Tailoring anti-platelet therapy – one size will not fit all

Keywords::

• Anti-platelet treatment
• Bleeding complication
• Prasugrel
• Tailoring treatment
• TRITON trial

The economist Milton Freidman’s observation that “there is no such thing as a free lunch” can be applied to anti-platelet therapy as appropriately as it was to 20th century economics. In current practice no clinical trial has as yet shown a reduction in ischemia-driven major adverse coronary events (MACE) without an increase in hemorrhagic side effects. For example the Antithrombotic Trialists Collaboration demonstrated a 60% increase in extra-cranial bleeding in patients treated with aspirin vs. placebo for secondary prevention in high-risk patients1. Similarly in the landmark CURE trial of clopidogrel after an acute coronary syndrome, clopidogrel use was associated with a 69% increase in bleeding compared to placebo2. In both trials, patients with a history of stroke seemed particularly prone to the bleeding complications of potent anti-platelet drugs. Indeed subsequent analyses of bleeding after an acute coronary syndrome have repeatedly identified a history of stroke along with age, female sex, weight, hypertension and renal function as risk factors for such complications.

Therefore it came as no surprise that the TRITON trial revealed that although the overall population derived a 19% reduction in MACE compared to clopidogrel, there was a significant excess of bleeding which was particularly prevalent in patients over 75 years, under 60 kgs and with a history of stroke3. The authors subsequently presented what was termed a ‘core cohort’ in which these three groups of at-risk patients were excluded demonstrating not only an improvement in MACE reduction (26% vs. 19% in the full cohort) but a reduction in the excess bleeding risk associated with prasugrel vs. clopidogrel4. In a similar fashion, the PLATO trial demonstrated an almost identical excess of bleeding to that observed in the TRITON trial (27%) in patients after an acute coronary syndrome treated with ticagrelor vs. clopidogrel5. However in the PLATO trial there was no clear group that seemed to be at a particularly higher risk of bleeding with ticagrelor. Therefore in clinical practice it might be harder to predict which patients were more likely to have a bleeding complication when treated with ticagrelor than with clopidogrel.

In the current issue of Current Medical Research and Opinion Wilcox et al. present a further analysis of the TRITON study restricting follow-up to 12 months rather than the original trial primary end-point of 15 months (performed because of the licensed treatment duration)6. This earlier time-point analysis demonstrates a further improvement in MACE compared to the original primary end-point reduction in what is now termed ‘the license population’ (previously the core cohort). However, more importantly the bleeding excess seen in the original time-point analysis has now all but disappeared. It was entirely predictable that an earlier cut of the data would reduce the difference in bleeding risk observed between prasugrel and clopidogrel as a landmark analysis previously published from the TRITON trial demonstrated a very gradual accumulation of risk over time and the bleeding risk curves diverging only slowly7. Therefore what these subsequent analyses of TRITON very importantly demonstrate (which is very much in keeping with all previous and subsequent anti-platelet trial data) is that certain groups benefit more than others from potent anti-platelet therapy and that the duration of therapy is also crucial in determining outcome.

What is clear therefore from these data is that there are winners and losers with anti-platelet treatment. The ADEPT registry fascinatingly demonstrated that whilst patients who were non-responders to clopidogrel did indeed experience a higher rate of stent thrombosis after percutaneous coronary intervention than responders, the rates of major bleeding were lower than observed in clopidogrel responders. The net result of the divergent rates of these adverse outcomes between the two groups was that there was no difference in mortality. It would therefore seem highly plausible that routine use of potent anti-platelet drugs for all (as ticagrelor has been promoted) in a clinical population is likely to do as much harm as good, i.e. all the MACE reduction will be offset by harm induced by bleeding excess.

So how are physicians to put these data into clinical practice? Routine platelet function testing as evidenced by ARCTIC, TRIGGER and GRAVITAS appears to be of little clinical use although on-going trials with more potent anti-platelet are ongoing. Until these trials are completed the available data should encourage clinicians to select patients based on either a formal or informal bleeding assessment. Indeed the ‘license cohort’ analysis presented in the current issue demonstrates that simply avoiding patients under 60 kgs, over 75 years and without a history of stroke can almost completely negate the bleeding excess seen with prasugrel and provide maximum ischemia benefit. Additionally newer stent technologies such as the Biofreedom (Biosensors SA), Synergy (Boston Scientific) and Combo (Orbus Neich) might provide an opportunity to shorten the duration of dual anti-platelet therapy further. Furthermore the Global Leaders trial (which is still recruiting) questions the whole premise of dual anti-platelet therapy after stenting. However, before these data are available the current evidence base strongly supports clinicians ‘tailoring’ anti-platelet therapy to provide maximum benefit and minimize harm.