Why not try a Wolff-kishner-reduction, using hydrazine? It oughta get that ketone out of there, replacing it with a hydrogen atom. The one possible problem would be if the meth-CAT will react with another mol of meth-CAT (reductive amination)...
Perhaps one can protect the nitrogen atom using Ac2O, for instance, though the yield probably sinks dramatically.
-Stanley