Download PDF
Thromb Haemost 2024; 124(06): 595-597
DOI: 10.1055/a-2216-5263
Letter to the Editor

Guided Anti-P2Y12 Therapy in Patients Undergoing Percutaneous Coronary Intervention

Authors

  • Marco Cattaneo 

    1   Divisione di Medicina Generale II, ASST Santi Paolo e Carlo, Dipartimento di Scienze della Salute, Università degli Studi di Milano, Milan, Italy
    2   Fondazione Arianna Anticoagulazione, Bologna, Italy

  • Alessandro Squizzato 

    3   Research Center on Thromboembolic Disorders and Antithrombotic Therapies, ASST Lariana, University of Insubria, Como, Italy

  • Simone Birocchi

    1   Divisione di Medicina Generale II, ASST Santi Paolo e Carlo, Dipartimento di Scienze della Salute, Università degli Studi di Milano, Milan, Italy

  • Gian Marco Podda 

    1   Divisione di Medicina Generale II, ASST Santi Paolo e Carlo, Dipartimento di Scienze della Salute, Università degli Studi di Milano, Milan, Italy
Further Information
   Permissions and Reprints

We read with interest the comprehensive editorial by Landolina and collaborators,[1] commenting our systematic reviews with meta-analyses of randomized controlled trials (RCTs) addressing the issue of optimizing anti-P2Y12 therapy in patients undergoing percutaneous coronary intervention.[2] We thank the authors for their appreciation of our study, but wish to clarify better some points, which were probably not well illustrated in our manuscript.

Although we agree with Landolina and collaborators that the exploration of subgroups is a strength of our study, we believe that its major strength lies on the stringent methodological approach in distinguishing among RCTs with different designs. Previous systematic reviews with meta-analyses focused on guided therapy (GT) with P2Y12 antagonists in general terms, lumping together RCTs testing different strategies of GT, with different accuracy and precision in the identification of patients potentially benefiting from switching from standard to alternative treatments.[3] [4] We believe that the general question of whether GT is useful[3] [4] is impractical, as it fails to identify which type of guided strategy, if any, should be implemented. RCTs that correctly tested the value of GT strategies compared the results obtained in patients randomized to GT with those obtained in patients randomized to unguided therapy. Two types of GT strategies were tested. In one type, alternative therapies were given based on the presence of genic polymorphisms affecting the pharmacologic efficacy of anti-P2Y12 treatment (Genotype-GT). In the other type, alternative therapies were given based on the degree of reactivity of platelets in patients on anti-P2Y12 treatment, detected by platelet function tests (PFTs) (PFT-GT). In our study, we meta-analyzed these two types of RCTs separately. Surprisingly, our study was the first one and only analyzing the value of PFT-GT, despite the fact that this is the first and most widely used strategy for GT. No GT strategy affected the incidence of bleeding. Only Genotype-GT reduced the incidence of major adverse cardiovascular events (MACE), while PFT-GT had no statistically significant effect.[2] Subgroup analyses showed that the results were not affected by the prevalence of patients with acute coronary syndromes, not confirming the assumption of some authors.[5] [6] In addition, our subgroup analyses showed that both types of GT significantly reduced the incidence of MACE in RCTs performed in China, while they were without statistically significant effect in the Rest of the World (RoW).[2] This difference was not due to differences in statistical power, which was higher in RCTs performed in RoW,[2] but likely attributable to the high prevalence in South/East Asian populations of gene mutations affecting the response to clopidogrel.[7] [8] [9] [10]

