Overdosed America (17 page)

One might think that a new drug earns its place among preferred therapies only after it has been shown to be superior, or at least equal, to the best available therapies. Often this is not the case. Expensive brand-name drugs are frequently tested against a placebo (meaning no therapy) even when effective alternative therapies are already in use. Evidence of being significantly more effective than no treatment is
sufficient for the FDA
to approve new drugs and for doctors to prescribe them in place of older, established, and usually less expensive treatments. A study of OxyContin, a long-acting form of oxycodone (commonly known as Percocet), provides a particularly dramatic example.

The study was designed to test
OxyContin’s ability to provide relief
to patients who were having “moderate to very severe” pain following knee replacement surgery. Patients were randomized into two groups: those assigned to the treatment group received OxyContin twice daily as “preemptive” pain medication, in a dose equal to six Percocet tablets over each 24-hour period. The patients assigned to the control group (also having moderate to very severe pain) were given twice-daily preemptive doses of a placebo. Patients in both groups could request a single Percocet tablet every four hours if they were uncomfortable, and the preemptive doses of pain medication were adjusted based on patients’ requests for additional medication—meaning that the patients in the treatment group who were having breakthrough pain received a higher dose of OxyContin and those in the control group received a higher dose of an inert pill. Can you guess which group had more pain? A hint: the people in the OxyContin group averaged the equivalent of more than 10½ Percocet tablets per day, while the people in the control group averaged 2½ Percocet tablets per day. The results of this study, published in the
Journal of Bone and Joint Surgery,
concluded: “Preemptive use of controlled-release oxycodone [OxyContin] during rehabilitation following total knee arthroplasty [replacement] leads to improved pain control, more rapid functional recovery, and a reduced need for inpatient rehabilitative services.”

In other words, treatment of moderate to very severe pain after knee replacement surgery with preemptive doses of OxyContin is superior to treatment with preemptive doses of nothing. Based on this study, the drug manufacturer earned the right to claim that treating post–knee replacement surgery patients with OxyContin significantly decreases their pain, facilitates their rehabilitation, and shortens the time spent in rehabilitation centers. Does this mean that routine “preemptive” treatment with OxyContin is more effective than routine “preemptive” treatment with other shorter-acting (and much less expensive) pain medication? This study leaves that question unanswered.

The next step in designing a clinical trial is to determine the characteristics of the people to be included in the study. Ideally, people included in a trial reflect the population of patients to whom the results will be applied—those most likely to use the drug or device being tested. But this is not always the case, as we’ve seen with the studies supporting the use of both Vioxx and Pravachol. Many studies choose a population that is younger and fitter than the target population, and therefore less likely to show side effects. An
editorial in the
Canadian Medical Association Journal
pointed out that only 2.1 percent of all patients in studies of anti-inflammatory drugs are over the age of 65; yet senior citizens, the editorial points out, “are among the largest users” of these drugs and are more likely to have serious complications from them. The editorial also indicts research on drugs for Alzheimer’s disease: A study to determine the effectiveness of one such drug, Aricept, restricted the range of patients to age 65 to 74 and excluded people with other medical problems besides Alzheimer’s disease, thus minimizing the likelihood of side effects. The problem is that the vast majority of patients for whom this drug will be prescribed are older than this, and therefore the results of the study do not apply to them. As the editorial pointed out, “If frail older patients are going to be targeted for dementia therapy we need to study this group in clinical trials to ensure the safe administration of the drug.”

Studies of cancer drugs
are similar. Nearly two-thirds of all cancer patients are 65 or older, but only one-quarter of the people in cancer studies have reached 65. Most of these studies exclude older patients by requiring that participants be able to take care of themselves independently or be able to work. Conducting research on only the strongest subset of cancer patients is not a good way to find out how to treat most cancer patients. Perhaps older people will have more severe reactions or derive less (or possibly more) benefit from the cancer therapies being tested. In any event, the systematic inclusion of unrepresentative patients in clinical studies may be good for the profits of the commercial sponsors of the studies (at least in the short term), but it is not good for the people who will receive care based on this distorted “science.”

Even when studies are designed to enroll the right mix of patients, make fair comparisons between drugs, and measure valid end points, there is still no guarantee that they won’t be prematurely stopped. That’s what happened in
a study sponsored by Pharmacia, ironically titled CONVINCE
. The study compared Pharmacia’s blood pressure drug Covera, a long-acting reformulation of an older calcium channel blocker, with far less expensive standard therapies, a beta blocker (atenolol) and a diuretic (hydrochlorothizide). This huge study included 16,600 patients and was planned to continue for five years, yet it was stopped two years early. The results up to that point showed that the sponsor’s more expensive drug was slightly less effective at preventing the complications of high blood pressure than the less expensive drugs. According to an editorial in JAMA, the decision to stop the study went against the recommendation of its own data and safety-monitoring board. Ignoring its own experts, the CONVINCE study quit while it was behind. What rationale did the sponsor give for stopping the drug trial? According to the JAMA editorial, “business considerations.” What were they? We may never know, but I’d hazard a guess that it had something to do with the fact that the
sponsor’s drug, costing about $1.50 per day
, was proving to be no better than the generic drugs that cost as little as 15 cents per day. Nonetheless, doctors continue to prescribe calcium channel blockers for hypertension more than any other drugs, believing them to be in some way better. The marketing is evidently still “convincing,” even if the scientific evidence is not.

