Editors at the New England Journal of Medicine have retracted a paper published in 1998 against the wishes of the authors because it is believed to contain a doctored version of a figure published by another group, in another journal eight years earlier. The retraction appeared in the 11 July issue of NEJM, accompanied by the editorial note, “because of the unmistakable similarity between the figures (the cellular aspects of Figure 1 and Figure 2 are fully superimposable in transparencies), we must retract publication of the article by Barbaro et al.”

The work, by Giuseppe Barbaro at the University la Sapienza in Rome, focuses on dilated cardiomyopathy in HIV-infected patients and presence of the virus in heart cells. The figure in question is an in situ hybridization of an HIV RNA probe in a heart biopsy from an infected patient. The HIV detection method used was the same as that in the earlier paper by Grody et al. published in 1990 in the American Journal of Cardiology—one of the first to document HIV infection in heart cells. One of the authors of the Grody paper alerted NEJM editors to the similarity in February.

Barbaro and co-authors have since submitted a letter to NEJM saying that they “do not agree with a retraction” and “strongly defend the scientific value of the article.” Barbaro has accused the editors of taking their decision without proper consultation. “I asked that those involved could meet with external experts named by all parties,” Barbaro told Nature Medicine, “so that the materials could be validated.” However, NEJM executive editor, Gregory Curfman, said the editors “conducted [their] own investigation using a consultant outside of the journal's office.” According to Curfman, the figure was inverted top-to-bottom and side-to-side and pixels were added, suggesting increased HIV RNA, that were not in the original figure, indicating to the editors “intentional doctoring.”

It has long since been known that HIV patients have increased risk of cardiomyopathy due to viral infection and inflammation. “For those of us in the field,” says Steven Lipshultz, an expert on the subject who has published a several papers with Barbaro, but not the one in question, “the beauty of the [Barbaro et al.] study was the size and design.” The study followed 952 asymptomatic HIV-positive individuals for a mean period of over 5 years. “But the findings have been confirmed over and over again in adults and children,” says Lipshultz. There are more than 2,000 papers on HIV in heart disease. “It is hard to know what to do with this particular one,” he says.