If a structure already exists a new submission for that structure (i. e. same isomer) will take the old drawing. In some cases this is a problem, e.g. if the new structure has explicit hydrogens. One could argue that a drawing should always be complete, e. g. use wedges and therefore no different drawing is needed, but this may not be realistic.
In this case, the choice which drawing to use (new or old) should be at the reviewer's discretion.