EGCG disrupts amyloid oligomers

As I mentioned in my post on the recent paper by Kukar et al., the disruption of amyloid plaques has been an ongoing focus in Alzheimer's disease research. However, plaques and inclusions are of concern in many diseases, and as a result there is a great deal of interest in finding molecules that can either dissociate, or prevent the formation of, amyloids of many different kinds of proteins. In the most recent issue of Nature Structural and Molecular Biology there is a paper suggesting that (-)-epigallocatechin gallate (EGCG) may be able to interfere with multiple proteins that form β-rich aggregates.

EGCG is a chemical found in green tea (although it is doubtful you could realistically drink enough green tea to absorb the concentrations used in this study). Previous studies had suggested that it altered the aggregation behavior of α-synuclein (αS) and huntingtin. So, Ehrnhoefer et al. use highly purified EGCG in a number of experiments to determine what effect it had on αS and the amyloid-β peptide (Aβ) (1). What these proteins have in common, besides the fact that their aggregation is associated with disease, is that the single proteins take on a β-strand structure that assembles into fibrils.

Ehrnhoefer et al. find that EGCG interferes with some aspect of this process, reducing the formation of the fibrils while inducing the formation of some alternate oligomeric structure. In the case of αS, the result is a spherical oligomer (Figure 1), although the gel filtration results indicate that a large spectrum of oligomeric states is formed at lower concentrations (the trace suggests that these oligomeric forms are interconverting during elution). NMR and other data show that EGCG associates directly, but non-specifically, with the protein backbone, and that the compound has strongest affinity for the C-terminus of αS, which may play a role in preventing aggregation. The EGCG-treated oligomers had reduced β-strand content (as assayed by CD). Treatment with EGCG appeared to reduce αS toxicity in cultured cells, although this was measured strictly in terms of cell death.

Similar results were seen with Aβ—addition of EGCG reduced the formation of fibrils and the toxicity of amyloids towards cultured cells. Again, the oligomers formed in the presence of EGCG could be quite large, and took a spherical shape.

Based on these data, the authors propose that EGCG binds preferentially to unfolded proteins and interferes with the formation of regular β-strand structure. The EGCG-bound proteins are unable to form fibrils, and therefore EGCG oligomers compete with fibrils for monomers, slowing the formation of the latter. The net effect is to divert these unfolded proteins out of amyloidogenic pathways and into alternate oligomeric structures, which appear to be nontoxic, or at least less toxic.

Can EGCG or a derivative be turned into a drug to treat Alzheimer's disease, or a general treatment for amyloidoses? This is an uncertain proposition. As the authors of a commentary (2) in the same issue of NSMB note, EGCG's nonspecific assault on amyloids may damage some normal structures built on this architecture. Moreover, because EGCG seems to bind unfolded regions nonspecifically, it has the potential to interfere with any of the numerous signaling proteins that possess such regions. The potential for side effects is very high, and the continued viability of cultured cells in the presence of EGCG, while reassuring, is not a particular reason to believe the compound is safe at high concentrations in the human nervous system.

The promiscuity of EGCG's interactions with unfolded regions poses another problem, in that all these proteins will act to interfere with EGCG's action on its intended target. Fairly high ratios of EGCG were necessary in these assays, and they mostly involved purified proteins. In vivo, all unfolded proteins will act to titrate EGCG out of plasma, meaning that significant quantities of this (or any other non-specific molecule like it) would need to be used in order to achieve the desired effect. This again raises the likelihood of side effects.

Of course, the most severe complication arises from the nature of amyloidoses themselves. Although the obvious presence of plaques and inclusions naturally leads us to suspect that these aggregates are the agents causing the disease, increasing evidence suggests that amyloids are merely the endpoints of some other process that is the actual culprit. In the case of Alzheimer's disease, for instance, a recent article in Nature Medicine has provided very strong evidence that soluble Aβ dimers are the dominant contributors to Alzheimer's pathophysiology (go check out Ashutosh's excellent discussion of this article over at The Curious Wavefunction for more information). Now, given that these dimers do eventually form amyloid, it seems likely that they have β structure in their pathogenic form, which EGCG will probably disrupt, but this is not guaranteed. Small molecules like EGCG that prevent deposition into amyloid may actually exacerbate the problems they are meant to solve. Further research is needed to establish that the EGCG oligomers of αS and Aβ are not toxic in vivo.