Often the medical researchers who carry out company-sponsored studies are not even allowed to see all of the data from the studies they are working on. These researchers are left in the position of analyzing and including in their articles only the data that the drug or device manufacturers have allowed them to see. In May 2000,
Dr. Thomas Bodenheimer brought many of these issues to light
in an important article in the
New England Journal of Medicine
titled “Uneasy Alliance: Clinical Investigators and the Pharmaceutical Industry.” One researcher quoted in the article explained that controlling access to the data allows drug companies to “provide the spin on the data that favors them.”

The September 2001 joint statement issued by the editors of major medical journals weighed in heavily on this important issue: “we strongly oppose contractual agreements that deny investigators the right to examine the data independently.. . . Such arrangements not only erode the fabric of intellectual inquiry that has fostered so much high-quality clinical research, but also make medical journals party to potential misrepresentation.” Practicing doctors count on the articles in medical journals to present and interpret the complete data, thus providing the “scientific evidence” that they trust to guide their patient care decisions. If even the researchers who write the articles have access to only the data that the corporate sponsors allow them to see, how can anyone have confidence in the “scientific evidence” published in the medical journals? And how can anyone have confidence in the medical care that is based upon results that have been censored to serve commercial interests?

The editors revised the guidelines of the International Committee of Medical Journal Editors to make explicit the recommendation that researchers have control over their data, analysis, and publication of their work.
A follow-up study
was done a year later to see if the new guidelines were being honored in university-based research contracts. It turns out that, despite their highly unusual public statement, the medical editors might as well have been whispering into the wind. The study found that their recommendations had not been implemented, and concluded that “academic institutions rarely ensure that their investigators have . . . unimpeded access to trial data.”

Before any medical article is accepted for publication in a respectable journal, it is peer-reviewed. Independent experts are called upon to evaluate the study’s data, and to concur (or not) with the authors’ analyses and conclusions. Most doctors believe that this peer-review process guarantees the integrity and completeness of the scientific evidence presented. But peer reviewers see only the data that have been included in the article—not all of the data the authors had access to and certainly not all of the data from the study. Readers of medical journals cannot assume that the process of peer review ensures fair and impartial presentation of research results.

Yet another way that drug companies can make sure that research results are written to best represent their interests is to hire ghostwriters to write the original draft of the article after a clinical trial is completed. As described by
Melody Petersen in the
New York Times,
the ghostwriter submits his or her draft to the drug company, which then passes it along for final approval to the authors of record—often busy doctors who are happy not to have to labor over the first draft. The problem with this system is that the drug companies get to infuse their perspective into the results from the very beginning. Dr. Linda Logdberg, a former medical ghostwriter, explained that drug companies “will drop a doctor if they don’t think he will be particularly malleable.” The
New York Times
article says, “The result . . . is marketing masquerading as science.” According to a study published in JAMA,
11 percent of the articles published
in peer-reviewed medical journals are written by “ghost authors.” (And 19 percent of the articles named “honorary authors” who had not contributed enough to the research and writing to justify being listed as authors.)

Even when studies that do not support the use of a sponsor’s product are published in medical journals, there is still a chance that well-funded marketing and public relations efforts will be able to protect drug sales.
The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attacks Trial (ALLHAT) study
,
*
for instance, compared the effectiveness of four drugs in preventing complications from high blood pressure. The study had been designed to measure important outcomes (heart attack and the broader category of cardiovascular disease—heart disease, stroke, other vascular disease, and the need for cardiac procedures to open blocked arteries). The study was to continue for four to eight years, but a part of it was
stopped prematurely
because the people who had been assigned to take one of the brand-name blood pressure drugs, Cardura—manufactured by Pfizer—were developing significantly more cardiovascular complications (particularly congestive heart failure) than the people taking a diuretic. At the time the results were published in JAMA, in April 2000, about
$800 million worth of Cardura
was being sold worldwide each year. The diuretic that was proving more effective than Cardura at preventing the complications of high blood pressure cost about one-seventh as much.

According to
a report in the
British Medical Journal,
as soon as they learned of the disastrous results for their drug, Pfizer hired damage-control consultants. The consultants discovered that most doctors simply weren’t aware of this research, and weren’t aware that Cardura ought not to be their first choice for the treatment of high blood pressure. So Pfizer simply kept quiet.

The American College of Cardiology (ACC), however, responded to the findings published in the JAMA article by issuing a press release, posted on its website, recommending that doctors “discontinue use” of Cardura. But within hours, the ACC downgraded its warning, recommending only that doctors “reassess” their use of Cardura. What happened? A confidential memo from Pfizer to the ACC requested a “clarification” of the ACC’s original press release. Bear in mind that Pfizer contributes more than $500,000 each year to the ACC.

The next round of results from the ALLHAT study came out two years later and contained more bad news for the manufacturers of brand-name blood pressure medicines. Again the low-cost diuretic was shown to be equal to or better than the higher-cost drugs—this time a calcium channel blocker (Norvasc) and an angiotensin-converting enzyme (ACE) inhibitor (Zestril and Prinivil). If medical practice were truly “evidence-based,” these results would have been a major problem for the manufacturers of the far more expensive but not as effective brand-name drugs. But not such a problem if the game is really hardball dressed up as evidence-based medicine. A strategic
marketing consultant for the pharmaceutical industry
was quoted in the
British Medical Journal
as saying, “So you’ve got one study that says yes, you should [use a diuretic], then starting the day after, you’ve got a $10 billion industry . . . and 55 promotional events . . . for an ACE inhibitor coming back in and saying ‘Here’s why my ACE inhibitor is safe and here’s why you should be using this.’ I mean, it’s promotion. Can ALLHAT stand up to that?” Almost certainly not.