Interview with an eminent clinical-researcher: Michel Ovize

Can you outline for us the key steps of your scientific career?

I trained as an interventional cardiologist. In the mean time, I had an early interest in translational research. I had the opportunity to spend two years in RA Kloner’s lab in L.A. (1990-92), with K.Przyklenk as a mentor, in the early days of pre-conditioning.  When back in Lyon, France, along with my cardiologist activity, I progressively built a research team, including both a basic science group and a clinical group. Our main goal was to be able to transfer cardioprotective interventions to the patients. This is what we did in 2005 when showing for the first time in a phase 2 trial that ischemic (angioplasty) postconditioning was able to reduce infarct size in STEMI patients. We are now continuing this research with larger clinical trials, e.g. like the CIRCUS study that was presented at the ESC meeting in London.

What is, for you, the best definition of “reperfusion injury”?

It can occur in any organ submitted to a period of ischemia followed by reperfusion (heart, liver, kidney, brain…). Reperfusion injury can be reversible (e.g. myocardial stunning, arrhythmias) after a short period of ischemia or irreversible after a prolonged period of ischemia. Irreversible reperfusion injury is the death of cells that were still alive at the end of the ischemic period, that occurs after reflow, and that can be prevented by a protective intervention applied after the onset of the ischemic phase.

Are there currently therapeutic strategies to reduce reperfusion injury?

Many interventions have been shown to prevent reperfusion injury in experimental preparations, the “generic” ones being: preconditioning, postconditioning and remote conditioning. Several drugs can mimic this in animal models. Phase II trials have shown that some of these drugs might be able to reduce infarct size in acute myocardial infarction patients. Unfortunately, so far, none of these therapies have been shown to improve clinical outcomes in phase III trials.

You were among the first to demonstrate that cyclophilin D-induced mitochondria permeability transition pore plays a central role in reperfusion injury. How did you come to the “cyclophilin D” story?

It has been long known (since the early nineties) that cyclosporine, a potent inhibitor of CypD, can prevent reperfusion injury in experimental preparations. We could show, together with some other labs, that cyclosporine can inhibit mitochondrial permeability transition pore opening and reduce infarct size when administered after reperfusion in in vivo preparations. This was reinforced by the demonstration that CypD-KO mice develop smaller myocardial and cerebral infarcts after a prolonged ischemia.

You conducted a first pilot clinical study targeting cyclophilin D in acute ischemia with the immunosuppressant cyclosporine A.  What were the obstacles in translating preclinical ischemia reperfusion protocols to the clinic?

The major obstacle we all face is that the process to set up a clinical trial is very slow and requires large amounts of money. But this first pilot trial was (to me) surprisingly easy, probably because the rationale was quite strong, cyclosporine was already used in clinical practice, and the benefit/risk ratio was a priori good.

This first clinical study gave very positive and encouraging results, whereas the second, larger, clinical trial called CIRCUS (does Cyclosporine ImpRove Clinical oUtcome in ST elevation myocardial infarction?) was neutral?  What are your explanation(s) for that?

This was a surprise and of course a disappointment. This is unfortunately a common observation for treatment of acute MI. The first possibility is what we classically call a type I error. It means that the first study, because of its small size (n=60 patients), possibly selected a group of “good responders to cyclosporine”. This phenomenon is highly unlikely when using larger population like in CIRCUS (n=970). While I can admit that cyclosporine would not impact clinical outcomes, the very puzzling observation of CIRCUS is that we did not observe any effect of cyclosporine on infarct size. I strongly believe that cyclosporine is efficient due to so many demonstrations in animal models by several labs. And we have the endogenously protected CyPD-KO mice. One possibility is that experimental works are, by design, exaggerating the positive effect of this drug. Another (more likely to me) possibility would be that, for some unknown reason, cyclosporine did not reach its mitochondrial target quickly enough to attenuate reperfusion-induced myocardial damage. Indeed, the time window for protection is quite narrow in these clinical conditions, and different confounders might modify the pharmacokinetics of cyclosporine. Another option is that ischemia-reperfusion could modify (through post-translational modifications) the molecular target (cyclophilin D) of cyclosporine after several hours of ischemia, so that this drug would no longer access its site of activity.

Do you think that CIRCUS will close the door to cyclophilin D as a cardioprotective target?

Not necessarily. Cyclosporine is a nonspecific inhibitor of CypD. It is possible that more specific agents will do better in the future. We obviously need to better understand : 1) the exact role of CypD in PTP and non-PTP functions, 2) the molecular composition and regulation of PTP opening, 3) other phenomena involved in reperfusion injury. It might well be naïve to imagine that a single drug, targeting a single signaling pathway or mechanism, would be able to limit the consequences of a phenomenon as complex as ischemia-reperfusion.

Why have attempts to translate novel cardioprotective strategies from preclinical studies to the clinic so often disappointed?

Several reasons:

• experimental models are (consciously or unconsciously) designed to show what we expect to show, thereby emphasizing all the positive effects of a given intervention
• experimental models are always highly controlled: e.g determinants of infarct size, duration of ischemia, animal strains, etc..
• nothing related to ischemia-reperfusion injury is controlled or even measurable in clinical settings, so there is a much higher degree of variability. Although the pathophysiology is likely very comparable to that of animal models, the underlying disease condition(s) (e.g. hypertension, diabetes,..) and co-treatments may impact ischemia-reperfusion injury and influence the efficacy of the tested intervention
• although there is a strong link between infarct size and adverse clinical events occurring after acute MI, there is a large gap between clinical events and surrogate endpoints.

What is the near future in cardioprotection?

Despite recently reported negative trials, the future of cardioprotection is still positive. There is compelling experimental evidence that reperfused tissues can (and must) be protected. We probably need to build more realistic animal models, try to better control clinical conditions, only test drugs/interventions with very strong preclinical demonstration of efficacy, and possibly combine different protective intervention to hopefully get a positive effect.