Faiez Zannad moderates a fascinating discussion between Milton Packer and Steve Nissen about the PARAGON-HF trial in heart failure with preserved ejection fraction. Nissen was on the FDA advisory panel for the sacubitril/valsartan label.
Nissen: “Part of the problem is the metric that we’re using, which is echo ejection fraction, is not a very precise technique” “We ought to be reading echos as showing normal, mild, moderate and severe, put it in those buckets”
Packer: “Clinical trialists really actually like to see p values with multiple zeros after the decimal place”
Then they discuss TOPCAT and the sites in Russia and Georgia. The patients there seemed different; adherence was low and event rates were low. Did they even have the disease at all? Nissen voted no on spironolactone for heart failure with preserved ejection fraction. Nissen argues you can’t throw away data; it’s about intention-to-treat. What can we learn from this trial despite what Packer calls, “execution failure”?
Nissen: “You don’t get to decide that some patients you randomize were good patients and some were bad”
Packer: “Quality control in this trial failed. And it didn’t fail at one site. It failed on a very significant scale” “Do I believe in the result in the Americas? I do. But just because I believe in it, doesn’t mean that it is actionable”
They also touch upon standards – what about COVID trials and oncology trials.