The Importance of Publishing Negative Clinical Studies

Pharmacology Watch

Sources of funding for pharmaceutical research has come under scrutiny in the last decade as academic and government sources of funding have become increasingly scarce and the pharmaceutical industry has become the main source of research dollars. But the issue of objectivity has been raised, and some have even suggested that negative studies, that is studies that show a drug in an unfavorable light, may never be published. The American Medical Association has recently tackled this issue and has asked the department of Health and Human Services to establish a public registry of all clinical trials in United States. The registry would include information regarding the design of the study and the questions to be addressed. The registry would also contain data about the study results, both positive and negative. Some members of Congress have indicated interest in pursuing legislation to create such a registry, and even large pharmaceutical companies such as Merck and GlaxoSmithKline support the concept. But despite the AMA’s valid concerns, several negative studies have been newsworthy in the last 2 months. This issue of PharmWatch highlights a few of those.

Cognitive Effects of Estrogen Therapy

Two studies in the Journal of the American Medical Association suggest that estrogen alone therapy may be associated with a decline in cognitive function in post-menopausal women and may increase the risk of dementia. Both studies are follow-ups from the Women’s Health Initiative Memory Study (WHIMS) which had previously shown that estrogen plus progesterone therapy increases the risk of dementia in postmenopausal women. The first study was a follow-up of nearly 3000 women randomized in a double-blind fashion to conjugated estrogen, conjugated estrogen plus progesterone, or placebo. In the estrogen alone wing, 28 women taking estrogen developed probable dementia vs 19 assigned to placebo (HR, 1.49; 95% CI, 0.83-2.66). Similar rates were noted in the estrogen plus progesterone wing. When data were pooled for both estrogen and estrogen plus progesterone, the overall hazard ratio for dementia was 1.76 (95% CI, 1.19-2.60; P = .005). Increased risk of mild cognitive impairment was also noted in the estrogen alone group and the estrogen plus progesterone group. When the data were pooled, the hazard ratio for mild cognitive impairment was 1.25 (95% CI, 0.97-1.60). This study showed that there is no difference between estrogen alone vs estrogen plus progesterone therapy in the risk of dementia or mild cognitive impairment, and in fact, both therapies increase the risk of both these end points (JAMA. 2004;291:2947-2958). The second study asked whether estrogen alone alters global cognitive function in postmenopausal women. During a mean 5.4 years of follow-up, nearly 3000 women were randomized to 0.625 mg of conjugated estrogen or matching placebo per day. The women were assessed annually with the Modified Mini-Mental State Examination. The data showed that testing scores were 0.26 units lower among women assigned to conjugated estrogen compared to placebo (P = .04). When the data for estrogen alone was pooled with estrogen plus progesterone, the decrease was 0.21 (P = .006). The adverse effect of hormone therapy was more pronounced in women with low baseline cognitive function. The authors conclude that for women age 65 and older, hormone therapy, including estrogen alone therapy, had an adverse effect on cognition (JAMA. 2004;291:2959-2968). As pointed out in the accompanying editorial (JAMA. 2004;291:3005-3007), this study did not look at women who took estrogen in the years immediately following menopause. Previous observational data have suggested that there is a critical period just after menopause during which estrogen may be neuro-protective (JAMA. 2002;288:2123-2129). However, these current studies seem to conclusively show that neither estrogen nor estrogen plus progesterone are neuroprotective for older women.

Vitamin Therapy and Restenosis

Vitamin therapy to lower homocysteine levels has been touted as an effective way to prevent restenosis after coronary angioplasty. A new study, however, suggests that vitamin combination may actually increase the risk of restenosis in these patients. In a double-blind, placebo-controlled study from Germany and the Netherlands, 636 patients who had undergone successful coronary stenting were randomized to a combination of 1 mg of folic acid, 5 mg of vitamin B, and 1 mg of vitamin B12 intravenously, followed by daily oral doses of the 3 vitamins for 6 months; or to placebo. In a follow-up, the mean luminal diameter was significantly smaller in the vitamin group and placebo group (P = .008), and the extent of luminal loss was greater (P = .004). The restenosis rate was also higher in the vitamin group than the placebo group (34.5% vs 26.5%, P = .05). A higher percentage of patients in the vitamin group also required target vessel revascularization (P = .05). The authors conclude that contrary to previous findings, the administration of folate, vitamin B-6, and vitamin B12 after coronary stenting, may increase the risk of instent stenosis (NEJM. 2004;350:2673-2681).

Echinacea and the Common Cold

Echinacea purpurea, the commonly prescribed herbal remedy, may have no effect on the common cold, according to a new study. In this randomized, double-blind, placebo-controlled trial, 128 patients with early symptoms of the common cold were randomized to 1 mg of Echinacea or lactose placebo 3 times per day for 14 days or until cold symptoms were resolved, whichever came first. No statistically significant difference was observed between treatment groups for either a total symptom score (P range for symptoms = .29-.90) or mean individual symptom scores (P range = .09-.93). The time toward resolution of symptoms is not statistically significant between the 2 groups (Arch Intern Med. 2004;164:1237-1241). The authors admit, however, that testing different preparations and dosing ranges of Echinacea may be needed to confirm these findings.

Effects of Paxil in Children Under 18

GlaxoSmithKline has been accused of suppressing negative data about its antidepressant paroxetine (Paxil), showing that it is broadly ineffective in children and adolescents, and could increase the risk of suicidal behavior. The accusation comes in the form of a lawsuit from New York Attorney General Eliot Spitzer, who filed the suit in early June accusing the company of fraudulently suppressing the data. In response, Glaxo has published several studies on its web site, and states that these studies had previously been published in journals or presented at scientific meetings. The company also reiterates that paroxetine is not approved for treatment of patients 18 years or younger, and states that they do not promote off-label use of their products. The British firm has released data from 9 pediatric trials, as well as the bibliography of public communications derived from the studies, and letters to United States physicians summarizing the data. As mentioned earlier, GlaxoSmithKline, has stated publicly, it’s support of the American Medical Association’s proposal to create a national registry of all proposed pharmaceutical studies. More information is available at

FDA Actions

Schering has received approval from the FDA to market a new low dose estrogen patch for the treatment of osteoporosis. The patch, which is dime sized, is applied once a week, and delivers 14 micrograms per day of estradiol. It will be marketed this summer under the trade name Menostar.

This supplement was written by William T. Elliott, MD, FACP, Chair, Formulary Committee, Kaiser Permanente, California Division; Assistant Clinical Professor of Medicine, University of California-San Francisco. Telephone: (404) 262-5416. E-mail: In order to reveal any potential bias in this publication, we disclose that Dr. Elliott reports no consultant, stockholder, speaker’s bureau, research, or other financial relationships with companies having ties to this field of study.