Determinants of the variability of thromboxane inhibition by aspirin in patients with cardiovascular disease

Overview analysis of indirect comparisons in clinical trials - showing saturation of the antithrombotic efficacy of aspirin at low-doses(75-150 mg daily)- sustains that suppression of platelet cyclooxygenase(COX)-1-dependent thromboxane(TX)A2 biosynthesis is the mechanism of action of aspirin in cardioprotection. Heterogeneity of COX-1 suppression by aspirin has been detected in cardiovascular (CV) disease and might contribute to aspirin failure to prevent clinical events. Among the possible mechanisms, the recent recognized capacity of platelets to make COX-1 de novo might be involved in the variable response to aspirin. Thus, we performed a clinical study in CV patients (n=201) and healthy subjects (n=31) chronically treated with low-dose aspirin (cardioaspirin, Bayer, 100 mg for at least 7 consecutive days) to investigate the determinants of the variability of TXA2 inhibition both ex vivo and in vivo. We assessed: i)serum TXB2 ex vivo, a capacity index of platelet COX-1 activity, ii)TXA2 biosynthesis in vivo, through the measurement of 11-dehydro-TXB2, a major enzimatic metabolite of TXB2; iii) time-dependent recovery of TXA2 biosynthesis in vitro in washed platelets stimulated with thrombin(0.5 U/ml) and fibrinogen(0.38 mg/ml) for 15 min and 24h. We enrolled 201 CV patients(130 males, 70±9y, mean±SD): among them 56.7% were diabetic, 43.3% non diabetic. Serum TXB2 levels in patients were 1.33 (0.015-156) ng/ml (median and range); only 6.5% of patients showed serum TXB2 levels higher than 10 ng/ml, i.e. the upper limit value for an adequate inhibition of platelet COX-1 by aspirin in healthy subjects. Serum TXB2 levels detected in diabetic patients were not significantly different from those of non diabetic patients [1.45(0.015-31.8) vs 1.3(0.04-156) ng/ml, respectively]. Moreover, in both groups, a similar percentage of patients showed serum TXB2 values higher than 10 ng/ml, i.e. 5 and 6% respectively. The addition of aspirin in vitro (50 microg/ml) was associated with a significant (P<0.01) reduction of serum TXB2 levels which were lower than 10 ng/ml in all patients except one who still generated 16 ng/ml. In washed platelets of aspirin-treated patients stimulated with thrombin and fibrinogen, TXB2 recovered in a time-dependent fashion. At 15 min and 24 h, it averaged 9.5 and 17.7ng/ml, respectively (P<0.01). The finding that the irreversible inhibition of platelet COX-1 by aspirin was partially recovered in a time-dependent fashion may suggest the occurrence of de novo synthesis of COX-1 in vitro. In a subgroup of patients(n=53), we verified whether mRNA levels of COX-1 may be a determinant of the extent of inhibition of platelet COX-1 by aspirin using real time-PCR (polymerase chain reaction). The levels of COX-1 mRNA were 1.8(0.37-8.3) relative units. In patients with serum TXB2 levels higher than 10 ng/ml, COX-1 mRNA levels [1-83(1.25-6) relative units] were not different from those detected in the rest of the population. Moreover, COX-1 mRNA levels did not correlate with serum TXB2 levels. Enhanced urinary excretion of 11-dehydro-TXB2 was detected in patients vs healthy subjects treated with aspirin [223(41-949) vs 102(0.4-250) pg/mg creatinine, rispectively, P<0.001]. Similar values of 11-dehydro-TXB2 were measured in diabetic and non diabetic patients: 228 vs 184 pg/mg creatinine; both groups showed 11-dehydro-TXB2 levels significantly(P<0.01) higher than those detected in healthy subjects. Urinary 11-dehydro-TXB2 levels significantly correlated with circulating white blood cell count (rs=0.48, P<0.05) suggesting the contribution of extraplatelet sources to increased TXA2 biosynthesis in vivo detected in CV patients treated with aspirin. No correlation was found between urinary levels of 11-dehydro-TXB2 and a systemic marker of inflammation (i.e. C Reactive Protein) and markers of glicemic control (i.e. glycemia values and % glycosylated hemoglobin). In conclusion, in CV patients, aspirin resistance (i.e. less sensitivity of COX-1 to inactivation by aspirin) is a rare phenomenon. Low-dose aspirin failed to cause an appropriate suppression of platelet COX-1 for low bioavailability, in some patients. We detected a more frequent occurrence of enhance synthesis of TXA2 in vivo from extra-platelet sources. These results open the way to study whether the coadministration of a TX receptor antagonist with aspirin would improve cardioprotection.