Significant doubts over scientific claims made about Procter & Gamble's osteoporosis drug Actonel were raised this week after a report by The Times Higher last year that Sheffield University researchers had put their names to findings without having carried out independent analyses of the drug-trial data.
In interviews due to be broadcast this week on You and Yours, the Radio 4 consumer affairs programme, and in others given to The Times Higher , experts argue that there is insufficient evidence to warrant the conclusions on Actonel published by Sheffield's Bone Metabolism Research Unit. The claims have helped boost the drug's standing against its main rival, Fosamax, which is made by Merck and Co.
Actonel and Fosamax act to reduce the risk of fractures by increasing bone mineral density and reducing bone "turnover" - the rate at which bones break down and repair themselves. It is generally understood that Fosamax, the market leader, reduces bone turnover most effectively. But research authored by Richard Eastell, head of the Bone Metabolism Unit, has claimed that there is a "plateau effect" - a point beyond which any reduction in turnover does not lower fracture risk - which suggests that Actonel is not necessarily less effective than its rival.
Professor Eastell reported the conclusions, which P&G disseminated widely, at the 2001 conference of the American Society for Bone and Mineral Research, then again a year later at the annual meeting of the International Osteoporosis Foundation. In 2003, they were published in the Journal of Bone Mineral Research ( JBMR ). Now, senior academics have expressed concern over the conclusions.
Bill Fraser, head of the metabolic bone disease department at the Royal Liverpool University Hospital, was at the 2001 meeting. He tells Radio 4:
"When I first saw the data, I wondered if there was a problem with the measurements. I was concerned that the blood and urine tests weren't sensitive enough, not good enough to be able to make the conclusions that Richard had come to."
Martin Bland, professor of health statistics at York University, analysed the JBMR paper. He tells You and Yours : "It is a strange analysis because they say they're not going to do what you'd think of as a normal statistical analysis, but something that doesn't allow them to actually test whether anything is happening."
Asked about a plateau effect, Professor Bland says: "Well, their statistics don't demonstrate it - it is just a picture of a plateau effect. There's nothing to tell you how they got it from the data. You just have to take it on trust."
He says he would not have passed the paper for publication. Jane Hutton, chair of Warwick University's statistics department, told The Times Higher :
"There is no evidence in this paper to justify the claim that there is a threshold for decrease in bone resorption (turnover) and the associated risk of fractures."
This week P&G said: "This is not about our product's safety or fracture-reduction benefits, which have been proven in one of the largest clinical trials programmes for an osteoporosis therapy.
"We stand by our conclusions published in the JBMR... Any conclusions reached in exploratory analysis are, and should be, open to legitimate academic and scientific debate and further study. In fact, an independent review of the data is ongoing, which we believe will confirm the validity of the statistical approach used, and the conclusions reached, on the basis of the database studied."
Professor Eastell declined to comment, but he had indicated to colleagues that he had worked closely with the P&G statisticians to identify the best approach to analysing the data.