Worst Pills, Best Pills

An expert, independent second opinion on more than 1,800 prescription drugs, over-the-counter medications, and supplements

A Review of Eszopiclone (LUNESTA): A Not-So-New Sleeping Pill

Worst Pills, Best Pills Newsletter article July, 2005

The Food and Drug Administration (FDA) approved eszopiclone (LUNESTA) as a “new” sleeping pill on December 15, 2004 despite a risk of cancer associated with the drug. The drug is marketed by Sepracor, Inc. of Marlborough, MA.

We put “new” in quotes because eszopiclone is derived from zopiclone, a drug that has been available in 85 foreign countries since 1987 (though never in the U.S.). Chemically speaking, zopiclone is a mixture of two molecules with identical atoms that are arranged...

The Food and Drug Administration (FDA) approved eszopiclone (LUNESTA) as a “new” sleeping pill on December 15, 2004 despite a risk of cancer associated with the drug. The drug is marketed by Sepracor, Inc. of Marlborough, MA.

We put “new” in quotes because eszopiclone is derived from zopiclone, a drug that has been available in 85 foreign countries since 1987 (though never in the U.S.). Chemically speaking, zopiclone is a mixture of two molecules with identical atoms that are arranged differently, called isomers. One of these molecules is eszopiclone. The two isomers that make up zopiclone are mirror images of each other.  

The marketing of drugs made in this way — by utilizing a single isomer of an existing drug that contains mirror isomers — is becoming more popular with drug manufacturers. As new drug pipelines run dry, companies need a way to ensure that they have a supply of patented drugs that ensure their ability to continue monopoly pricing practices and charge the public whatever the market will bear. A single isomer drug rarely offers a therapeutic advantage over the mixture of isomers from which it is derived — usually there is no such advantage. (see Selling New Drugs Using Smoke and Mirror (Images) in Worst Pills, Best Pills News March 2003 ). However, the single isomer is patentable distinct from its multiple-isomer origin. Thus, the drug company’s monopoly can continue.

Eszopiclone belongs to the family of drugs known as nonbenzodiazepine sleeping pills. This family also includes the widely prescribed drugs zaleplon (SONATA) and zolpidem (AMBIEN). Each of these drugs works in the same way. They act similarly to the older group of sleeping pills and tranquilizers called benzodiazepines, a group that includes products such as flurazepam (DALMANE), diazepam (VALIUM), and oxazepam (SERAX). Eszopiclone, zaleplon, and zolpidem could also be called nonbenzodiazepine-benzodiazepines — they affect the body in a  manner similar to benzodiazepines but have different chemical structures. In addition, like the benzodiazepines, all three are controlled substances with a risk for dependence.  

The FDA posted its reviews of the data submitted by eszopiclone’s manufacturer to support the approval of the drug on the Internet in early May 2005. We often rely on FDA reviews of drug company studies as sources for our articles because it is not always possible to know with certainty if a peer - reviewed medical journal article is science or promotion. The publicly available FDA reviews often give a totally different, often more negative picture of the therapeutic value of new drug compared to the general media and medical journals.

The FDA medical officer who reviewed eszopiclone recommended that the drug not be approved. Her major concerns were cancers in patients in clinical trials conducted before the drug was approved, tumors in laboratory animals, and drug-induced damage to DNA that could lead to the development of cancer. This recommendation is not surprising. Zopiclone, the mixture that contains eszopiclone, was denied approval in the U.S. because of its association with cancer.

The medical officer’s two immediate supervisors, the Team Leader and Division Director, concurred with her recommendation not to approve the drug. One of the reviewing pharmacologists also recommended against approval. The Director of the Division of Neuropharmacological Drug Products wrote in a February 20, 2004 memo:

I think that the mouse tumors (and, to a lesser degree) the rat pulmonary findings, are true findings, and pose at least a potential risk for humans, for the reasons given above. Given this conclusion, it is, of course, impossible to predict how significant a risk (if any) this poses for humans, but it would appear small. I would argue, however, that a risk of this sort (carcinoma), in the setting of recently approved drugs[zaleplon and zolpidem] without such a potential risk, for the indication insomnia, for a treatment with no evidence of a benefit of any sort compared to other available treatments, is too great a risk to justify approval[emphasis added].

