Under regular conditions of plotting an ROC curve (in medical research), is there any circumstance where we get an AUC >1? If so, how do we interpret it?
As far as I know the values of AUC can be anywhere between 0 and 1 (perfect) not more, not less. I suggest that you recheck your analysis of it's more than 1.
By definition, no. You can't have more than 100% sensitivity and more than 100% specificity. It has to be between 0 (including 0, but that would be an awful predictor) and 1 (a perfect predictor).