Identifying Determinants to Tailor Aspirin Therapy
Identifying Determinants to Tailor Aspirin Therapy
In the early 1970s, the antiplatelet effect of aspirin was initially described by generic assays available at the time, exploring primary hemostasis or platelet function, such as the Ivy bleeding time and the ADP-induced optical aggregation, respectively. By slightly prolonging the bleeding time and reducing ADP-induced platelet aggregation, aspirin was classified as a 'weak' antiplatelet drug, producing only a 'minor hemostatic defect'. A few years later, Smith and Willis showed that aspirin blocked prostaglandin generation from human platelets, and Hamberg et al. identified TXA2 as the biologically active compound synthesized from arachidonic acid (AA) in platelets and inhibited by aspirin. In the mid-1970s, Majerus et al., by using radiolabeled H-acetyl aspirin, showed that this drug rapidly and irreversibly acetylates a 85-kDa platelet protein fraction, containing an AA-binding, active site, corresponding to the COX enzymes. In the mid-1990s, Loll et al. resolved the crystal structure of aspirin-bound COX-1. The best characterized PD of low-dose aspirin as an antithrombotic drug relies upon its capacity to permanently acetylating a serine residue and blocking the active site of COX enzymes in platelets.
The human pharmacology of low dose (up to ~160 mg/day), oral aspirin as an antiplatelet agent in humans, was described by different groups in the 1980s. Patrono et al. established a biochemical, ex vivo assay, directly reflecting the activity of COX enzymes from peripheral platelets, rather than the degree of protein acetylation, easing the characterization of aspirin pharmacology. Until then, the methods used in humans ex vivo, were quite lengthy, requiring mixing H-acetyl aspirin into the blood of aspirin-treated subjects and subsequent protein extractions or based on in vitro optical platelet aggregation, followed by prostanoid extraction and measurement. Moreover, H-acetyl aspirin explored the degree of protein acetylation, rather than the residual COX-dependent enzymatic capacity of generating TXA2. Aside from practical advantages (minimal blood volume and preanalytical handling), the principle of the assay relies on a physiological reaction: the maximal generation of endogenous thrombin during whole blood clotting at 37°C, strongly triggers phospholipases and AA release from platelet membranes, thus, feeding COX activity and TXA2 synthesis (Figure 3). TXA2 is then rapidly (32 s) and nonenzymatically hydrolyzed in aqueous milieu to TXB2, which is stable, biologically inactive and measurable in serum without further purification by standard immunometric methods. Thus, this assay reflects the maximal biosynthetic capacity of low-dose aspirin target – that is, platelet COX enzymes in a physiological environment. Approximately 20 years after the description of this method, the EMA acknowledged serum TXB2 as the only surrogate assay required for the authorization of new modified-release aspirin formulations. Importantly, serum TXB2 is nonlinearly correlated with other functional platelet assays (Figure 4). In other words, platelet TXA2 biosynthetic capacity should be almost completely suppressed, down to very low absolute values (~1 ng/ml in healthy subjects), in order to fully block platelet function. On this basis, it is somewhat surprising that only a minority of papers dealing with aspirin resistance have used the only available method directly reflecting the enzymatic activity of aspirin molecular targets – that is, serum TXB2 (Figure 2). The majority of the studies utilized assays assessing platelet function with specific (AA) or nonspecific agonists (collagen, ADP and epinephrine).
(Enlarge Image)
Figure 3.
Pharmacokinetics and pharmacodynamics of aspirin. Thrombin, generated in vivo or ex vivo by several chemical or physical stimuli, activates its protease activated receptors increasing intraplatelet calcium, which triggers PLA2-dependent cleavage of arachidonic acid from plasma membranes. Arachidonic acid is the enzymatic substrate of COX-1 and -2. COX-1-dependent arachidonic acid path in platelets generates mainly TXA2, which amplifies platelet activation by binding to their platelet receptors. TXA2 both in vivo or ex vivo is nonenzymatically hydrolyzed to TXB2, which is biologically inactive but stable and can be measured in ex vivo assays, or undergoes further hepatic enzymatic biotransformation in vivo. Aspirin, absorbed in the stomach and small intestine, exerts its pharmacodynamic effect, that is, the acetylation of a Serine (Ser) 529 residue of COX-1, in the portal blood and is biotransformed to inactive salicylic acid by intestine, plasma and liver esterases. On average, its systemic bioavailability is approximately 50% of the administered dose. Once in systemic circulation, aspirin reaches bone marrow MKs and inhibits COX-1 and -2 of MKs and developing platelets.
MKs: Megakaryocytes.
Modified with permission from [173].
(Enlarge Image)
Figure 4.