Diseases associated with protein aggregation continue to pose a challenge precisely because we have such a poor handle on their pathogenesis. Ehrnhoefer et al. clearly demonstrate that EGCG possesses the ability to alter the behavior of amyloidogenic unfolded proteins. While that may imply that it has promise as a broad-spectrum drug to attack these diseases, the promiscuity of its action is a cause for concern from the perspective of dosage and side effects. And, because soluble oligomers may well be the pathogenic species in many (if not all) of these diseases, our optimism about this approach must be tempered with an awareness that the actual effect of EGCG may be to enhance, rather than diminish, the toxicity of the relevant protein targets.

1. Ehrnhoefer, D.E., Bieschke, J., Boeddrich, A., Herbst, M., Masino, L., Lurz, R., Engemann, S., Pastore, A., Wanker, E.E. (2008). EGCG redirects amyloidogenic polypeptides into unstructured, off-pathway oligomers. Nature structural & molecular biology, 15(6), 558-566. DOI: 10.1038/nsmb.1437

2. Roberts, B.E., Shorter, J. (2008). Escaping amyloid fate. Nature Structural & Molecular Biology, 15(6), 544-546. DOI: 10.1038/nsmb0608-544

Read the rest...

Zetia and Vytorin update

A few months back I mentioned a study called ENHANCE from the makers of Vytorin and Zetia that showed these drugs had no effect. That study has again entered the spotlight thanks to the release of the results and a pair of commentaries in the New England Journal of Medicine. The published results are essentially as was previously reported: addition of ezetimibe to an existing simvastatin regimen was of no benefit to patients, either in terms of reducing a measure of atherosclerosis or in the frequency of adverse events. The article (citation below) is worth checking out if you are interested.

As I mentioned in the previous post, these results are not definitive. The group under study was a high-risk group, but the fact that most of the participants were taking statins already significantly complicates the interpretation of the results. Is the lack of improvement with the addition of ezetimibe simply a case of diminishing marginal returns?

The really curious thing here is that ezetimibe did what it was supposed to do. Patients who took it had lower LDL levels than patients who didn't. Yet this reduction in LDL did not improve the measured outcomes. Does this mean that the measurement (intima-media thickness) was not appropriate? Or do these results imply that there are deficits in our understanding of the interplay between LDL, statins, and arterial disease? Additional studies using other observables are underway, but will not be completed until 2011. Until those results come in, the American College of Cardiology recommends that ezetimibe be used only when other approaches (notably statins) have failed to control LDL. I want to reinforce that these results do not mean that ezetimibe is dangerous, or even that it does not help prevent cardiovascular disease. The present situation is one of uncertainty. If you are presently taking ezetimibe as Vytorin or Zetia, continue to do so until you have a chance to discuss your medications with your doctor.

You can read the study here:
Kastelein, J.J., Akdim, F., Stroes, E.S., Zwinderman, A.H., Bots, M.L., Stalenhoef, A.F., Visseren, F.L., Sijbrands, E.J., Trip, M.D., Stein, E.A., Gaudet, D., Duivenvoorden, R., Veltri, E.P., Marais, A.D., de Groot, E. (2008). Simvastatin with or without Ezetimibe in Familial Hypercholesterolemia. New England Journal of Medicine, 358(14), 1431-1443. DOI: 10.1056/NEJMoa0800742 OPEN ACCESS

The commentaries have DOIs 10.1056/NEJMe0801608 and 10.1056/NEJMe0801842. Both are open access.

PalMD of the Denialism Blog has another commentary on this issue.

Read the rest...

What will cure your child's cough?

Continuing with the recent mini-theme of pharmaceutical agents going awry, it seems appropriate to revisit last week's FDA advisory that children under the age of 2 should not be given over-the-counter cough medicines. This advisory had been expected for several months, of course, and the companies involved had already pulled most of these products from the shelves voluntarily. The FDA has not yet resolved the question of what to do about cough medicines for children aged 2-11, but because of the risks of accidental medication and the existence of serious questions about the actual effectiveness of these medications, there is a substantial probability that the agency will decide to ban these as well. Of course, this would seem to leave parents in a bit of a bind: what will they turn to now that nasty-tasting cold medicines are gone?

Of course, the situation isn't that dire: after all, cold medicines do not cure colds: they merely treat symptoms. No cold medicine can actually diminish the length of time a child is sick. The main upshot of this decision will be to put more of a burden on parents, who will have to deal with illnesses using more traditional means. At the same time, I am reminded of a recent study indicating that traditional means might work just fine, at least in limited cases. In December, a report in Archives of Pediatrics and Adolescent Medicine stated that honey was more effective in easing cough symptoms in children aged 2-18 than the popular cough suppressant dextromethorphan (DM).