The recommendations of four qualified FDA scientists were overridden by an official in the upper administration of the Center for Drug Evaluation and Research, the section of the FDA that approves new drugs. Even this official who made the decision to approve eszopiclone conceded that it had “no advantage over alternatives.” His rationale for approving the drug was:

My conclusion that the drug can be considered approvable is based on my view that 1) the carcinogenicity findings with the racemate [mixture of isomers containing eszopiclone] are very weak in the first place and 2) are simply not present in the single enantiomer [eszopiclone by itself] studies. I do not believe that an unreplicated weak finding should lead us to conclude that there is a human risk.

This official did not discuss that eszopiclone can cause damage to DNA, or genotoxicity. Additionally, in the single isomer studies of eszopiclone, the dose was one-half that in the animal studies of the two isomers combined — a finding the official also did not address.

The FDA is not the only organization that sees no therapeutic difference between the non-benzodiazepine and short-acting benzodiazepine sleeping pills. The British National Institute for Clinical Excellence (NICE), an organization created to make sure that public money is spent wisely for drugs in the United Kingdom, reviewed zaleplon, zolpidem, and zopiclone had in April 2004. Their conclusion:

It is recommended that, because of the lack of compelling evidence to distinguish between zaleplon, zolpidem, zopiclone or the shorter-acting benzodiazepine hypnotics, the drug with the lowest purchase cost (taking into account daily required dose and product price per dose) should be prescribed.

Since eszopiclone and zopiclone are, for all practical purposes, the same drug, it is clear that eszopiclone is nothing special compared to sleeping pills already on the market.

Sepracor, eszopiclone’s manufacturer, submitted six clinical trials to the FDA in support of the drug’s approval. In four of these trials the primary endpoint, the outcome used as the basis for the drug’s approval, was Objective Latency to Persistent Sleep, or the time it takes to fall to sleep. The FDA medical officer found that one of these four studies had “serious problems” and that it “cannot provide clinically meaningful results.”

In the remaining three trials for which time to sleep was the primary endpoint, the average time to fall asleep in patients taking 3 milligrams of eszopiclone ranged from 18.0 to 19.3 minutes. In patients given the placebo, the range was from 33.0 to 40.8 minutes. The difference in these three trials between eszopiclone and placebo ranged from 15.0 to 21.5 minutes.

One of these three trials also compared 3 milligrams of eszopiclone to a placebo and also to 10 milligrams of zolpidem. On average, patients taking eszopiclone fell asleep in 18.3 minutes. Patients on zolpidem took 16.6 minutes. This is an average difference of only 1.7 minutes, favoring zolpidem.

The editors of The Medical Letter on Drugs and Therapeutics, an independent source of drug information written for pharmacists and physicians, reviewed eszopiclone in their February 28, 2005 issue. Their less-than-enthusiastic conclusion was that:

The main difference between all of them [eszopiclone, zaleplon, zolpidem], except for half-life, may be that the manufacturer of Lunesta sponsored a 6-month trial and submitted the results to the FDA, while the other 2 manufacturers did not.

There is a major difference between eszopiclone, zaleplon, and zolpidem when compared with generic oxazepam (SERAX). Oxazepam is the short-acting benzodiazepine that we recommend in Worst Pills, Best Pills if you must use a sleeping pill. There is a remarkable difference in their prices.

The table below lists the retail cost of a 30 day supply of generic oxazepam, eszopiclone, zolpidem, and zaleplon as posted on the Web site of a popular Internet pharmacy.

By using generic oxazepam, a drug with a therapeutic value equal to eszopiclone, zolpidem, and zaleplon, you could save from $86.00 to $72.00 per month.

Eszopiclone will sell and taxpayers and patients will waste millions of dollars when equally effective, less expensive alternatives such generic oxazepam are available. How will this happen? The Wall Street Journal reported that “Lunesta is backed by a $60 million advertising campaign” and it “is off to one of the fastest prescription starts on record. In the seven days through May 9, it had 42,508 new prescriptions, giving it a 9.8% share of new insomnia-drug prescriptions.”

What You Can Do

There is no medical reason that you should pay eszopiclone’s exorbitant price. If you do need sleeping pills, which should be used only for short periods of time, consider generic oxazepam.


Retail Cost of a 30 Day Supply of Sleeping Pills

 Oxazepam capsules, 10 milligrams  $12.99 
 Eszopiclone (LUNESTA) tablets, 3 milligrams  $98.99
 Zolpidem (AMBIEN) tablets, 10 milligrams  $87.99
 Zaleplon (SONATA) capsules, 10 milligrams  $85.27