Editorials
PHILO – Ensuring Trial Results Are Not Lost in Translation –
2015 Volume 79 Issue 11 Pages 2326-2328
Details
2015 Volume 79 Issue 11 Pages 2326-2328
Guideline recommendations support the use of dual antiplatelet therapy in patients after acute coronary syndrome (ACS) or percutaneous coronary intervention (PCI). In particular, prasugrel and ticagrelor have demonstrated superior cardiovascular (CV) efficacy as compared with clopidogrel in patients after ACS, but with increased bleeding.1,2 Yet, there remains continued uncertainty regarding the relative efficacy and safety of these more potent P2Y12 inhibitors in East Asian populations. In part, this debate stems from a paucity of phase III clinical trial data to support the use of these novel P2Y12 inhibitors in patients of East Asian ethnicity.3
Article p 2452
However, a key question is whether there is a priori reason to believe that individuals of East Asian descent might respond differently to antiplatelet therapies than Caucasian patients. Also, if such variability exists, it is critical to better understand whether these differences are explained by baseline differences, such as lower body weight or comorbidities, or whether these differences have a genetic basis. Such findings could have relevant implications for patients worldwide regardless of ancestry.
As with many established pharmacotherapies, interethnic variability in the response to P2Y12 inhibitors has been reported; in particular, higher on-treatment platelet reactivity has been observed in East Asians treated with clopidogrel as compared with Caucasian patients.3 In part, this finding may be explained by a higher prevalence of the CYP2C19 loss-of-function (LOF) allele in East Asian (65%) vs. Caucasian (30%) patients, as the CYP2C19 enzyme is required for the bioactivation of clopidogrel to its active metabolite.4,5 Yet, despite the diminished pharmacodynamic response to clopidogrel, many studies have reported that Asian patients have a similar or lower risk of CV events after ACS or PCI than Western patients, a phenomenon often described as the “East Asian paradox.”3 In contrast, many studies have reported a higher risk of bleeding in East Asian patients treated with clopidogrel, thereby raising concerns that East Asian patients may have a different “therapeutic window” with regard to platelet reactivity.3,6
The concerns about net clinical benefit are perhaps even more pertinent for the newer and more potent P2Y12 inhibitors, ticagrelor and prasugrel. In contrast to clopidogrel, greater platelet inhibition has been observed in East Asian patients treated with prasugrel and ticagrelor when compared with Caucasians. In phase I studies of prasugrel, Asian subjects exhibited higher exposure to the active metabolite and increased platelet inhibition when compared with Caucasian patients even after normalizing for body weight.7,8 Based on concerns about excess bleeding, a lower maintenance dose of prasugrel (3.75 mg daily) was evaluated in Japanese clinical trials and subsequently approved in Japan.9,10 Similarly, higher drug exposure and greater inhibition of platelet aggregation have been described in Japanese patients treated with ticagrelor, as compared with Caucasians, in single and ascending dose studies.11,12 Neither ticagrelor nor prasugrel is believed to be significantly affected by the CYP2C19 LOF allele.
Yet, clinical outcome studies of the third-generation P2Y12 inhibitors in Asian patients are relatively few. A post-hoc analysis conducted from the PLATO trial13 demonstrated that the relative efficacy and safety of ticagrelor was comparable in patients of Asian (South Asian or East Asian) descent (n=1,106) when compared with non-Asian patients. Of interest, the crude incidence of bleeding was also similar across Asian and non-Asian patients despite differences in baseline demographics. The comparable safety signal with ticagrelor is perhaps even more notable, given the higher prevalence of the CYP2C19 LOF allele in Asian patients, because it might be hypothesized that an exaggerated bleeding signal with ticagrelor would be seen on the basis of diminished pharmacodynamic response to clopidogrel in the control arm in Asia relative to other patients worldwide.
