Cost-Effectiveness of Platelet Function Testing in Dual Antiplatelet Therapy Decision-Making after Intracranial Aneurysm Treatment with Flow Diversion

INTRODUCTION

Over the past decade, the adoption of Flow Diversion (FD) for the management of intracranial aneurysms (IAs) has shown consistent expansion.¹ Flow diverters’ elevated metal coverage ratio necessitates the post-procedural application of dual antiplatelet therapy (DAPT) to mitigate the risk of thromboembolic complications (TECs) after stent placement. However, there is no consensus in the literature as to what the optimal postoperative DAPT regimen should be.

Across the US, the combination of Clopidogrel and aspirin represents the predominant choice for DAPT following FD.² However, 16 to 52% of patients exhibit Clopidogrel resistance, potentially elevating their risk of TECs to as much as 17%.^3–5 To address those concerns, up to 90% of US institutions now implement various platelet function testing (PFT) to determine patients’ responsiveness to Clopidogrel before FD stent placement.² Currently, the VerifyNow test, which quantifies platelet reactivity units (PRU), stands as the prevailing PFT in neurointervention due to its practicality, allowing for bedside implementation.³

However, not only does PFT bear a cost, but alternative antiplatelet medications such as Prasugrel or Ticagrelor, which may be prescribed in cases of Clopidogrel resistance, are often more expensive for patients, potentially affecting compliance to DAPT.⁶ In some cases, health insurance companies are reluctant to approve higher costs for these alternative antiplatelet therapies unless PFT has been done.

This study aims to evaluate the cost-effectiveness of various strategies for determining the most appropriate postoperative DAPT regimen following FD intervention for unruptured IAs, focusing on comparing universally applied DAPT protocols and those guided by VerifyNow PFT outcomes.

METHODS

Results

Deterministic Sensitivity Analysis

The deterministic sensitivity analysis involves adjusting the values of specific variables that may be relevant to the outcome, conducting this adjustment individually for one-way sensitivity analysis and in pairs for two-way sensitivity analysis, while maintaining the constancy of other variables. This process evaluates if the preferred strategy shifts when these variable ranges are altered. The variables noted to have the most considerable impact on incremental NMB were TEC and bleed mortality rates, along with disability rates after a TEC or bleed (Figure 2A). Given that variation in complication rates is the primary differentiating factor between regimens, other than cost, this comes as no surprise. An additional variable that we considered for a sensitivity analysis was the cost of platelet function testing itself, given that this would be a determining factor at many institutions (Figure 2B). Key parameters, such as treatment costs, effectiveness, and discount rates, were individually altered within plausible limits to evaluate their impact on the ICER. Threshold analyses identified the critical points at which parameter changes would alter the preferred strategy.

• Download figure
• Open in new tab

Figure 1 – Simplified decision tree used to model five different DAPT regimens after FD. TEC, thromboembolic complications. mRS, modified Rankin Scale.

• Download figure
• Open in new tab

Figure 2A – Two-way sensitivity analysis varying the rate of mortality after a hemorrhagic complication (x-axis) and the rate of mortality after an ischemic complication (y-axis).

• Download figure
• Open in new tab

Figure 2B – One-way sensitivity analysis varying the cost of platelet function testing. A higher NMB is more cost-effective. The values on the x-axis and y-axis are USD.

Probabilistic Sensitivity Analyses

A probabilistic sensitivity analysis was conducted, simulating a cohort of 10,000 patients across various iterations. Within this framework, five distinct follow-up strategies were evaluated using the Monte Carlo simulation method. This analysis revealed that PFT-PSG continued to be the most economically viable option, demonstrating the greatest NMB for any WTP threshold, as depicted in the cost-effectiveness acceptability curve (Figure 3). Further, Monte Carlo acceptability analysis indicated that PFT-PSG was identified as the most cost-effective strategy in approximately 70% of iterations, considering the standard WTP threshold of $100,000 per QALY (Figure 3).

• Download figure
• Open in new tab

Figure 3 – Probabilistic sensitivity analysis: cost-effectiveness acceptability curve at various WTP values.

