Mark Pine

The first clinical trial of human embryonic stem cells was given the green light by the FDA on Friday. As has been widely reported, this would be the first time that the use of human embryonic stem cells (hESC) to treat an otherwise intractable medical condition has been attempted. hESC have been the focus of much bitter controversy during the last decade related to the moral issues surrounding them. If the trial succeeds, then the promise of this long-anticipated therapy will have been realized for one actual disease, and the therapeutic claims of the proponents of the therapy will have been partially validated.

The investigation, carried out by the developer of the cells, Geron, will involve people who have sustained severe spinal cord injuries but have not suffered severing of their cords. That kind of trauma results in bleeding and inflammation at the site of the injury, and paralysis of muscles and loss of sensation in the body below the injury. The patients admitted to the study must have been injured in the thoracic (chest) region of the spine and they must be experiencing the most severe form of this type of injury with complete loss of motion and sensation in their lower bodies.

This initial trial will be small, including up to 10 patients, who will have been injured one to two weeks before receiving the stem cells. Studies in animals suggest that once several weeks or months have passed since the injury, scar formation would prevent effective treatment.

The stem cells will be administered to patients by injection through a syringe placed in a specially designed positioning device to ensure that the cells are delivered to the site of the spinal cord injury. Once they have been injected with the cells, the patients will receive an immunosuppressive drug for 46 days to minimize inflammation and stem-cell rejection.

Primarily intended as a “phase I” study of the safety of stem cell therapy, the trial protocol includes a long series of neurological examinations of the participants. Secondarily, the study will assess the return of their sensory and motor functions. The volunteers will continue to be followed for 15 years.

Geron derived the hESC from discarded embryos created as by-products of in-vitro fertilization (IVF) treatments. Using the cells, the company established immortal cell lines, which is possible because embryonic cells express a protein called telomerase that permits unlimited cell division. The company took some of the hESC and used molecular techniques to transform them into nerve tissue precursor cells called progenitor oligodendrocytes.

The study is designed so that the patients in the trial will have suffered the kind of spinal cord injury that causes loss of oligodendrocytes, a type of nerve tissue cell that supports the neurons that transmit impulses along the spinal cord. The underlying rationale of the trial is that by injecting precursor oligodendrocytes, the therapy will replace the kind of cell that was lost as a result of the injury.

Before attempting the trial in humans, Geron tested the rationale, effectiveness, and safety of the treatment in laboratory rodents with the same kind of spinal cord injury. The animal tests indicated that the therapy would likely be effective and safe in humans.

If Geron’s phase I trial shows that hESC therapy is safe and potentially effective, the study would be followed by larger phase II trials of up to 100 patients, and after that, by phase III trials of several hundred patients. The large trials would be used to definitively establish the safety and efficacy of the stem cell treatment and to learn more about how to administer it and which patients benefit from it.

If patients suffering from severe, permanent spinal cord paralysis and anesthesia recover some motion and sensation as a result of the new treatment, it would be a spectacular medical advance. Moreover, the accomplishment would encourage research into hESC therapies for other kinds of intractable injuries and diseases. For example, Geron has established stem cell lines to potentially be used for heart disease, diabetes, osteoporosis, arthritis, and liver disease, in addition to the nerve tissue cells being tested in the current trial.

As women age, their bones often lose calcium and fractures occur more frequently, a condition called osteoporosis. So it seems logical for older women (and men with osteoporosis) to take calcium supplements. Logical … but perhaps not a good idea.

The BMJ (British Medical Journal) published yesterday a meta-analysis (analysis of combined data) of five clinical trials involving 8,000 people who took calcium supplements or placebo pills. The researchers from universities in New Zealand, Scotland and the United States compared the rates of heart attacks among the participants, who were followed for an average of about four years. Those taking the calcium had a 31% higher rate of heart attacks than those taking the placebo.

The researchers also presented a larger analysis involving eleven clinical trials with 12,000 people. Although these trials provided less detailed information about the participants than first five, the results were similar. Those taking calcium had a 27% higher rate of heart attacks. One important caveat: All the trials compared calcium alone to placebo. Calcium is often taken along with vitamin D, and the results may or may not be applicable to the combination.

Cardiologists from two universities in the U.K. reviewed the implications of the meta-analysis in an accompanying editorial. They noted that no increase in mortality rate accompanied the increase in the rate of heart attacks, raising the possibility of a spurious finding, since an increase in heart attacks should result in an increase in mortality. But they also pointed out that although calcium supplements improve bone mineral density, they have not been found to reduce bone fractures. In view of the possibility that they may contribute to heart disease, it would seem wiser not to take them.

