Month: September 2015

Trial on antidepressant neglected suicide attempts

The antidepressant paroxetine was reported in 2001 to be effective and safe for adolescents. Now, this trial was re-evaluated following an open call in the Britsh Medicinal Journal (BMJ), as IFLScience reports. The new study was made possible with the help of GlaxoSmithKline, who initiated the original work and made the data accessible for re-evaluation. Paroxetine has been disputed over the last years, which was scaringly justified by the new results.

It came out that not only paroxetine is not beneficial to adolescents, but also that 11 persons from the 2001 study taking paroxetine attempted suicide or showed self-harming behavior, in comparison to only one person in the control group. This had been ignored by the researchers. It also had been ignored that parent- or self-rating by the patients of paroxetine did not show a significant difference from placebos. Here, we have a point that is valid in whole science.

“The investigator assessments always end up looking more favorable to the drug than those from the patients,” Jureidini told IFLScience.

Scientists are considered to be neutral and unbiased while they develop theories and prove or disprove them. But of course, this cannot be true, as scientists – along with all other human beings – have expectactions and are influenced by their opinions. That a researcher will judge results tendentially in favor for his new theory is not the problem, because it is just human. But it is apparently not responsible to legalize a medicine based on a trial wich was not double-checked.

Even though the danger of paroxetine is now revealed, not earlier than 14 years after the initial study, the BMJ call shows how important double-checking of clinical trials is. It should be not too difficult to legalize a drug only if its effectiveness and danger potential has been confirmed by two independent studies. Or, like in this case, the data set being analyzed separately.

A “TripAdvisor” for chemical probes

If medicinists want to test a new drug, they can literally choose between hundreds or thousands of reported molecules. But the real problem they are here facing is the high number of ill-suited molecules, that are not properly described, e.g., compounds that target enzymes other than the desired one, or have unwanted side-effects. Finding a suitable drug for a biomedical study can thus take endless hours before the study itself has even started.

Chemical biologists have now used crowdfunding to start an internet platform that recommends chemical probes, as reported in Science News. This action is in my opinion a very exciting act of self-empowerment, based on the strong impression that the self-correction mechanisms in scientific publishing are not sufficent. I think, one underlying problem might also be the reproducibility crisis, science is still facing. Once a new compound is published, its reproduction (and cross-checking) by other labs is not feasible anymore, since the work would be not original. Problems in reproduction therefore usually occur when the compound is supposed to be used for an application.

Maybe ChemicalProbes.org has the potential to provide an alternative metric for science, based on the applicability of drugs and drug-like molecules.