Discussion

Cost-effective analyses, like our study utilizing a decision tree model, help understand the value of medical interventions in terms of cost per quality-adjusted life year (QALY) gained. Our study demonstrated that PFT-Prasugrel (Clopidogrel for responders and Prasugrel for non-responders detected on PFT) is the most cost-effective DAPT strategy for post-FD treatment of UIAs. Universal Prasugrel, PFT-Ticagrelor, and universal Ticagrelor were also considered cost-effective compared to universal Clopidogrel, assuming a WTP threshold of $100,000/QALY. Prasugrel and Ticagrelor are associated with higher annual drug costs, compared to Clopidogrel. However, these higher costs were offset in corresponding regimens by their better effectiveness in improving thromboembolic complications, which is critical in patients treated with FD. These findings conclude that the use of PFT to guide the choice of DAPT regimen following FD is not only clinically effective but also cost-effective.

Our study is the first of its kind in evaluating the cost-effectiveness of PFT to guide DAPT regimen choice for patients post-FD treatment for unruptured IAs. Previous studies in cardiovascular literature, however, have studied the cost-effectiveness of DAPT. Coleman et al. compared similar DAPT strategies in patients with acute coronary syndrome (ACS), demonstrating that platelet reactivity assay (PRA) (e.g., VerifyNow) driven Prasugrel and Ticagrelor were cost-effective to universal Clopidogrel.⁶ In this study, universal Ticagrelor and Prasugrel were also found to be cost-effective compared to their respective PRA-driven regimens. These results were most sensitive to differences in drug costs and drug-specific relative risks of death. It should be noted that whilst our study did not focus on a specific age group, Coleman et al. focused on a defined cohort of 65-year-old ACS patients.⁶ Additionally, the group’s economic analysis was primarily based on model assumptions and hypothetical cohorts, which might limit the generalizability of the results. Other cardiovascular studies have considered cost-effectiveness via genotype testing-driven DAPT regimens, checking for loss-of-function (LOF) alleles, after ACS and percutaneous coronary intervention. Limdi et al. found that utilizing both universal Ticagrelor and genotype-guided escalation of Ticagrelor was more cost-effective than universal Clopidogrel.¹⁵ A secondary analysis in this study, which evaluated genotype-guided de-escalation, showed higher effectiveness compared to nonguided de-escalation but was not cost-effective under most WTP thresholds. In addition to drug prices, the cost-effectiveness of genotype-guided strategies in this study could be sensitive to the prevalence of LOF genes. A second genotype-guided economic evaluation model by Kazi et al. showed genotype-guided therapy using Ticagrelor or Prasugrel for patients with LOF alleles is cost-effective compared to universal application of these drugs.¹⁶ In this study, genotype-guided Ticagrelor had an ICER of $30,200 per QALY compared to Clopidogrel. Temporal differences exist between our study and the evaluation by Kazi et al., who discuss lifetime costs and outcomes, compared to the shorter 3-month focus we used which is the typical amount of time that DAPT regimens are required post-FD.¹⁶ These studies evaluating the cost-effectiveness of DAPT regimens emphasize that personalized approaches to managing DAPT, whether through genetic testing or PFT, can lead to better economic and clinical outcomes. However, the specific focus on different patient populations, interventions, and biomarkers for personalization suggests that findings from each study are best applied within their respective contexts.

Despite our findings’ clear benefit of PFT, there is no consensus on the cut-off point for defining Clopidogrel hyporesponsiveness. Cut-off values of P2Y12 reaction units (PRU) of 85–208 have been described in cardiac literature as adequate Clopidogrel response, with PRU < 85 reported to be associated with a higher risk of bleeding events and PRU > 230 associated with significant morbidity and mortality.^6,17 PRU studies in neurosurgery literature report adequate Clopidogrel response to fall between the ranges of PRU 60–240, with dose changes required at either end of the scale: for PRU < 60, lower doses are advised; for PRU >240, increasing Clopidogrel dose or switching to either Prasugrel or Ticagrelor is advised. To further reduce complication risk, additional Clopidogrel stratification has been employed, with Clopidogrel response defined as extreme response (PRU ≤ 15) and hyperresponse (PRU 16 – 60).¹⁸ In addition to Clopidogrel responsiveness, PRU responsiveness to Prasugrel has also been studied, with dose changes effected according to cut-off values. Higashiguchi et al. and Suyama et al. reported three different Prasugrel doses in patients post-FD treatment, with a 20 mg loading dose and daily 3.75mg for PRU > 210, daily 3.75mg for PRU 60 – 210, and daily 1.875 mg for PRU < 60.^8,19 Variability in pharmacodynamic response to antiplatelets is also seen with Ticagrelor, where the established twice-daily 90 mg dose can lead to PRU values < 40, converting Clopidogrel hyporesponders to extreme responders.^18,20

