Milnacipran (SAVELLA) Fails to Alleviate Fibromyalgia Pain, Has Safety Concerns

The Food and Drug Administration (FDA) approved the drug milnacipran (SAVELLA) earlier this year to treat fibromyalgia despite the fact that studies have failed to show that the drug alleviates pain associated with fibromyalgia.

Milnacipran was approved in January 2009, becoming the first drug approved only for fibromyalgia, a chronic condition in which patients experience pain in many places in their body. However, two other drugs — pregabalin (LYRICA) and duloxetine (CYMBALTA) — have FDA approval to treat fibromyalgia.

Drug approval based on weak data

Scientists have found no rigorous ways to measure fibromyalgia thus far, partly because it is a medically unexplained syndrome. The studies submitted by the drug’s co-sponsors (Cypress Bioscience and Forest Laboratories) to the FDA for the approval of milnacipran exhibit not only dubious evidence of effectiveness but also questionable statistical manipulation on the part of the drug’s co-sponsors. By re-evaluating the results of an earlier trial based on findings in a second trial, Cypress Bioscience and Forest Laboratories managed to make it seem as though the unremarkable results seen in the drug’s pre-approval studies were more favorable than they actually were, thus favoring the drug’s approval.

The director of the FDA division responsible for the approval of rheumatology products wrote the summary review of milnacipran. His review explains that although the results of the first of the two required pre-approval trials showed no statistically significant differences between groups taking a placebo (sugar pill) and those receiving the drug, the FDA allowed the sponsors to change both the protocol for the second study as well as the analytical method of looking at missing data, opening the gates for more favorable-appearing results.

After managing to "achieve" a successful outcome in the second study by changing the methods for calculating results, drugmakers Cypress Bioscience and Forest Laboratories then reanalyzed the first study and found that, lo and behold, it, too, was positive. This is known as a post-hoc analysis and, in general, is never allowed because of the possible bias introduced, but for reasons never explained in the FDA documents, it was accepted in this case. This allowed the drug manufacturer to obtain a more favorable result that increased its chances of FDA-approval.

Data manipulations help drugmaker "improve" drug’s results

Although the primary complaint of patients with fibromyalgia is pain, neither the drug’s co-sponsors nor the FDA could find a statistically significant improvement in pain when looked at by itself.

To achieve data that supported the case for approving the drug, drug company researchers determined that, for their purposes, the best markers of improvement were two combinations of results:

• a high level of self-described general improvement in function and quality of life combined with a 30 percent improvement in pain;
• a high level of self-described general improvement in function and quality of life combined with a 30 percent score improvement in a self-reported questionnaire on patient function. (Patient function includes general activities of daily life, such as housework, driving a car and shopping.)

In the first clinical trial, only 27 percent of patients taking milnacipran reported general improvement and an improvement in pain, compared to 19 percent of patients taking a placebo who claimed to experience the same improvements.

Similarly, in the same trial, only 20 percent of patients taking milnacipran reported general improvement and an improvement in patient function, compared to 12 percent of patients taking a placebo who claimed to experience the same improvements.

Although researchers were able to generate a statistically significant result only when they manipulated data to combine "pain" with "patient function," there was still only a very slight difference between patients taking a placebo and patients taking the drug – at most 7 percent to 8 percent. That means that more than 90 percent of people in the trials had no benefit whatever yet were subjected to a wide variety of adverse effects. The second study had similar results.

Fibromyalgia is considered a chronic condition; yet even with all these efforts at achieving success, patients felt relief only after taking the drug for three months.

Europe rejects drug as unsafe and ineffective

The European Medicines Agency (EMEA), the European version of the U.S. FDA, agrees with Public Citizen’s analysis that milnacipran is both ineffective and unsafe and recently turned down approval for both milnacipran and pregabalin to treat fibromyalgia (the makers of duloxetine have not submitted an approval application to consider the drug for this indication).

With respect to milnacipran, the EMEA concluded that "the effect of the medicine was marginal" and that a lack of data on long-term effects existed.

With respect to pregabalin, the EMEA concluded that there were "no consistent or relevant reductions in pain or other symptoms in the short-term studies, and the maintenance of pregabalin’s effect was not shown in the longer study."

The milnacipran drug label lists 14 warnings and precautions, including suicide risk; serotonin syndrome (a potentially life-threatening condition); increases in blood pressure and heart rate; seizures; liver toxicity; abnormal bleeding; and withdrawal symptoms.

Withdrawal symptoms occur because milnacipran can be addictive. That means that patients have to discontinue the drug slowly, which can be a problem when side effects develop.

One of the major problems to occur is an increase in blood pressure — 20 percent of patients with normal blood pressure before starting milnacipran became hypertensive as a result of taking it, compared to 7 percent of placebo patients.

Gastrointestinal side effects also were common. In a study in healthy, fasting volunteers given a single 100 milligram dose, vomiting occurred in 58 percent and nausea in 71 percent. Even after a single 50 milligram dose, approximately 50 percent of those given the drug had nausea and 10 to 13 percent had vomiting.