A seemingly small mistake in a paper in The New England Journal of Medicine (NEJM) landed a Danish physician-researcher in hot water last month after a German company threatened to sue him for potential losses that could run in the millions of dollars. The exchange prompted media consternation in Denmark over whether academic freedom was being censored, but the researcher, Anders Perner of Copenhagen University Hospital has corrected the error, which occurred in the publication of a study of a widely used drug to prevent shock, and thereby averted legal action. Still, the episode has shone a light on a therapy that some researchers say may be doing more harm than good, despite its widespread use.
At issue is hydroxyethyl starch (HES), a synthetic derivative of regular starch that has been used globally for decades to prevent shock in patients who have lost large amounts of blood. It is also used to treat patients with sepsis, whose blood leaks out of their capillaries. "HES binds liquid, and the idea is that this keeps up the blood volume," explains Tobias Welte, a pulmonologist at the Hannover Medical School in Germany who has studied the compound.
On 27 June, Perner published a randomized clinical trial in NEJM that compared the effects HES has in septic patients to treatment with an alternative called Ringer’s acetate. The results were not good for HES: After 90 days, 201 of 398 patients in the HES group had died, compared with 172 of 400 patients in the group treated with Ringer’s acetate. Patients treated with HES were also more likely to need renal-replacement therapy and to suffer from severe bleeding. The study was “well-designed and well-conducted,” says Greg Martin, an intensive care medicine expert at Emory University School of Medicine in Atlanta.
On 9 July, however, Perner received an e-mail from German pharma company Fresenius Kabi, one of the biggest manufacturers of HES, saying that in the NEJM article he had misidentified the compound used in his study. Perner had used Tetraspan, produced by B. Braun in Melsungen, Germany, and commonly described as HES 130/0.42. (The first number gives the molecular weight of the chemical; the second is an indication of the number of hydroxyethyl groups in the starch.) While the methods section names the company and the drug, it is referred to as HES 130/0.4 in the original article and its title. Those are the specifications of Fresenius’s HES product, Voluven.
The e-mail, Perner says, included a threat by Fresenius to take legal steps to recoup any financial losses the company would suffer because of the false information. In a market worth billions, that could be a lot of money, says Welte. Perner says Fresenius also demanded an explicit statement that the first version of the paper was wrong.
Perner and NEJM have changed the online version accordingly and Fresenius says that's the end of the story. “It was our aim to have the scientific misinformation in the title and the text of the article in NEJM corrected,” a company spokesperson wrote in an e-mail. “This has happened so we see no need for legal action.”
Fresenius says that its product is different from the one Perner used not only in the hydroxyethyl ratio, but in other respects as well; for instance, in contrast to the compound tested in the trial, it is not produced using potato starch.
But Perner maintains it was not a mistake, saying the differences are inconsequential, and the data from his study are likely to apply to Voluven as well. The variation in the hydroxyethyl ratio is minimal and unlikely to have an effect on the drug's action, he argues. “We rounded to 0.4 because the ranges of the HES products made precision less relevant,” he says, pointing out that the hydroxyethyl ratio ranges from 0.4 to 0.44 in Tetraspan and 0.38 to 0.45 in Voluven. "The paper is more precise now," he concedes, "but from a scientific point of view, I think, it makes no difference." Welte agrees. "For all intents and purposes, these compounds are identical. If you do a trial with aspirin you do not have to write whether you used the drug produced by Bayer or someone else either,” he says.
The debate about the safety and effectiveness of the various forms of HES has been going on for some time. Welte was involved in a 2008 study, also published in NEJM, which concluded that HES solutions should be avoided "until long-term studies with adequate numbers of patients show that a particular HES solution is safe in critically ill patients.“ That study used yet a different compound (HES 200/0.5), “but you can lay the data of the new study on top of ours and they are almost identical,” Welte says.
HES compounds were first developed by Fresenius in 1974, and others soon followed. While the compounds received regulatory approval in the United States and throughout Europe, Perner contends that their safety was never adequately assessed by modern standards. “When side effects occurred, the companies made new, smaller molecules but launched on the back of the older products,” he says. Others share his concern. “In my opinion, there is sufficient evidence of adverse effects with starch solutions in sepsis to restrict use of all starches until additional safety and efficacy data are available," says Martin. Regulatory agencies have ignored evidence of potential side effects and failed to act, adds Welte.
While there has been concern about the compounds since 2001, the risk-benefit analysis has remained positive, the German medicine regulatory agency BfArM says in a statement to ScienceInsider. The new results from Perner’s study are being evaluated, a spokesperson for the agency says, but no decisions will likely be made until the results of another clinical trial, expected in August, are presented. That study, led by John Myburgh at The George Institute for Global Health in Sydney, Australia, uses Fresenius's compound.