The BMJ also published last January a meta-analysis of clinical trials of supplements of vitamin D with or without calcium. Pooling seven clinical trials involving 70,000 participants, Danish medical researchers found that the combination of Vitamin D and calcium reduced the overall rate of bone fractures by 8% and the rate of hip fractures by 16%. Among the patients who took at least 10 mcg of vitamin D with calcium, the rate of hip fracture fell by 26%. Taking vitamin D alone, however, had no effect on fractures.

The reduction in the rate of hip fractures from taking 10 mcg vitamin D plus calcium is substantial, particularly since hip fractures result in considerable disability and increase mortality. Still, the outcome of a benefit-risk analysis that balances the potential benefit of reducing hip fractures against the potential harm of increasing heart attacks is not clear. Although the meta-analysis published by the BMJ yesterday did not investigate heart attacks in people taking calcium along with vitamin D, the possibility that the same increase would also occur with the combination is cause for concern.

The authors of the editorial take the view that calcium supplements “seem to be unnecessary in adults with an adequate diet. Given the uncertain benefits of calcium supplements, any level of risk is unwarranted.” For the time being, the better course would seem to be eating well, including adequate amounts of food containing calcium, but not taking more of the mineral in supplements.

News reports on Avandia (NYT, LAT, Bloomberg) say that an FDA advisory panel of experts from outside the agency voted yesterday in the majority to withdraw or severely restrict the use of the medication, which lowers blood sugar levels in diabetics but also exacerbates the risk of cardiovascular disease. In contrast, the Washington Post described the panel’s judgment as “mixed, with the majority in favor of keeping the drug on the market.” The tallies showed that a plurality of the experts (12) voted to withdraw the drug, and the next largest group (10) voted to restrict its use and strengthen warnings, but the remaining experts voted only for stronger warnings (7) or to continue the present arrangement (3).

Avandia [generic name: rosiglitazone, RSG], which is made by GlaxoSmithKline Plc (GSK), is under a cloud because of studies that show it increases the risk of heart attacks and cardiovascular disease. One such analysis found that Avandia raised the combined risk due to the drug of heart attacks, strokes, heart failure or death by more than half compared with another similar drug called Actos [generic name: pioglitazone, PIO]. The study was conducted by David Graham, a pharmaco-epidemiologist at the FDA, along with other scientists at the agency, and was published in JAMA, the AMA journal, last month.

Several news reports indicated that the availability of Actos, which works by the same mechanism as Avandia to lower blood sugar but is safer, influenced some advisory committee experts to recommend withdrawing or restricting Avandia. Actos is made by Takeda Pharmaceutical Co.

Moreover, GSK has apparently known of Actos’ superior safety for many years. Published in the NY Times along with the main article are several GSK documents dating from 2001 that detail discussions within GSK concerning a comparison between Avandia and Actos.

In 1999 GSK took it upon itself to study Takeda’s Actos to learn how it compared to its own Avandia. But company officials believed that doing a head-to-head study of both drugs at the same time would “entail very high risk” to its own drug. Instead they tested Actos alone in a group of patients that was “very similar to the Avandia clinical trials population.” One reason for doing this may have been that GSK would have been required by FDA regulations to report the results of any study of its own drug, but since Actos is Takeda’s drug, GSK would not be under a strict requirement to report.

The results of GSK’s study of Actos showed clear superiority of that drug in regard to some important safety parameters. When the results of the Actos trial were compared to those of the Avandia trials, Avandia raised the level of LDL cholesterol, a risk factor for cardiovascular disease, by 11% more. Similarly, Avandia raised the level of triglycerides, another cardiovascular risk factor, by over 20% more. The results of both comparisons fell at the upper limit of what GSK was prepared to accept as the likely outcome of doing a head-to-head clinical trial. The results also favored Actos over Avandia in regard to two other cardiovascular parameters, the levels of HDL cholesterol and fatty acids. GSK decided not to do a direct comparison of the two drugs at that time.

Commenting in an email on the results of the Actos study, Martin I Freed, a pharmaceutical R & D executive of GSK, wrote:

These [results] put Avnadia [sic] in quite of negative light when folks look at the response of the RSG monotherapy arm. It is a difficult story to tell and we would hope that these do not see the light of day.

