In fact, epinephrine probably has more evidence for its use than any other drug in ACLS for full arrest.
Which is to say: next to none. There has never been a randomized, placebo controlled trial into epi in cardiac arrest, and unfortunately due to entrenched practice there is not likely to ever be one.
In fact there is no strong evidence to suggest that any pressor provides benefit in cardiac arrest.
Epi, as mentioned, is given for it's alpha agonist effects. Increased peripheral vascular resistance improves intra-aortic pressures, which then increases coranary perfusion pressure, which is shown (in dogs) to be a strong predictor of ROSC in arrest.
It is not given for it's beta effects. It's beta effects increase myocardial O2 demand, increase the likelihood of ventricular arrythmias, increase intrapulmonary shunt and decrease the ability of the heart to relax (and therefore pump should ROSC occur).
There is a nasty downward spiral when giving epi in cardiac arrest: the longer the arrest continues, the more epi is required to maintain CPP and the worse the myocardial dysfunction is subsequently.
Epinephrine also exacerbates neurological dysfunction post arrest by causing decreased cerebral blood flow and globally increased oxygen demand that cannot be met by CPR.
This is why there is so much interest in vasopressin in cardiac arrest as it is pure alpha, no beta. Unfortunately this has yet to lead anywhere useful either and there is no strong evidence to suggest that one, the other or both are of any use.
We need to concentrate on good quality CPR with minimal interruptions; defibrillation; and post-resus care. Post Cardiac Arrest Syndrome is something that is being investigated further, and it would not surprise me if in the future we are not only cooling and haemodiluting whilst maintaining MAP (as we are now) but also adding things like low dose steroids to combat post-resus adrenal insufficiency.