Failure of PFT-GT is likely due to the known inaccuracy of PFTs, the instability of the platelet reactivity phenotype, and lack of standardization of preanalytical variables.[11] [12] [13] [14] [15] [16] [17] [18] [19] This interpretation is corroborated by the comparison of the results of PFT-GT with those of RCTs on therapy of “high on-treatment platelet reactivity” (HTPR-Therapy). These RCTs randomized patients with HTPR to standard clopidogrel therapy or alternative therapy with prasugrel, ticagrelor, or double-dose clopidogrel. Thus, they are of no value in the assessment of the utility of GT and should therefore be ignored in this setting, for the simple reason that they did not compare GT with unguided therapy. We deemed relevant to analyze also these RCTs in our meta-analyses,[2] because they could help interpreting the data of RCTs on PFT-GT, as they used the same PFTs and the same alternative therapies, but followed up only a selected small proportion of patients in whom PFTs detected HTPR, rather than the whole patient population that was screened by PFTs ([Fig. 1]). At variance with the negative results of PFT-GT RCTs, HTPR-Therapy RCTs showed that alternative treatments significantly reduced the risk of MACE ([Fig. 1]). These results confirm that these alternative therapeutic regimens are more effective than standard-dose clopidogrel, as it had been previously shown by phase III RCTs,[20] [21] [22] and would very likely have been obtained independently of the chosen criteria for patient selection. This interpretation is supported by the observation that, at variance with Genotype-GT and PFT-GT RCTs, the results of HTPR-Therapy RCTs were apparently not influenced by the background prevalence of unfavorable gene mutations, because the point estimates of relative risk (RR) of MACE were similar in RCTs performed in China (0.56) and RoW (0.58), with the lack of statistical significance in RoW likely due to insufficient statistical power.[2] If the accuracy of PFT were 100%, PFT-GT RCTs would have obtained very similar results to those of HTPR-Therapy. Indeed, the good results obtained in patients with HTPR by switching them to alternative therapies that decreased their platelet reactivity would have been observed also in patients who did not display HTPR already when they were on standard treatment ([Fig. 1]). However, due to the aforementioned pitfalls of PFTs, both groups of patients, with and without HTPR, were most likely nonhomogeneous, including a large proportion of patients without and with HTPR, respectively ([Fig. 1]). Misclassification of patients would preclude poor responders to clopidogrel (displaying HTPR) from being treated with more efficient regimens, while good responders (not displaying HTPR) would unnecessarily switch to these alternative regimens, thus eventually leading to dilution and underestimation of the potential clinical benefit of GT in the overall population. Obviously, only the results of PFT-GT RCTs reflect the effects of monitoring patients by PFTs in real life, because they consider the clinical outcomes of all patients undergoing PFTs screening, not only of those treated with more efficient drug regimens.

Zoom
Fig. 1 Differences between RCTs on HTPR-Therapy and RCTs on PFT-GT and potential influence on the results of real-world accuracy and ideal 100% accuracy of PFTs. Left panels refer to the results of the RCTs on HTPR-Therapy and PFT-GT that have been meta-analyzed by Birocchi et al.[2] The total number (100) of patients represented in the lower panel represents 100% of the total population enrolled in the RCTs on PFT-GT. In 34 of these, PFTs detected HTPR, reflecting the total percentage of such patients (34%) in all PFT-GT RCTs meta-analyzed by Birocchi et al.[2] For a better comparison, the same number (34) of patients diagnosed as displaying HTPR in HTPR-Therapy RCTs is shown in the upper panel. The composition of all groups of patients is not homogeneous in terms of presence of “HTPR” and “No HTPR,” reflecting the known inaccuracy and pitfalls of platelet function testing (see the text). We arbitrarily set at 60% the accuracy of PFTs, which are likely very accurate for detecting the few patients with platelet reactivity at the extreme HTPR and LTPR values of the distribution, but are very inaccurate for distinguishing between HTPR and No HTPR in the vast majority of patients distributed around the cut-off value distinguishing between the two platelet reactivity phenotypes. The choice of a more optimistic value of accuracy (e.g., 70%) would not affect the message that our simulation wishes to convey. Right panels refer to the theoretical differences in the results obtained by RCTs, if the accuracy of PFTs were 100% (see the text): the point estimates should be very similar to that calculated in the meta-analysis of HTPR-Therapy RCTs.[2] HTPR, high on-treatment platelet reactivity; LTPR, low on-treatment platelet reactivity; PFT, platelet function test; PFT-GT, platelet function test guided therapy; RCT, randomized controlled trial; RR, relative risk.

In conclusion, both strategies of GT proved effective in China, likely due not only to the high prevalence of unfavorable gene mutations but also to the particularly high risk for MACE associated with them in populations of East/South Asia.[23] [24] In the RoW, the dubious efficacy of genotype-GT reflects the marginal role played by gene mutations in the response to clopidogrel in these countries.[25] The unquestionable failure of PFT-GT in the RoW to reduce MACE (RR = 1.08) should be adequately emphasized, considering that PFT-GT is still implemented in many centers in Western countries, absorbing economic resources and increasing the workload of health care workers.



Publication History

Received: 30 October 2023

Accepted: 21 November 2023

Accepted Manuscript online:
22 November 2023

Article published online:
18 December 2023

© 2023. Thieme. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany