Lack of reproducibility of assessment of aspirin responsiveness by optical aggregometry and two platelet function tests

The term aspirin-resistance describes the failure of aspirin to inhibit thromboxane A2 production. Many new tests have become available for potentially measuring aspirin responses but some are non-specific and do not isolate COX-1 activity. We previously demonstrated that agreement between two tests (PFA-100 (R) and VerifyNow (R)-ASA) and light transmission aggregation (LTA) was no greater than would be expected by chance. In this study we re-tested the same patients using identical methods after 1 year to determine whether poor agreement might have been due to assessment in the acute phase and whether the results of the individual tests are consistent over time. Platelet function by all three tests was re-tested in the 72 patients who were alive and still receiving low dose ASA therapy one year after the first tests were performed. On re-testing the prevalence of ASA non-responsiveness compared with baseline was 10% vs 17% by the VerifyNow (R)-ASA test, 25% vs 22% by the PFA-100 (R), and 1% vs 5% by LTA. Agreement between the tests at 1 year remained poor (kappas: 0.02-0.17) and only one patient was identified as a non-responder by all three tests, in keeping with the theoretical differences between the tests. Within test comparisons of baseline vs 1 year showed moderate agreement for the PFA-100 (R) (kappa = 0.44, 95% CI 0.19-0.68, p = 0.0006), a fair agreement for VerifyNow (R)-ASA (kappa = 0.34, 0.04-0.64, p = 0.12) and poor agreement for LTA (kappa = 0.14, -0.11 -0.39, p = 0.24 for ADP; kappa = 0.09, -0.21-0.39, p = 0.41 for arachidonic acid). Agreement between the three tests in identifying aspirin non-responsiveness remained poor in patients who had been taking aspirin for at least 1 year follow-up. Reproducibility over time was no greater than chance for LTA and only moderate for VerifyNow (R)-ASA and PFA-100 (R). Lack of consistency over time in identification of apparently non-responsiveness individuals is likely to substantially undermine any ability of these tests to predict risk of recurrent vascular events.