Does PEGASUS Have the CHARISMA to Change Practice?

The results of the PEGASUS-TIMI 54 (Long-Term Use of Ticagrelor in patients with Prior Myocardial Infarction) trial were presented to a capacity audience at the opening session of the March 2015 American College of Cardiology meeting.  But after hearing the much anticipated results and reading the simultaneously published manuscript we’re left wondering if PEGASUS is a landmark clinical trial or simply a trial that validates what we have known for years.¹

To understand the PEGASUS trial, we need to flashback to 2006 and the CHARISMA (Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization) trial.²  The CHARISMA trial randomized 15,603 patients with either risk factors for or clinically evident cardiovascular (CV) disease — including coronary artery disease (CAD), peripheral arterial disease (PAD), and cerebrovascular disease — to receive dual antiplatelet therapy with clopidogrel 75mg and aspirin 75-162mg daily or placebo and aspirin.²  After a median follow-up of 28 months, the rates of the primary composite endpoint (CV death, non-fatal MI or non-fatal stroke) were similar in the dual antiplatelet therapy and aspirin alone arms, thus questioning the benefit of DAPT in stable patients.  However, there was significant heterogeneity in the results when comparing patient sub-populations.  The CHARISMA investigators later reported that in the 9,478 patients with prior MI, ischemic stroke, or symptomatic PAD, the primary endpoint was observed in significant fewer patients treated with dual antiplatelet therapy when compared to aspirin alone (7.3% versus 8.8%; Hazard Ratio (HR) 0.83; 95% CI 0.72-0.96) albeit at a cost of increased moderate bleeding (2.0% versus 1.3%; HR 1.6; 95% CI 1.16-2.20). ³  There was no difference in severe bleeding. The investigators appropriately tempered their recommendations by stating that some patients with stable atherosclerotic disease might benefit from intensive antiplatelet therapy but “future trials will need to validate” the results of this subgroup analysis.  Unfortunately, future trials with clopidogrel and aspirin never materialized, leaving clinicians with a promising hypothesis but no sound evidence to apply in practice. 

That brings us now to the PEGASUS trial which gives clinicians much needed insight into the role of intensive antiplatelet therapy in patients with stable CAD following a myocardial infarction.¹  Patients were included in the PEGASUS trial if they were 1-3 years post-MI and had specific high-risk characteristics (age 65 years or older, diabetes mellitus, second prior spontaneous MI, multi-vessel CAD, chronic renal dysfunction).  Patients were excluded if the provider planned to extend P2Y12 receptor antagonist therapy or had a history of bleeding, ischemic stroke, or intracranial bleeding. Patients were randomized to ticagrelor 90mg, ticagrelor 60mg, or placebo dosed twice daily.  All patients received aspirin 75-150mg once daily.  The primary endpoint was similar to that in the CHARISMA trial, a composite of CV death, MI or stoke. The trial was powered based on an anticipated 20% difference in event rates in the 90mg ticagrelor and placebo arms.  Authors estimated that 1,360 events would be needed to provide 90% power.  

A total of 21,162 patients were randomized and followed for a median of 33 months.  The primary composite efficacy endpoint was observed in 7.85%, 7.77%, and 9.04% of patients randomized to ticagrelor 90mg, ticagrelor 60mg, and placebo, respectively.  This equates to a statistically significant 15% – 16% relative risk reduction with ticagrelor plus aspirin versus aspirin alone (see table 1).  These positive findings were offset by significantly higher rates of TIMI major bleeding in the ticagrelor groups (see Table 1).