Non-linear relationship between serum TXB2 and assays measuring ex vivo or in vivo platelet function. The figure represents the non-linear relationship between individual serum TXB2 measurements and the correspondent ARU (A) or urinary metabolites of TXA2(B) in a group of healthy subjects on or off aspirin (open circles) and in essential thrombocythemia patients on aspirin (closed circles). The best fitting in both data sets, including all measurements, was non-linear; the horizontal dotted bars on the x-axis of both panels identify the normal range of serum TXB2 in the absence of aspirin treatment.
ARU: Aspirin-responsive units.
Data not shown from [6,146].
The main features of once-daily, low-dose aspirin pharmacology in (healthy) humans are: irreversible, cumulative and saturable inhibition of platelet TXA2 generation upon repeated, once-daily dosings, which plateau within 7–10 days; inhibition of platelet's COX activity largely occurring in the portal blood, before liver first-pass; following aspirin withdrawal, a 24–48 h lag is observed before the reappearance of nonacetylated COX protein or TXA2 biosynthesis in circulating platelets (Figure 5); and the full recovery of platelet COX activity after aspirin withdrawal takes approx 7–10 days. Overall, these characteristics reflect the interaction between a short-lived (20 min) drug with two longer-lived cell targets: circulating platelets and megakaryocytes. In fact, the combination of a limited ability of platelets to synthesize new COX protein under normal conditions, hampering the replacement of the acetylated enzyme, together with platelet lifespan, explains the cumulative and saturable effect of once-daily low-dose aspirin, mainly exerted at a presystemic level. In addition, the delay in recovery of peripheral platelets' COX activity upon drug suspension, reported independently of the techniques used (radiolabeled aspirin and serum TXB2), probably reflects the systemic bioavailability of aspirin and the acetylation of COX-1 and -2 in megakaryocytes and possibly, proplatelets. Thus, at least in conditions of normal megakaryopoiesis, during the first 24–48 h after aspirin withdrawal, platelets with acetylated, nonfunctioning COX enzymes likely enter the peripheral circulation from bone marrow sinusoids (Figure 5). Interestingly, this lag interval in the recovery of TXA2 biosynthesis from circulating platelets and the effect on megakaryocytes was observed in different mammalian species, and experimental models. Thus, a fully-effective, once-daily, low-dose aspirin regimen requires a coordinated action of three components: a short-lived drug, a normal lifespan of circulating platelets and normal physiological megakaryocyte/proplatelets activity.
(Enlarge Image)
Figure 5.
Lag in the recovery of serum TXB2 after aspirin withdrawal. Following aspirin withdrawal in aspirin-treated healthy subjects, a delay has been observed in the recovery of serum TXB2. This delay is likely due to acetylation of bone marrow megakaryocytes and aspirin bioavailability.
MKs: Megakaryocytes.
Modified with permission from [6].
In addition to the COX pathway, AA in platelets can be metabolized through the 12-LOX enzyme to 12-HETE. The role of the 12-LOX pathway and 12-HETE generation while COX activity is blocked by aspirin in humans in vivo remains controversial. While a direct, specific inhibition of aspirin on 12-LOX has been consistently ruled out by different groups, indirect effects on 12-LOX activity may derive from the residual salicylate moiety or from negative feedback exerted by the TXA2 receptor. The in vivo relevance of these observations is currently unknown.
Compared with other antiplatelet agents such as thyenopyridines, prodrugs requiring a multistep (CYP450- or esterase-dependent) biotransformation, the PK of aspirin is less complex (Figure 3). In the stomach, the low pH facilitates the absorption by passive diffusion of plain aspirin, largely undissociated, but aspirin is also absorbed in the upper small intestine. Acetylsalicylic acid preparations undergo presystemic metabolism and approximately 50–70% of an oral dose reaches the systemic circulation, while the remaining fraction is inactivated – that is, de-acetylated to salicylic acid – by the carboxylesterases in plasma and the liver (first-pass effect). The hepatic human carboxylesterase-2 isozyme mainly accounts for the first-pass effect and it is also present in the gut where it may contribute to aspirin de-acetylation. In addition to hepatic human carboxylesterase-2, other esterases are present in the peripheral blood including cholinesterases, intraerythrocyte hydrolases and other undefined esterases called 'aspirin esterases'. The residual salicylic acid is virtually unable to inhibit TXA2 generation from human platelets when aspirin is given at low doses.
After a single oral-dose of plain aspirin, plasma peak concentration is reached within 30–40 min and circulating aspirin has a half-life of 15–20 min. Enteric-coated (EC) formulations take a longer time to peak and acetylsalicylic acid is detectable for longer intervals in the peripheral blood, depending on the formulation.
