Studies were undertaken by Yellon's laboratory to determine whether human muscle can be pre-conditioned and whether there is a role for the adenosine receptor. Atrial muscle from human right atrial appendage of patients undergoing cardiac bypass surgery was placed in an organ bath. The trabeculae were attached to a fixed post and the other end to a force transducer to measure functional recovery. The human muscle was directly subjected to a preconditioning protocol of 3 minutes of hypoxia and 7 minutes of re-oxygenation prior to a lethal ischemic insult of 90 minutes.

Similar signalling pathways to that seen in the experimental models were seen in studies in Yellon's laboratory focusing on protein kinase C (PKC). The PKC inhibitor chelerythine was given to ascertain whether any observed effect can be abolished. In addition, DiC8, a PKC analogue, was given to stimulate PKC. In human muscle, 30% functional recovery was regained after 90 minutes of hypoxia and 2 hours of re-oxygenation in the control (Fig. 2). About a 70% functional recovery was regained after preconditioning with 3 minutes of hypoxia and 7 minutes of re-oxygenation. When DiC8 was substituted for preconditioning to activate PKC about 50% functional recovery was obtained. The protective effect was abolished with chelerythine.

That K_ATP was shown to be downstream of PKC was observed when DiC8 was given to activate PKC and this was blocked by the K_ATP channel blocker. Preconditioning and cromakaline provided good functional recovery, about 62% and 58%, respectively. DiC8 alone provided functional recovery of about 50%. However, the protective effect was abolished with DiC8 was given in the presence of glibenclamide.

Opioid receptors have also been shown to play a role in triggering preconditioning in animal hearts. A number of receptors, including the mu, delta, and kappa, exist in the heart. The distribution of opioid receptors in the human heart favors the mu and delta subtypes. Using human atrial and ventricle muscle Yellon's group was able to show that in both muscles the mu and delta receptors were expressed abundantly, compared to the kappa receptor.

Studies in low risk patients undergoing coronary artery bypass grafting (CABG) was done by Yellon's group to determine whether patients undergoing CABG needed additional myocardial protection and, if so, whether preconditioning can be of any benefit either prior to, or during, bypass surgery. Cardioplegia and cross clamp fibrillation are used to protect the myocardium during CABG. However, there are problems with both. Cardioplegia is not uniformly distributed and the time of cross clamp fibrillation is limited. Both techniques also involve whole body cooling.

Patients were randomized to preconditioning (n=17) or control (n=16) groups. The preconditioning group received 3 minutes of cross clamping of the aorta followed by 2 minutes of reperfusion followed by another 3 minutes of cross clamping. The control patients did not receive this preconditioning stimulus. All patients then underwent 3 separate periods of cross clamp fibrillation (10 minutes) during which the graft was sown on.

Troponin T was used to measure tissue necrosis; blood samples were examined after 72 hours. The patients in the control group had significantly greater troponin T levels compared to the preconditioning group (about 1.4 µgm/ml vs about 0.2 µgm/ml; p<0.05) (Fig. 3). This study provided proof of concept that preconditioning in patients undergoing CABG results in a direct benefit in terms of less troponin T.

To evaluate the effects of adenosine in preconditioning before bypass surgery, a blinded randomized study using an experimental adenosine A₁ receptor agonist (GR79236x) was performed. Patients were randomized to four groups (10 per group): control, cold cardioplegia, preconditioning (2 x 3-minute bursts of ischemia, each followed by reperfusion), and an adenosine A₁ receptor agonist (given over 10 minutes followed by wash out).

There was an indication that the adenosine A₁ receptor agonist was having some beneficial effect. The level of troponin T release was similar to that in the preconditioning group, although this did not reach statistical significance. Whether the study was underpowered is unclear. In the control group, there was a significant increase in troponin T compared to the preconditioning group, showing again preconditioning confers protection by reducing troponin T release. The troponin release was similar in the cardioplegia and control groups.