120 retractions from one journal since June, and the number that matters more

Elsevier's International Journal of Biological Macromolecules has retracted 120 papers since June for manipulated peer review in guest-edited special issues, with at least one more to come, Retraction Watch reported on 12 August 2026. The retraction notices describe "[s]ystematic, coordinated, and widespread manipulation of the peer-review process."

The facts

An Elsevier spokesperson told Retraction Watch the publisher became aware in 2025 of possible systematic manipulation, along with plagiarism, image duplication and authorship problems, in articles published in special issues, and widened the investigation from there. The retractions "pertaining to this peer-review manipulation scheme are now almost complete," the spokesperson said, and guest editor checks have been tightened.

Two further numbers give the scale some shape. Nearly all of the retracted papers were published in 2024 and 2025, and Retraction Watch reports that nearly all listed authors with affiliations in China. And the journal's output grew steeply over the same period: from nearly 5,000 articles in 2023 to more than 9,600 in 2024 and more than 10,300 in 2025, with over 4,000 so far in 2026. Elsevier did not respond to Retraction Watch's question about the spike.

Why 120 is the less interesting number

A bulk retraction reads like a system working. It partly is. But two things in the reporting are more useful to a working researcher than the headline count.

The first is the yield. Fabian Wittmers, a postdoctoral researcher at Oregon State University, has been sending batches of concerning papers to this journal since April 2025 and has flagged 352, both to the journal and on PubPeer. Exactly one of them appeared among the 120 retracted for peer review manipulation, and 11 more were retracted for other reasons. That is roughly one retraction for every 29 papers a single person took the trouble to document. The manipulation was found, in the end, by a publisher-side investigation rather than by the reports.

The second is what happens next in the literature. These papers spent one to two years being read and cited before the notices appeared. A retraction notice fixes the publisher's page. It does not reach into the reference list of every review, meta-analysis or thesis that already cited the work, and it does not travel reliably across the bibliographic databases most people actually search from.

What this changes for a reader

Concretely: if you are citing work from this journal published in 2024 or 2025, especially from a special issue, check the current status of each item rather than trusting a stored PDF or a reference manager entry captured at the time. The Retraction Watch Database is free and is now maintained with Crossref, so this is a lookup rather than a research project.

More generally, this is an argument for treating a special issue as a different kind of object from a regular issue. Guest-edited collections have been the repeated site of these failures: Hindawi suspended special issue publishing in 2023 amid manipulated peer review, a Springer Nature journal retracted 34 special issue papers in 2024, and Annals of Operations Research retracted an entire special issue the same year. That is a pattern in the mechanism, not a run of bad luck at unrelated publishers.

Where the standard now sits

COPE's retraction guidelines were revised to version 3 in September 2025, and one of the additions was a section on batch retractions, on the basis that they "might be needed when there is evidence of systematic manipulation of the publication process in one or more journals." COPE also publishes a separate flowchart for assessing suspected paper mill activity. The guidance is catching up with a failure mode that is now routine enough to have its own procedure, which is itself the story.

What to watch

Whether the tightened guest editor checks Elsevier describes are published in any auditable form, and whether the journal's publication volume moves in 2026 and 2027. Volume is not evidence of anything on its own, but a journal that doubled its output in a single year and then retracted 120 papers from that period has given readers a reason to watch the number.