The FDA advisory committee meeting, a conference open to the public that took place on Tuesday and yesterday, was unusual for the amount of controversy on display. Conflicting opinions within both the agency and the advisory committee were aired, with some of the participants disputing and others affirming the safety of Avandia. Perhaps that may indicate that the FDA, under its new commissioner, Margaret Hamburg, plans to increase the completeness, transparency and accuracy of the information it makes available to the nation.

New techniques to observe the presence of Alzheimer’s disease (AD) in the central nervous systems of living humans are currently in development and have made news recently. At present, the diagnosis of AD can only be made for certain at autopsy by finding the characteristic plaques in brain tissue. To develop and test potential medicines, however, it is necessary to be able to observe plaques in living people, in order to determine whether drugs can prevent or slow down the process of plaque formation.

One new technique involves the use of a radioactive dye that binds to plaques in the brains of AD patients. Yesterday, Gina Kolata, a science reporter for the NY Times, reported the advance, which makes it possible to observe plaques in the brains of AD patients, who are administered the dye and then undergo PET scans. Dr. Daniel Skovronsky, the researcher who developed the dye and has formed a company to market it, will describe his work at the International Conference on Alzheimer’s Disease next month in Honolulu.

Skovronsky’s technique is unique among diagnostic technologies in current development for AD. It has passed a crucial clinical test, that of proving the accuracy of the technique by showing that PET scans of human patients who have received the dye actually do identify AD plaques in their brains. The trial was done in hospice patients who had not long to live and who consented to receive the dye and undergo the scans and then have their brains autopsied after death. By May of this year, 35 patients had been studied, the target number for completing the study. The results showed that the autopsies confirmed the accuracy of the scans.

This month at the Society of Nuclear Medicine’s 57th Annual Meeting, researchers from the University of Leipzig announced the results of a clinical trial of another radioactive dye, florbetaben, in AD patients imaged with PET scans. Like Skovronsky’s imaging agent, the Leipzig researchers’ dye binds to plaques and can be observed with the scans in living patients.

In this case, however, the patients studied were not near to dying, and no autopsies were done as part of the trial. Rather, the accuracy of the technique was assayed by administering the compound to a group of 81 patients with probably AD, diagnosed using clinical signs of dementia, such as memory loss. As a control, florbetaben was also administered to a group of 69 healthy volunteers, and the presence of plaques on the PET scans was compared between the groups. All the subjects were 55 years or older.

The results of the trial were said to demonstrate the accuracy of the scans. However, since the methodology of the trial lacks a definitive diagnosis of the AD patients, which can only be made at autopsy, the results cannot be considered proof.

Researchers at the Cedars-Sinai Medical Center, the Weizmann Institute of Science in Israel and the University of Southern California have described a third technique. In early development, this technology may prove to be the easiest to use and least costly, since it does not rely on administering synthetic dyes and doing PET scans.

The medical scientists used curcumin, a component of the spice, turmeric, which is a fluorescent dye and binds to AD plaques also. They imaged the retinas of deceased patients with AD and those with presumed AD based on clinical findings. Using the curcumin dye and high-resolution optical imaging, they detected plaques in the retinas of patients but not in a group of aged-matched control subjects without the disease [correction per study author 7/2/10].

Although the retina of the eye is not a part of the brain, it is part of the central nervous system, and it is the only part of the CNS that can be visualized directly without surgery. The scientists showed that plaques also form in the retinas of AD patients, as part of the disease process.

The scientists also tested the curcumin technique in mice genetically modified to express a murine equivalent to human AD. Plaques were detected in the retinas and the brains of the mice, and the retinal plaques appeared even before those in the brains did. They scientists will present the results of their work at the international conference in Honolulu next month.

Any trial of a potential therapy to slow the progression of AD in living humans or to arrest or reverse the disease process would have to rely on techniques like these to image plaques in the brains of living patients. In such a trial, groups of patients with AD would be followed for months or years, and the patients treated with the potential therapy would be compared with the patients treated with a placebo. Both groups of patients would undergo serial imaging during the period of the trial, and the number of plaques would be determined and compared between the patients receiving the investigational treatment and those receiving the placebo. If a therapy worked, the patients in the treatment group would show, on average, fewer plaques than the patients in the placebo group.

“Exercise Fights Anxiety and Depression,” USA Today headlined yesterday. The article cited scientific studies of exercise treatment of depression and quoted a psychologist, who has co-authored a guidebook on treatment, urging therapists to write actual prescriptions of exercise for their patients.

A new meta-analysis of exercise therapy by researchers at Arizona State University featured prominently in the USA Today report. The authors of the study pooled the results of 58 clinical trials of depression with 3000 patients. They found an overall effect size of an 80% reduction in depression scores from exercise therapy.

