Immortal Time Bias - Deceased Control Patients Can’t Be Hospitalized, page-25

The thing that interests me is.... why?

Leaving out heart attack, as that's a wildly complex topic on its own, I'm going to focus on strokes.

What is the mechanism behind stroke reduction? I think it would be good to know what sort of strokes were seen in the control group, and how prevalent they actually were. They must have been quite prevalent to have found significance with a population size such as it was.

People with heart failure are at risk of cardioembolic strokes for a number of reasons (where a clot shoots off from the heart, generally left side).

When you've got an ejection fraction of 20% there's a lot of blood hanging out in the left ventricle. This blood is at risk of forming clots, and those get shot up to the brain (and elsewhere in the body). This is why patients with certain levels of heart failure get put on blood thinners such as warfarin or apixaban.

If someone has an LVAD, they're also already on blood thinners (prevents clotting around the device).

As these patients were on maximal therapy already, I would have to assume most would be on blood thinners as is? Or, was the cohort of NHYA 2 patients not on these blood thinners yet? What was their ejection fractions? And are we properly normalising for people on blood thinners (were more of our therapy arm on anticoagulants or were they similar numbers).

Now people with heart failure are at risk of other types of strokes as well. As many heart failure cases are due to ischaemic heart disease, these patients are vasculopaths. They've got evidence of atherosclerosis, and it's safe to assume they've got it through their body, not just in the coronaries.

They're at increased risk of microvascular strokes, but these are often sub clinical (don't present as a stroke), and rather build up over time to cause cognitive impairment (this is a very simplified version of a complex topic). Are these the type of strokes prevented? Did they do imaging of the patients to see if there was evidence of this?

Atherosclerosis is an inflammatory disease. Presumably, based on the theorised mechanism of MSCs these could act on systemic inflammation to reduce formation of plaques or increase plaque stability?

But this was not a systemic infusion, like in ARDS. This was a localised injection. How much systemic response are we expecting from this local injection? Would they have had better results with an infusion?

Or are these results another example of a number of very complex disease processes that may or may not be affected by a very complex therapy with an unknown mechanism?

Just some thoughts. Don't have any answers for them. But these may be things that people on the FDA are thinking as well.