A Triad in the Pharmacology of Low-dose Aspirin
In the early 1970s, the antiplatelet effect of aspirin was initially described by generic assays available at the time, exploring primary hemostasis or platelet function, such as the Ivy bleeding time and the ADP-induced optical aggregation, respectively. By slightly prolonging the bleeding time and reducing ADP-induced platelet aggregation, aspirin was classified as a 'weak' antiplatelet drug, producing only a 'minor hemostatic defect'. A few years later, Smith and Willis showed that aspirin blocked prostaglandin generation from human platelets, and Hamberg et al. identified TXA2 as the biologically active compound synthesized from arachidonic acid (AA) in platelets and inhibited by aspirin. In the mid-1970s, Majerus et al., by using radiolabeled H-acetyl aspirin, showed that this drug rapidly and irreversibly acetylates a 85-kDa platelet protein fraction, containing an AA-binding, active site, corresponding to the COX enzymes. In the mid-1990s, Loll et al. resolved the crystal structure of aspirin-bound COX-1. The best characterized PD of low-dose aspirin as an antithrombotic drug relies upon its capacity to permanently acetylating a serine residue and blocking the active site of COX enzymes in platelets.
The human pharmacology of low dose (up to ~160 mg/day), oral aspirin as an antiplatelet agent in humans, was described by different groups in the 1980s. Patrono et al. established a biochemical, ex vivo assay, directly reflecting the activity of COX enzymes from peripheral platelets, rather than the degree of protein acetylation, easing the characterization of aspirin pharmacology. Until then, the methods used in humans ex vivo, were quite lengthy, requiring mixing H-acetyl aspirin into the blood of aspirin-treated subjects and subsequent protein extractions or based on in vitro optical platelet aggregation, followed by prostanoid extraction and measurement. Moreover, H-acetyl aspirin explored the degree of protein acetylation, rather than the residual COX-dependent enzymatic capacity of generating TXA2. Aside from practical advantages (minimal blood volume and preanalytical handling), the principle of the assay relies on a physiological reaction: the maximal generation of endogenous thrombin during whole blood clotting at 37°C, strongly triggers phospholipases and AA release from platelet membranes, thus, feeding COX activity and TXA2 synthesis (Figure 3). TXA2 is then rapidly (32 s) and nonenzymatically hydrolyzed in aqueous milieu to TXB2, which is stable, biologically inactive and measurable in serum without further purification by standard immunometric methods. Thus, this assay reflects the maximal biosynthetic capacity of low-dose aspirin target – that is, platelet COX enzymes in a physiological environment. Approximately 20 years after the description of this method, the EMA acknowledged serum TXB2 as the only surrogate assay required for the authorization of new modified-release aspirin formulations. Importantly, serum TXB2 is nonlinearly correlated with other functional platelet assays (Figure 4). In other words, platelet TXA2 biosynthetic capacity should be almost completely suppressed, down to very low absolute values (~1 ng/ml in healthy subjects), in order to fully block platelet function. On this basis, it is somewhat surprising that only a minority of papers dealing with aspirin resistance have used the only available method directly reflecting the enzymatic activity of aspirin molecular targets – that is, serum TXB2 (Figure 2). The majority of the studies utilized assays assessing platelet function with specific (AA) or nonspecific agonists (collagen, ADP and epinephrine).
(Enlarge Image)
Figure 3.
Pharmacokinetics and pharmacodynamics of aspirin. Thrombin, generated in vivo or ex vivo by several chemical or physical stimuli, activates its protease activated receptors increasing intraplatelet calcium, which triggers PLA2-dependent cleavage of arachidonic acid from plasma membranes. Arachidonic acid is the enzymatic substrate of COX-1 and -2. COX-1-dependent arachidonic acid path in platelets generates mainly TXA2, which amplifies platelet activation by binding to their platelet receptors. TXA2 both in vivo or ex vivo is nonenzymatically hydrolyzed to TXB2, which is biologically inactive but stable and can be measured in ex vivo assays, or undergoes further hepatic enzymatic biotransformation in vivo. Aspirin, absorbed in the stomach and small intestine, exerts its pharmacodynamic effect, that is, the acetylation of a Serine (Ser) 529 residue of COX-1, in the portal blood and is biotransformed to inactive salicylic acid by intestine, plasma and liver esterases. On average, its systemic bioavailability is approximately 50% of the administered dose. Once in systemic circulation, aspirin reaches bone marrow MKs and inhibits COX-1 and -2 of MKs and developing platelets.
MKs: Megakaryocytes.
Modified with permission from [173].
(Enlarge Image)
Figure 4.
