of COX-1. This supported a contribution to the insufficient aspirin response of both accelerated turnover and more reactive platelets. Importantly, administration of 2 or 3 doses of 100 mg aspirin per day were highly effective in suppressing platelet COX-1 function and in reducing variability in drug response. Can platelet count help guide the decision of which patients with ET should receive aspirin more than once daily? The authors conclude that the discriminatory power of platelet count is insufficient to reliably predict nonresponders to once per day aspirin dosing. They argue that although sTxB2 would ideally be measured to distinguish responders from nonresponders, this currently is impractical in clinical routine. Instead, they suggest treating all patients with aspirin 100 mg twice daily, although this might be associated with a small increase in the risk of bleeding complications in patients who would have responded adequately to once daily dosing. Whereas further investigation of long-term outcomes of an optimized aspirin therapy in ET is warranted, this research highlights the need for a deeper understanding of both genetic and non-genetic sources of variation in drug response as we move toward a more individualized drug therapy. Open access funding enabled and organized by Projekt DEAL. No funding was received for this work. S.F. declares no competing interests for this work. T.G. serves as an associate editor for Circulation Genomic and Precision Medicine.