I've seen second editions of philosophical articles. For example, Quine's From a Logical Point of View is a collection of articles, some of them updated. But I have never seen a second edition of an article in journals. Are books the only option for this?
Here are the options I've seen:
I definitely agree with Bryan's suggestion like "a 15-year follow up on XYZ" or a review/survey paper.
You can sometimes republish the 2nd edition of the paper in another journal, often because your previous paper had an impact, or the original journal was terminated and thus very hard to be obtained by researchers. This serves the need of other researchers and could be done if you arrange this with publishers.
Republication in another journal in a distantly related field, in cases your paper has multi-disciplinary values. Be sure to check with the publisher's policy: you might need to significantly update your paper.
Republication in a book or paper collection. This is the easiest way.
Republication in a conference proceeding. Be careful that some conferences allow paper published elsewhere while many do not allow.
Depends on how much you updated the original paper, here are some addition options:
Addendum to the original paper in the same journal
Corrigendum to the original paper in the same journal
Comment, note, or correspondence to the original paper in the same or another journal. Be careful that some journals accept comments or notes, some do not.
Papers are certainly more one-off than books; they are meant to be a permanent archival record of some study.
However, some studies, especially long-term longitudinal studies, will produce multiple papers. Some of the data are shared paper to paper, and the new combined dataset is just published as a new paper. It needs not be in the same journal, but typically will have some reference to the ongoing nature in the title, something like "a 15-year follow up on the XYZ study".
The other case is in review papers or theoretical position papers, where one person may summarize the state of the field and/or their position on some open problem, and then do so again in the future. However, they won't make this an "edition 2" of the previous publication, instead it will stand on its own even if it references the older summary.