Do not adjust the platelet count in light transmittance aggregometry when predicting thrombotic events after percutaneous coronary intervention

Authors


Jurriën M ten Berg, Department of Cardiology, St Antonius Hospital, P.O. Box 2500, 3435 CM Nieuwegein, the Netherlands.
Tel.: +31 306099111; fax: +31 306034420.
E-mail: berg03@antoniusziekenhuis.nl

Dual antiplatelet therapy with aspirin and clopidogrel reduces thrombotic complications in patients undergoing percutaneous coronary intervention (PCI) [1,2]. A growing body of evidence demonstrates that the efficacy of dual antiplatelet therapy is highly variable and that high on-treatment platelet reactivity (HPR) is associated with the occurrence of atherothrombotic events [3–6].

‘Classical’ light transmittance aggregometry (LTA) is still considered to be the reference standard for quantifying the magnitude of on-treatment platelet reactivity, but it remains poorly standardized, and various parameters are used by different laboratories to determine the magnitude of platelet reactivity [7]. The adjustment of the platelet-rich plasma (PRP) to achieve a platelet count of 250 000 μL−1 has been proposed to standardize LTA in patients with bleeding diasthesis. As no clinical endpoint studies have previously determined whether LTA with either ‘native’ or adjusted (standardization of the platelet count to 250 000 μL−1) PRP is a better predictor of adverse events, it remains controversial whether adjustment of the platelet count is necessary for the monitoring of antiplatelet therapy in cardiovascular patients on aspirin and clopidogrel. Therefore, the aim of the present study was to evaluate the value of LTA in predicting atherothrombotic events, with both ‘native’ and adjusted PRP.

A prospective cohort study of consecutive patients undergoing elective PCI with stent implantation was performed [4]. All patients received clopidogrel treatment before PCI, and all patients were on aspirin at a dose of 80–100 mg daily for ≥ 10 days, unless they were on long-term anticoagulation with coumadins. The primary endpoint was defined as a composite of all-cause death, non-fatal myocardial infarction, definite stent thrombosis and ischemic stroke.

LTA was assessed on a four-channel APACT 4004 aggregometer (LABiTec, Arensburg, Germany). Samples were centrifuged for 10 min at 150 × g to obtain native PRP. Platelet-poor plasma (PPP) was obtained by centrifuging the remaining sample at 1500 × g for 10 min. Half of the amount of native PRP was adjusted (with PPP) to achieve a calculated platelet count of 250 000 μL−1. Patients with a platelet count < 300 000 μL−1 in PRP before adjustment were excluded. PPP was set as 100% aggregation, and maximal (peak) platelet aggregation (%) induced by ADP in a final concentration of 20 μmol L−1 was measured in PRP. To evaluate LTA’s ability to discriminate between patients with and without atherothrombotic events 1 year post-PCI, a receiver operator characteristic (ROC) curve analysis was performed for both adjusted and non-adjusted PRP. The optimal cut-off level was calculated by determining the smallest distance between the ROC curve and the upper left corner of the graph (i.e. the point with the highest sensitivity as well as specificity). Patients above the optimal cut-off were considered to exhibit HPR. The predictability of the parameters, that is, the ability of the test to correctly classify those with and without atherothrombotic event, was expressed as area under the curve (AUC).

LTA induced by 20 μmol L−1 ADP was performed in 1051 patients undergoing elective PCI with stent implantation. Of these, 753 had a platelet count in native PRP of > 300 000 μL−1. Owing to logistic demands or a low volume of PRP, PRP samples of 692 patients were adjusted to a platelet count of 250 000 μL−1. The latter cohort was used for the present analysis. The platelet count in native PRP was 418 600 ± 92 900 μL−1, as compared with 260 300 ± 23 000 μL−1 in adjusted PRP. The magnitude of on-treatment platelet reactivity was significantly higher when native PRP was used than when adjusted PRP was used (58.2% ± 14.0% vs. 49.2% ± 16.4%, P < 0.0001). In addition, the ROC curve-derived cut-off value was higher when native PRP was used than when adjusted PRP was used (67.0% vs. 58.7%).

At 1-year follow-up, the primary endpoint occurred more frequently in patients with HPR than in patients without HPR, with the use of both native PRP [30/200 (15.0%) vs. 33/492 (6.7%); odds ratio (OR) 2.45; 95% confidence interval (CI) 1.45–4.15, P = 0.001] and adjusted PRP [30/243 (12.3%) vs. 33/449 (7.3%); OR 1.78; 95% CI 1.05–2.99, P = 0.04] (Fig. 1). In addition, the predictability was similar in LTA with native PRP and with adjusted PRP (AUC 0.59; 95% CI 0.52–0.66 for both). The negative predictive values (NPVs) of both were high, and the positive predictive values (PPVs) were low (NPV of 93.3.% and PPV of 15.0% with native PRP; NPV of 92.7.% and PPV of 12.3% with adjusted PRP), which is in agreement with other platelet function studies linked to clinical outcome.

Figure 1.

 Kaplan–Meier analysis for the event rate for the combined primary endpoint in patients with and without high on-treatment platelet reactivity as measured in both native and adjusted platelet-rich plasma (PRP). In adjusted PRP, the platelet count was adjusted to a calculated platelet count of 250 000 μL−1. HPR, high on-treatment platelet reactivity according to the defined cut-off (i.e. > 67.0% aggregation in native PRP and > 58.7% in adjusted PRP). NPR, normal on-treatment platelet reactivity according to the defined cut-off.

Although LTA is still regarded as the reference standard method, this technique is poorly standardized, as no external quality assessment is available and no standard platelet function testing protocol has been unanimously adopted. Throughout the last two decades, several attempts have been made to increase the between-center comparability of LTA by standardization of: (i) platelet storage temperature prior to testing; (ii) stirring rate; (iii) centrifugation speed to obtain PRP and PPP; (iv) agonist sources; and (v) adjustment of the platelet count in PRP to a standard count [7–13]. However, the frequently used procedure of adjusting the platelet number in PRP is cumbersome, may affect platelet activation, and has been questioned because it does not reflect platelet function in vivo [7,14,15]. Moreover, the avoidance of the time-consuming step of platelet count adjustment would make LTA more applicable for the routine monitoring of antiplatelet therapy in clinical practice [14].

In the present study, the predictive value of both native and platelet count-adjusted PRP for the occurrence of adverse atherothrombotic events was evaluated. Although the ROC-based cut-off value for segregatation of patients with and without HPR was considerably lower when PRP was adjusted, the two procedures have equal ability to predict adverse clinical outcomes. Thus, the adjustment of platelet count does not provide additional information. LTA with native PRP is easier to perform and has a similar accuracy in predicting atherothrombotic events. Therefore, our advice is not to adjust the platelet count in PRP when predicting thrombotic events after PCI.

Disclosure of Conflict of Interests

The authors state that they have no conflict of interest.

Ancillary