Why press releases don’t tell the whole storyBMJ 2022; 379 doi: https://doi.org/10.1136/bmj.o2938 (Published 06 December 2022) Cite this as: BMJ 2022;379:o2938
It’s good to see that the full data from the trial of lecanemab for Alzheimer’s disease have now been published in a peer reviewed journal.1 When the results were first announced by Eisai and Biogen in September it was via an “investors relations” press release.2The BMJ did not report on these “results” because the journal has a new policy that means it won’t report on press releases that are not supported by information that allows proper scrutiny, such as a journal study report or a detailed research summary.3
But elsewhere in the media the pick up was immense. An internet search on lecanemab in Google News on 29 November, just before the study was published in the New England Journal of Medicine, generated over 39 000 links. Speculation about the benefits of lecanemab was captured in headlines even in highly respected news outlets, such as “Alzheimer’s-slowing drug labelled historic” from the BBC and “Alzheimer’s drug slows cognitive decline in key study” from the New York Times.
Now that the full paper is available it’s possible to report on the size of the effect of the findings and the data on side effects.4 We learnt that more than twice as many patients treated with lecanemab left the trial because of adverse events (6.9%) as in the placebo group (2.9%). Most importantly, the data showed that lecanemab made a tiny difference to cognitive deterioration on cognition scales over 18 months, which although statistically significant may not be clinically significant in that it isn’t evident to patients, carers, or doctors. This is all important information that helps contextualise the possible future use of lecanemab.
Eisai said, “The phase 3 trial results of the Clarity AD study is material information to Eisai’s investors. Regardless of the results, this information must be disclosed shortly after it is known per the rules of the stock exchange. According to Eisai’s corporate disclosure policy, we announced the topline data of the Clarity AD study in a transparent and timely manner in late September.”
Since The BMJ launched its policy to no longer report on unsubstantiated press releases the news desk has recorded two other occasions where press releases with trial data have preceded peer reviewed publication in a journal or as a preprint.
On 1 November Pfizer sent out a press release about results from the Matisse (Maternal Immunization Study for Safety and Efficacy) study, a phase III international trial of its bivalent respiratory syncytial virus (RSV) vaccine candidate for maternal immunisation.5 The Times ran the headline “Vaccine for mothers could save thousands of babies” and hailed a “game changing” vaccine. Because no underlying data were supplied in the press release The BMJ didn’t cover the story other than to refer to the study in a longer piece on the UK Medicines and Healthcare Products Regulatory Agency’s approval of nirsevimab, a long acting monoclonal antibody, to protect newborn babies against RSV.6
When The BMJ asked Pfizer why it thought it appropriate to release data on its RSV vaccine by press release, a spokesperson said, “There was a planned interim analysis and the topline data was considered material to the company. There is more data being collected and will be released via presentation or peer-reviewed paper.”
The most recent example of trial results being reported by press release only happened on 28 November. Cambridge University Hospitals NHS Foundation Trust announced findings from the Heal-Covid (Helping to Alleviate the Longer-Term Consequences of Covid-19) trial, an NHS funded platform or open ended trial that is aiming to find the best treatments to reduce the number of people who die or are readmitted after time spent in hospital for covid-19.7 The trial’s first finding, involving the anticoagulant apixaban, showed that it made no difference to the “number of days alive and out of hospital at day 60 after randomisation,” the press release said.
The Cambridge University Hospitals trust did not respond to a request for a comment.
Launching the new policy in July, Kamran Abbasi, The BMJ’s editor in chief, warned, “The press release has become the propagandist’s tool.” He added, “We saw it regularly at the height of the pandemic: grand announcements heralding successful new treatments and strategies, without a research paper or data analysis to support the hyperbolic claims.”
The use of press releases rather than the full data is an abuse of power, said Abbasi, that “minimises criticism and amplifies the messaging in the press release. It’s a mode of communication that serves companies and their shareholders and not patients or the doctors trying to treat them.”
On 8 December we added the response from Eisai, sent to The BMJ on 7 December after publication.