Review of the PRAMI Trial
Randomized Trial of Preventive Angioplasty in Myocardial Infarction
Background: The COURAGE trial was published in 2007. It compared up-front PCI to medical therapy alone in patients with stable CAD. Preventive PCI did not reduce the chance of dying or having a heart attack over a median follow up time of 5 years. The results rocked the cardiology world because for years prior to the publication of COURAGE, the standard of care called for revascularization of obstructive coronary stenosis. Despite what we would consider minor criticisms of COURAGE, the results have held over time as a preventive PCI strategy has failed repeatedly to reduce death or MI compared to medicine alone in subsequent large trials (BARI 2D, FAME 2, ISCHEMIA and ISCHEMIA-CKD) involving patients with stable CAD.
But what about patients with acute coronary syndromes who have, a clearly defined “culprit” lesion and stable coronary stenosis of a non-infarct vessel? On the surface, the answer might seem simple - treat the “culprit” lesion with PCI and leave the stable disease alone. Continue optimal medical treatment of stable CAD indefinitely with consideration of revascularization only if new symptoms arise.
But what if a stable coronary stenosis behaves differently in a patient with an acute coronary syndrome than in patients without it? Are these patients predisposed or particularly susceptible to acute plaque rupture and thrombogenesis to such an extent that they would benefit from a preventive revascularization strategy?
The Primary Angioplasty in Myocardial Infarction (PRAMI) trial sought to test the hypothesis that immediate preventive PCI of non-culprit vessels plus the culprit vessel compared to culprit vessel only PCI would improve outcomes in patients with a STEMI and coronary stenosis of a non-infarct related artery.
Patients: From 2008 through 2013, patients were enrolled from 5 coronary care centers in the United Kingdom. Patients could be any age with acute STEMI and multivessel CAD detected at the time of emergency PCI. The trial was limited to patients with STEMI because ST-segment elevation, unlike ST-segment depression, localizes the area of ischemia in the myocardium and an “infarct-artery” is usually easy to distinguish.
Clinically stable patients were considered for eligibility after undergoing PCI of the infarct artery while they were in the catheterization lab. They were eligible if successful PCI of infarct artery was performed and there was stenosis of 50% or more in one or more non-infarct arteries. Exclusion criteria included cardiogenic shock, previous CABG, had left main or significant disease in the ostia of both the LAD and circumflex vessels, or if the only non-infarct stenosis was a chronic total occlusion.
Baseline characteristics: The trial screened 2,428 patients and randomized 465 patients (19%) with 234 to preventive PCI and 231 to no preventive-PCI. The majority of patients were excluded for single vessel disease (1122/1922 [58%]). The average age of patients was 62 years and more than 75% were men. Close to 50% were current smokers. The infarct artery was anterior in 35%, inferior in 60% and lateral in 5%. Approximately 65% of patients had 2 vessel disease and 35% had 3 vessel disease.
Procedures: After completion of PCI in the infarct artery, eligible patients were randomized and those assigned to the preventive-PCI group underwent the procedure immediately in all non-infarct arteries with a coronary stenosis >50%. PCI was discouraged at a later date (sometimes this strategy is referred to as “staged PCI”) in the no preventive-PCI group unless it was symptom driven. Any patient in the trial with subsequent symptoms of angina that were not controlled with medicine was required to undergo objective assessment of ischemia to secure a diagnosis of refractory angina. Follow-up information was collected at 6 weeks and then yearly thereafter.
Endpoints: The primary endpoint was a composite of death from cardiac causes, nonfatal MI, or refractory angina. Secondary outcomes included the individual components of the composite endpoint along with noncardiac death and repeat revascularization.
Myocardial infarction was defined as symptoms of cardiac ischemia and a troponin level >99% URL. However, within 14 days after randomization, MI diagnosis also required ECG evidence of new STE or left bundle branch block and angiographic evidence of coronary artery occlusion (essentially this makes it so only in-stent thrombosis or spontaneous STEMI count and other causes of peri-procedural MI do not - this would bias the trial in favor of the preventive-PCI group).
Refractory angina was defined as angina despite medical therapy and objective evidence of myocardial ischemia (i.e., ischemia on ECG during spontaneous episode of pain or abnormal results on functional testing).
It was determined that 600 patients would be needed to achieve 80% power to detect a 30% relative reduction in the preventive-PCI group, at a 5% level of significance, assuming an annual rate of the primary outcome of 20% in the control group. Stopping criteria were prespecified if the results from the trial showed a primary outcome difference at the 0.001 level of significance.
Results: The trial was stopped early based on a significant difference (P<0.001) in the incidence of the primary outcome favoring the preventive-PCI group after the 2nd interim analysis was performed. At the time the trial was stopped, 465 patients were enrolled with 234 in the preventive-PCI group and 231 in the control group. Patients in the preventive-PCI group had an average of 1.29 stents placed in non-infarct arteries and these were primarily drug eluting stents (71%).
Compared to no preventive-PCI, preventive-PCI significantly reduced the primary outcome (9.0% vs 23%; HR 0.35; 95% CI 0.21-0.58) along with the individual components of nonfatal MI (3.0% vs 8.7%; HR 0.32; 95% CI 0.13-0.75) and refractory angina (5.1% vs 13.0%; HR 0.35; 95% CI 0.18-0.69). Cardiac death was also numerically reduced with preventive-PCI but did not reach statistical significance due to low power.
Conclusions: In patients with STEMI and non-infarct artery coronary stenosis >50%, preventive PCI significantly reduced a primary composite outcome of cardiac death, nonfatal MI and refractory angina in the PRAMI trial with an estimated NNT of 7 patients over 2 years. Individual components of the primary endpoint that were significantly reduced included nonfatal MI and refractory angina by similarly large margins.
These results may seem impressive at first glance but we urge extreme caution in their interpretation. First, this is a relatively small trial with a historically large effect size, especially when considering hard endpoints like cardiac death and nonfatal MI were included. Such results are often later found to be falsely positive when larger, confirmatory studies are conducted. Second, the trial was stopped early and early stopping is prone to yield false positive and/or exaggerated results. Third, inclusion of refractory angina in the primary endpoint, an endpoint susceptible to bias in an unblinded study (see earlier discussion of “faith healing” and “subtraction anxiety” in FAME 2; consideration also must be given to nocebo effects in patients who know they have “untreated blockages”), clouds the main findings by inflating the effect size and making the trial susceptible to large differences in underpowered endpoints before sufficient data can be accumulated on hard outcomes. For example, if the trial had sought to detect a conservative difference of 30% in a primary composite endpoint that only included cardiac death or nonfatal MI, based on an event rate of 12% in the control group (the actual event rate in the trial), over 2,200 patients would be needed for 80% power at a 5% level of significance. The estimated number of actual events would be around 230. However, only 47 events occurred in PRAMI making the results highly susceptible to noise.
While results of PRAMI suggest a beneficial role for preventive-PCI in patients with STEMI, more evidence is needed to confirm the results.
Thank you. More please.
great stuff.. thank you