Non-linear relationship between serum TXB2 and assays measuring ex vivo or in vivo platelet function. The figure represents the non-linear relationship between individual serum TXB2 measurements and the correspondent ARU (A) or urinary metabolites of TXA2(B) in a group of healthy subjects on or off aspirin (open circles) and in essential thrombocythemia patients on aspirin (closed circles). The best fitting in both data sets, including all measurements, was non-linear; the horizontal dotted bars on the x-axis of both panels identify the normal range of serum TXB2 in the absence of aspirin treatment.
ARU: Aspirin-responsive units.
Data not shown from [6,146].
The main features of once-daily, low-dose aspirin pharmacology in (healthy) humans are: irreversible, cumulative and saturable inhibition of platelet TXA2 generation upon repeated, once-daily dosings, which plateau within 7–10 days; inhibition of platelet's COX activity largely occurring in the portal blood, before liver first-pass; following aspirin withdrawal, a 24–48 h lag is observed before the reappearance of nonacetylated COX protein or TXA2 biosynthesis in circulating platelets (Figure 5); and the full recovery of platelet COX activity after aspirin withdrawal takes approx 7–10 days. Overall, these characteristics reflect the interaction between a short-lived (20 min) drug with two longer-lived cell targets: circulating platelets and megakaryocytes. In fact, the combination of a limited ability of platelets to synthesize new COX protein under normal conditions, hampering the replacement of the acetylated enzyme, together with platelet lifespan, explains the cumulative and saturable effect of once-daily low-dose aspirin, mainly exerted at a presystemic level. In addition, the delay in recovery of peripheral platelets' COX activity upon drug suspension, reported independently of the techniques used (radiolabeled aspirin and serum TXB2), probably reflects the systemic bioavailability of aspirin and the acetylation of COX-1 and -2 in megakaryocytes and possibly, proplatelets. Thus, at least in conditions of normal megakaryopoiesis, during the first 24–48 h after aspirin withdrawal, platelets with acetylated, nonfunctioning COX enzymes likely enter the peripheral circulation from bone marrow sinusoids (Figure 5). Interestingly, this lag interval in the recovery of TXA2 biosynthesis from circulating platelets and the effect on megakaryocytes was observed in different mammalian species, and experimental models. Thus, a fully-effective, once-daily, low-dose aspirin regimen requires a coordinated action of three components: a short-lived drug, a normal lifespan of circulating platelets and normal physiological megakaryocyte/proplatelets activity.
(Enlarge Image)
Figure 5.
Lag in the recovery of serum TXB2 after aspirin withdrawal. Following aspirin withdrawal in aspirin-treated healthy subjects, a delay has been observed in the recovery of serum TXB2. This delay is likely due to acetylation of bone marrow megakaryocytes and aspirin bioavailability.
MKs: Megakaryocytes.
Modified with permission from [6].
In addition to the COX pathway, AA in platelets can be metabolized through the 12-LOX enzyme to 12-HETE. The role of the 12-LOX pathway and 12-HETE generation while COX activity is blocked by aspirin in humans in vivo remains controversial. While a direct, specific inhibition of aspirin on 12-LOX has been consistently ruled out by different groups, indirect effects on 12-LOX activity may derive from the residual salicylate moiety or from negative feedback exerted by the TXA2 receptor. The in vivo relevance of these observations is currently unknown.
Compared with other antiplatelet agents such as thyenopyridines, prodrugs requiring a multistep (CYP450- or esterase-dependent) biotransformation, the PK of aspirin is less complex (Figure 3). In the stomach, the low pH facilitates the absorption by passive diffusion of plain aspirin, largely undissociated, but aspirin is also absorbed in the upper small intestine. Acetylsalicylic acid preparations undergo presystemic metabolism and approximately 50–70% of an oral dose reaches the systemic circulation, while the remaining fraction is inactivated – that is, de-acetylated to salicylic acid – by the carboxylesterases in plasma and the liver (first-pass effect). The hepatic human carboxylesterase-2 isozyme mainly accounts for the first-pass effect and it is also present in the gut where it may contribute to aspirin de-acetylation. In addition to hepatic human carboxylesterase-2, other esterases are present in the peripheral blood including cholinesterases, intraerythrocyte hydrolases and other undefined esterases called 'aspirin esterases'. The residual salicylic acid is virtually unable to inhibit TXA2 generation from human platelets when aspirin is given at low doses.
After a single oral-dose of plain aspirin, plasma peak concentration is reached within 30–40 min and circulating aspirin has a half-life of 15–20 min. Enteric-coated (EC) formulations take a longer time to peak and acetylsalicylic acid is detectable for longer intervals in the peripheral blood, depending on the formulation.