Table 1:^1,3  

The efficacy and safety of intensive antiplatelet therapy versus aspirin alone for extended treatment in post-MI patients:  PEGASUS and CHARISMA studies compared

^a Values presented as hazard ratios (95% confidence intervals) for intensive antiplatelet therapy versus aspirin alone

^b Number Needed to Treat (NNT) or Number Needed to Harm (NNH)

^c Rates of cardiovascular death, MI or stroke are derived from the post-MI cohort of the CHARSIMA symptomatic cohort analysis; bleeding rates are for the CHARISMA symptomatic cohort as a whole (not reported in post-MI group)

^d TIMI major bleeding: fatal bleeding, intracranial bleeding, clinically overt signs of hemorrhage associate with drop in hemoglobin of ≥5g/dL or hematocrit of ≥15%

^e GUSTO severe bleeding: fatal bleeding, primary intracranial hemorrhage or bleeding that causes hemodynamic compromise and requires blood or fluid replacement, inotropic support or surgical intervention

^f TIMI minor bleeding: Any clinically overt sign of hemorrhage (including imaging) that is associated with a fall in hemoglobin of 3 to <5 g/dL (or when hemoglobin is not available, a fall in hematocrit of 9 to <15%)

^g GUSTO moderate bleeding: Bleeding requiring transfusion of whole blood or packed red blood cells without hemodynamic compromise

PEGASUS seems to confirm the CHARISMA subgroup results.  More intense antiplatelet therapy in stable, post-MI patients reduced CV death, MI and stroke but at the cost of increased major bleeding.  To put these findings into perspective we need to examine a few clinically important issues.

Investigators powered PEGASUS to find a target 20% risk reduction in event rates⁴;  however, the observed RRR was only 15% with the 90mg dose of ticagrelor.  Because the number of events in the trial was greater than anticipated (1,558 versus 1,360), statistically significant differences were still realized.  But are these results clinically important?  If we treated 10,000 patients with ticagrelor 60mg twice daily for 1 year, we’d prevent 42 primary efficacy events and we would cause approximately 31 major bleeds.⁵   Similar NNT and NNH numbers are seen with ticagrelor 90mg twice daily versus placebo (see Table 1).

A second point to consider:  do the harms essentially offset the benefits?  Here clinicians and patients will need to have a candid discussion.  CV death, MI or stroke are likely more important outcomes to most patients when compared to major bleeding events.  But the bleeding events reported in Table 1 are not minor nuisances, such as mild bruising or gum bleeding while brushing your teeth.  Major bleeds can cause substantial harm — including, in the most extreme cases, permanent disability and death.  It is important to point out that ticagrelor did not impact CV or all-cause mortality, which may be the most important factor when considering use of an additional therapy for secondary prevention. 

These points are not intended to devalue the findings of the PEGASUS trial but should serve to temper any knee-jerk reaction to routinely place all patients with stable CAD and a prior-MI on ticagrelor in combination with aspirin.  Instead, it should prompt individualized discussions between providers and patients.

Clinicians, policy makers, and clinical practice guideline committees will face another important question.  If the statistically positive findings from PEGASUS validate the post-hoc subgroup data from CHARISMA, does clopidogrel therapy offer a credible benefit when combined with aspirin in patients with stable CAD for more than one year after an MI.³ This is an important question because clopidogrel appears to have a more palatable risk profile than ticagelor when comparing the results of CHARISMA and PEAGASUS (see table 1). Of course this comparison is not a direct one.  But it’s not like we are comparing apples and oranges.  The similarities between the post-hoc subgroup CHARISMA evaluation and PEGASUS is like comparing Macintosh and Red Delicious apples.  If one believes the benefits are similar and the risks are lower, coupled with generic availability and a lower price, clopidogrel becomes an attractive choice over ticagrelor for long-term intensive antiplatelet therapy in the post-MI setting.^1,3

So, is it time to use of long-term dual-antiplatelet therapy in patients beyond 1 year post-MI? Given PEGASUS and the DAPT trial (see the companion iForumRx commentary), the answer is “yes”.  However, “only in carefully selected high risk patients in whom the benefit of therapy will likely outweigh the risk.”  Unfortunately for clinicians and patients, there are no established criteria to aid in weighing this clinical decision.  Has the results of PEGASUS-TIMI 54 changed your recommendations for patients with stable CAD following an MI?  Will you use clopidogrel over ticagrelor when using extended dual antiplatelet therapy?