In my personal experience, the effect of exercise on mood is a powerful one. I tend to worry and obsess about many things, and although I’m not clinically depressed, I easily get feeling stressed and gloomy. But I also exercise 5-6 days a week by jogging, cycling or swimming. Beforehand I might feel like a wet washrag, saturated with filthy, scummy water. But after the intense aerobic activity, followed by a cleaning shower, I feel laundered, wrung clean, and set to dry in sunshine and fresh air. It’s great!

To check out the science further, I searched for clinical trials of exercise as therapy for depression on the PubMed website of the National Library of Medicine. The query yielded more than 50 reports, which described positive effects in groups of women, including those with post-partum depression, in men, in the elderly, in stroke patients, and in patients with hypertension, cancer and other diseases.

One of the better studies was conducted at the Cooper Institute in Golden, CO, in 80 young to middle-age patients with mild or moderate major depressive disorder. The protocol used a dose-response design, with four groups assigned randomly (without blinding) to no exercise or supervised aerobic exercise 3 or 5 days per week at a low-dose level of 7.0 kcal/kg/wk (about 450 calories a week for an average-weight person) or a “public health dose” level of 17.5 kcal/kg/wk (about 1100 calories a week for an average-weight person). After 12 weeks (the usual duration of clinical trials of depression), depression scores in the high-exercise group decreased by 47% on average and those in the no-exercise or low-exercise groups decreased by 29%-30%. The frequency of exercise (3 or 5 days per week) did not matter. The result was statistically significant.

Exercise as therapy for depression ought to be prescribed much more often than it is, in my view. While drug therapy is effective, there are many negative effects, and a very common one is weight gain, a long-term adverse effect, associated with deterioration of physical health. In contrast, exercise therapy improves physical fitness and can aid in weight loss, while it helps to treat depressed mood.

A commentary last month in JAMA, the journal of the AMA, urged the NIH to require reporting of more results of clinical trials of medications on its website, clinicaltrials.gov. The author of the opinion piece is James Dabney Miller, a scholar at the Johns Hopkins Bloomberg School of Public Health. He wrote that the sponsors of clinical trials are not required to report results of the trials at the present time. However, as mandated by Congress in a 2007 law, NIH will require that the sponsors post some results of their investigations on the clinical trial website starting this September.

Miller also said that NIH should insist that sponsors of clinical trials report results of trials of drugs not yet approved by the FDA—including trials of drugs already approved for another medical condition but not for the one that is the focus of the trials.

In recent weeks, I’ve put up several posts on this subject, including the last two. In this post, I want to consider some issues related to Miller’s position, which shed light on a problem for implementation of evidence-based medical treatment. Additionally, I want to discuss some reasons to move forward on Miller’s recommendation.

Most research on new drugs and other new medical treatments is sponsored and performed, directly or through hired researchers, by companies such as pharmaceutical manufacturers with a financial interest in profiting from a successful outcome of the research. Consequently, up to the present time, the sponsors often fail to publish findings that are negative on effectiveness or show harmful adverse effects.

A case in point is rosaglitazone, which I wrote about Friday. It causes serious adverse effects on the heart, which prompted the FDA to require a warning. Evidence of the problem was the subject of a study in the New England Journal of Medicine by Nissen and Wolski of the Cleveland Clinic. Nissen, a cardiologist, and his co-author obtained and reanalyzed publicly available data from 42 trials by the drug’s sponsor, GlaxoSmithKline (GSK). The data included summaries of adverse events but not the patient-level source data. Nevertheless, the authors were able to conclude that “…compared with placebo or with other antidiabetic regimens, treatment with rosiglitazone was associated with a significant increase in the risk of myocardial infarction and with an increase in the risk of death from cardiovascular causes…”

The fact that these medical experts not associated with GSK were able independently to examine summary data and demonstrate a serious problem with the drug shows the need for posting clinical trial results on a public website. The company did not come forward with a similar analysis and conclusion. The rosiglitazone case was a fortunate one, in that the problem could be discerned by independent experts from available summary information. But most often that would not be possible. Thus, a main reason for publicly posting source data would be so such experts could check and reanalyze it to verify—or dispute—the sponsor’s claims.

