Is it possible to withdraw authorship almost 2 years after the paper has been published in a journal? Reasons could be anything ranging from disagreement on data interpretation, new data contradictory to the original one, etc.
I wouldn't retract authorship, science is all about revising research based on new data or a better explanation. I'd suggest publishing a new paper, and explaining what you disagree with or have found better interpretations of from the old paper. A withdrawal of authorship (even if possible) would be interpreted as some form of manipulation of data in the prior paper.