Because few patients in PLATO were enrolled in East Asia, the PHILO trial, reported in this issue of the Journal,14 was designed to mimic the study protocol of PLATO but with patients exclusively enrolled in Japan (n=721), South Korea (n=44) and Taiwan (n=35) (Table); the study was conducted at the request of regulatory agencies. Although the trial was not adequately powered for clinical efficacy, Dr Goto and colleagues should be commended for conducting a rigorous trial in an understudied patient population. In PHILO, 801 patients hospitalized with ACS were randomly assigned to ticagrelor or clopidogrel. In contrast to PLATO, all patients were intended for PCI and were required to be on a background of low-dose aspirin (higher aspirin doses were allowed in PLATO and patients could be medically managed). The observed event rate in PHILO was lower than expected, thereby limiting the ability of the investigators to evaluate with certainty either safety or efficacy. At 12 months, overall major bleeding tended to be higher in ticagrelor-treated patients as compared with those treated with clopidogrel, although the difference was not significant (HR 1.54, 95% CI 0.94–2.53). Contrary to expectations, the incidence of CV events tended to be higher in ticagrelor-treated patients as compared with those on clopidogrel, although confidence intervals were also wide (HR 1.47, 95% CI 0.88–2.44).14
| PHILO | PLATO | |
|---|---|---|
| East Asian patients, n (% of trial population) | 801 (100) | 664 (3.6) |
| Age ≥75 years (%) | 25.8 | 15.5 |
| Body weight <60 kg (%) | 38.2 | 7.0 |
| Hypertension (%) | 74.3 | 65.4 |
| Dyslipidemia (%) | 75.3 | 46.7 |
| Diabetes (%) | 34.7 | 25.0 |
| History of prior MI (%) | 8.0 | 20.5 |
| History of prior PCI (%) | 10.9 | 13.4 |
| History of non-hemorrhagic stroke (%) | 6.9 | 3.9 |
| Persistent ST-segment elevation on entry ECG (%) | 59.0 | 37.6 |
| ST-segment depression (%) | 42.6 | 50.9 |
| Positive troponin test at study entry (%) | 75.8 | 85.7 |
| Index event: STEMI (%) | 51.8 | 37.7 |
| β-blocker use (%) | 10.5 | 89.5 |
| ACEI use (%) | 16.4 | 75.6 |
| Statin use (%) | 52.4 | 89.5 |
| PCI for index event (%) | 84.6 | 64.3 |
| CABG for index event (%) | 1.5 | 10.2 |
| Major bleeding (CABG and non-CABG-related) (%) | 10.3 for ticagrelor 6.8 for clopidogrel HR 1.54, 95% CI 0.94–2.53 |
11.6 for ticagrelor 11.2 for clopidogrel HR 1.04, 95% CI 0.95–1.13, P=0.43 (Asian pts HR 1.02, 95% CI 0.70–1.49) |
| Primary efficacy endpoint (Composite of CV death/ MI/ CVA) (%) | 9.0 for ticagrelor 6.3 for clopidogrel HR 1.47, 95% CI 0.88–2.44 |
9.8 for ticagrelor 11.7 for clopidogrel HR 0.84, 95% CI 0.77–0.92, P<0.001 (Asian pts HR 0.84, 95% CI 0.61–1.17) |
| All-cause mortality (%) | 2.5 for ticagrelor 1.8 for clopidogrel HR 1.42, 95% CI 0.54–3.74 |
4.5 for ticagrelor 5.9 for clopidogrel HR 0.78, 95% CI 0.69–0.89, P<0.001 (Asian pts HR 0.77, 95% CI 0.51–1.17) |
ACEI, angiotensin-converting enzyme inhibitor; CABG, coronary artery bypass grafting; CI, confidence interval; CV, cardiovascular; CVA, cerebrovascular accident; HR, hazard ratio; MI, myocardial infarction; PCI, percutaneous coronary intervention; pts, patients; STEMI, ST-elevation myocardial infarction.
So the challenging question is whether any conclusions can be drawn from the results of the PHILO trial. Investigators and regulators were undoubtedly hopeful that the results for East Asia in PHILO would be consistent with those in PLATO. However, the unfavorable trend for efficacy with ticagrelor in the current trial should not dissuade clinicians from the use of ticagrelor in Japan or East Asia. Importantly, it is reassuring that the efficacy and safety profile of ticagrelor in Asian patients in PLATO appeared to be consistent with those of patients worldwide. Therefore, as suggested by the study’s investigators, it is probable that the current findings are explained by the relatively small sample size and chance imbalances in patient characteristics across treatment arms. In particular, patients treated with ticagrelor tended to be older and have several comorbidities that are associated with excess CV risk. It would therefore be a disservice if the results of the current study discouraged Japanese clinicians from prescribing ticagrelor, as many East Asian patients could benefit from a therapy that is not influenced by the CYP2C19 LOF allele.
More broadly, the results of the PHILO trial raise important questions regarding the utility of country-specific clinical trials if the sample size is not sufficient to adequately test the hypothesis. Ideally, registration pathway clinical trials for novel therapeutics should enroll adequate numbers of patients from different regions, in particular East Asia, in order to demonstrate consistency with the overall trial results. However, this strategy is not without peril, as inconsistent results may be observed in a region, perhaps due to play of chance.15 Further, regional requirements for clinical trials may contribute to soaring clinical trial costs and delay patient enrollment. Nonetheless, it remains critical that early dose-ranging and subsequent phase III trials reflect the drug’s intended prescribing patterns so that informed decisions can be made regarding its clinical use. Therefore, PHILO should serve as a call to broaden the diversity of trial populations, as inclusion of globally representative populations will increase confidence that the outcomes of studies apply to a more heterogeneous population.
M.L.O’D. has received grant funding from AstraZeneca, Eisai, Merck and GSK.