Moreover, as Miller advised, NIH should also require posting of results of trials of already-approved and marketed drugs that are being investigated for new indications. The reason is that sponsors have no interest in publishing research that fails to demonstrate effectiveness or shows harmful adverse effects associated with an unapproved indication of a drug already on the market. This is because physicians have the right to prescribe such a drug even for an unapproved use. Oftentimes, a drug has acquired a reputation for effectiveness for an unapproved indication on the basis of a few suggestive but inconclusive studies or case reports. In such situations, company sales people may promote a drug for the unapproved indication, even though doing so is improper, and the company may benefit from increased sales of the drug whether or not clinical trials support the use.

As a result, if a company is profiting from such “off-label” sales of a drug, its most advantageous and least costly course of action may be to hold back information and trial results that do not support the use. In that situation, only outside experts analyzing publicly available source data would be able to determine whether the use is effective and safe, or otherwise.

Only one-third of clinical studies of medications compare the effectiveness of two or more drugs, and only 1% investigated cost-effectiveness of the medications studied. Physicians at Harvard and USC surveyed research on medicines published in six leading, prestigious medical journals between June 1, 2008 and September 30, 2009. The results of their investigation are reported in the latest issue of the Journal of the American Medical Association.

Health care costs are rising rapidly in the United States, with present expenditures amounting to 17% of GDP and increasing around 1% each year. Americans spend almost 50% more on health care per person than we would if we spent the same fraction of our per capita income as other developed nations.

Health-services researchers say a major factor contributing to our high health care costs is our greater use of high-cost medical treatments. In the context of the debate on reforming health care, the need for more research comparing the effectiveness of medical therapies has come to prominence. Last year’s economic stimulus legislation (American Recovery and Reinvestment Act of 2009) provided $1 billion for comparative effectiveness research. It is against this background that the JAMA report has appeared.

The authors identified 328 studies of medicines in humans in the 6 journals during the 16 months. Of these investigations, 104 (32%) met their criteria for comparative effectiveness (CE), which included only clinical trials, observational studies and meta-analyses of drugs approved by the FDA. To my mind, these findings are among the most significant:

• 90 of 104 (87%) of CE studies were funded partly or fully by non-commercial sources, including government.
• 126 of 224 non-CE studies (56%) were funded partly or fully by commercial sources
• Only 2 of 104 CE studies and none of the non-CE studies evaluated the cost-effectiveness of medicines.
• Studies with a positive effectiveness result that achieved statistical significance were more likely to occur in the non-CE group (63%) than in the CE group (39%).

In an editorial accompanying the report, Alec B. O’Connor of the Rochester School of Medicine and Dentistry pointed to a problematic role played by FDA regulations and the agency’s interpretation of them. For decades, the agency has approved new drugs if they are shown to be safe in use, as directed in the labeling, and also in the author’s words:

if the treatment demonstrates a statistically significant benefit over placebo (justifying label claims of efficacy), even if the new treatment appears to be less efficacious than established treatments or produces benefits of questionable clinical significance.

In practice, this has meant that many new drugs fall into the category of so-called “me-too” medications. If one pharmaceutical company (the innovator) discovers and markets a drug that generates enormous sales (a “blockbuster”), then other companies (the imitators) will follow suit and seek approval of other medicines in the same class. This kind of thing has occurred with antidepressant drugs, stomach acid inhibitors, cholesterol-lowering medications, and many other drugs that produce billions of dollars of profit for drug companies.

The imitator companies are required by the FDA to show only that the imitator drug is statistically significantly better than an inactive placebo. If these companies do comparative drugs trials in the course of drug development, they almost always perform non-inferiority trials. Such trials are designed to be much more likely to show no difference between two drugs than to show that one drug is superior. The non-inferiority trials support the imitator companies’ goal of showing that their drug is as good as the innovator drug.

Each new drug that comes to market, whether it is an innovative treatment or an imitator of one, requires the developer company to spend billions of dollars on research and marketing. Since there are many more imitator drugs than innovators, this dramatically drives up the costs of producing medicines.

O’Connor recommends that the FDA establish a requirement for well-designed comparative effectiveness trials within the drug approval process. Such trials would be designed to determine whether one drug is superior. He writes:

An FDA requirement for active-comparator trials could conceivably reduce new drug and device development due to the increased costs of trials including active comparators and the increased risk of discovering late in the development process that a new treatment is inferior to existing treatments.

Such an approach would save a lot of money at the same time as it would improve medical treatment, since it would mean that drug companies wouldn’t spend billions developing and marketing imitator drugs and the most effective treatments would get used. It would also increase the proportion of comparative effectiveness research funded from commercial sources, a supply which the JAMA report this week shows to have been lacking.