Additionally, universal Prasugrel was a cost-effective strategy and commonly preferred regimen at several institutions.² Resistance to Prasugrel, which has been reported in other cohorts, has been more linked to medication adherence and pharmacological interactions, rather than genetic variants.²¹ In contrast to Clopidogrel, Prasugrel is a prodrug that requires activation through intestinal esterase and, to a lesser extent, CYP2C19 and CYP2C9.²² Additionally, unlike Clopidogrel, studies evaluating genetic polymorphisms have not revealed any effects on the pharmacokinetics and pharmacodynamic effects of Prasugrel.²³ Possible explanations for this phenomenon include alternative activation through different hepatic CYP enzymes and higher potency as an irreversible ADP P2Y12 receptor antagonist.^3,24,25 Due to Prasugrel’s high potency, some initial concerns arose regarding the heightened risk of intracranial hemorrhage in patients with a history of ischemic stroke.^26,27 For outpatient neurovascular procedures, as unruptured IAs treated with FD, various cohorts have demonstrated a favorable safety profile for Prasugrel.^28–31 In the meta-analysis by Podlasek et al., comprising 49 studies and 2,526 patients undergoing FD treatment for unruptured IAs, hemorrhagic events were negligible for Prasugrel.¹² Similarly, in the cohort reported by Souyama et al., in which there was routine use of Prasugrel post-FD for 3-6 months, out of the 110 patients, only one had a hemorrhage after treatment.⁸ Overall, these findings underscore Prasugrel’s promising safety and efficacy profile in the context of FD, warranting further exploration in prospective clinical trials.

While insightful, our study is not without limitations. The reliance on retrospective cohorts and model-based simulations may not capture the full complexity of individual patient responses in a real-world setting. The heterogeneity of included studies also limits our retrospective approach in determining thromboembolic and hemorrhagic rates. The paucity of data from meta-analyses also led to our reliance on the best available retrospective cohorts, which has limitations. Retrospective analyses are subject to selection bias and confounding. This, along with the heterogeneity in the sample set and variation in the duration of DAPT, may produce findings that are not always generalizable. Moreover, we also omitted double-dose Clopidogrel as a DAPT regimen, which may be efficacious in a subset of individuals with Clopidogrel hyporesponsiveness.³² An additional limitation that was difficult to circumvent due to the lack of stratified reporting in the literature was the difference in the rates of various types of bleeding events and the mortality rates of each of those types, which also varies by type of regimen. There were not enough studies considering all these possible variations and elucidating each of these individual rates. Finally, our use of the decision tree model for economic valuations, which is contingent on the accuracy of the cost data and assumptions we embedded in the model, might not reflect the dynamic nature of healthcare economics and patient management. Notably, our decision tree model did not account for the influence of age, gender, or specific patient comorbidities. It also assumed that rates of thromboembolic and hemorrhagic events would remain constant over the three-month study period. Thus, future studies should aim to incorporate more diverse patient data and real-world outcomes, where available, to validate and refine the proposed cost-effectiveness framework.

Conclusion

We have demonstrated that PFT-guided therapy, using Prasugrel for Clopidogrel hyporesponsiveness, was the most cost-effective strategy guiding DAPT post-FD treatment of unruptured IAs. Furthermore, prescribing universal Prasugrel, PFT-guided Ticagrelor, or universal Ticagrelor were also cost-effective compared to prescribing universal Clopidogrel. Although this sheds light on the potential cost benefits of initiating Prasugrel DAPT upfront, future prospective, and ideally randomized clinical trials are necessary to evaluate the safety of this intervention post-FD.

Declarations