The study did not rely on objective measures of coughing frequency or intensity, but instead depended on the subjective reports of parents to determine the effects of the medicine. Children were treated with either honey-mimic DM (formulated to approximate the taste and consistency of honey), actual buckwheat honey, or nothing at all. While families receiving no treatment were not blind to this fact, the honey and DM treatment groups were blinded. Surveys filled in before and after treatment dictated who entered the trial and assessed the treatment results. The study involved a few more than 100 children from the Hershey PA area; the sample size was limited by the requirement that no cough medicine have been given to the child in the previous 24 hours.

The outcome is shown in the figure at right—higher scores on the survey reflect a worse condition (i.e. 6 = extremely frequent cough). "First night" reflects the pre-treatment condition, while "Second night" comes from the results of a telephone survey carried out after treatment. You'll notice right away that the claim honey is "better" than DM reflects a pretty narrow difference. For the most part, scores are not that different no matter what treatment was used. This is especially surprising in light of the fact that the "no treatment" option was not blind. One might reasonably expect that awareness of null treatment would produce a negative effect on the subjectively reported outcomes; however, this seems not to have happened. Still, there is a clear, if slight, advantage for the honey treatment in this study. At the same time, it does not appear that parents will be much worse off if they do nothing at all. Keeping in mind that cough suppressants only alleviate symptoms, and have no effect on the underlying infection, these results sharply call the benefits of DM treatments for children into question.

This is not actually as crazy as it sounds. As the authors explain, honey has well-known antioxidant and anti-microbial properties. The dark honey used in this study is likely to be even stronger in this regard. Also, the authors note that sweet placebos have often shown a comparable effect to DM, so physiological effects of the sugar may also play a role. Shelley Batts of Retrospectacle, who covered this study when it came out, suggests that the viscosity of honey may also have a role, as it "coats the throat".

There are some complaints to be made about the study: geographically and temporally it is rather limited, and factors such as income and nutrition seem to lack controls. DM might have benefits in children with poor nutrition or additional conditions that would not be picked up by this study. Moreover, the nature of the sample—cases of cough that were severe enough to cause a parent to visit the doctor but not severe enough to induce the parent to use an OTC medication—may be selecting for parents who overreact to a single bad night of an otherwise mild cough. Clearly a larger study with additional data-gathering and controls is desirable. The best situation would be to enroll families at the beginning of cold and flu season in a truly blind trial and track the coughs longitudinally. Also, "no treatment" should be replaced by something, perhaps a colored simple syrup, to preserve blindness, and a fourth group in which DM is mixed directly into honey should be added to assess the possibility that characteristics of the honey-mimic DM broke the blind. Cynically, one would also prefer that such a study not be funded by the honey board.

Despite these limitations, the results are quite striking. In light of this study and its own evolving but clear preference to err on the side of caution it is likely that the FDA will either ban children's cold medications or recommend significantly stricter labeling. This shift in attitude on the part of FDA may mark the beginning of a generalized swing away from medication and back towards traditional treatments for children with mild illnesses. However, the advent of the preference for medication did not arise from medication itself; rather it evolved in concert with social considerations such as the rise in families where both parents work. In the social context of the 21st century a prescription for "plenty of bed rest, lots of fluids, chicken soup, and honey for the cough", may not be as workable as it was 50 or even 20 years ago.

Paul IM, Beiler J, McMonagle A, Shaffer ML, Duda L, Berlin CM. "Effect of Honey, Dextromethorphan, and No Treatment on Nocturnal Cough and Sleep Quality for Coughing Children and Their Parents." Arch. Pediatr. Adolesc. Med. 161 (12) 2007 p. 1140-1146.

Read the rest...

Uptake blockers do not improve on statins

It has not been a very good couple of weeks for big pharma in the public eye. They've been taking a pounding in the political arena since the primary season finally got into full swing. Last week, we learned that today's medicines, rather than funding tomorrow's miracles, mostly finance power lunches with your cardiologist. And this week, Merck and Schering-Plough, after much delay, finally announced the results of a major study, called ENHANCE, that did not demonstrate any benefit to using the highly-touted drugs Vytorin and Zetia rather than a generic statin.

Because almost everyone who owns a TV knows that Vytorin blocks both sources of cholesterol, and that "Zetia works differently", the results seem particularly damning. The companies involved did not help this appearance by holding back the results so long, and indeed doing so may have exposed them to lawsuits from shareholders and customers. The study was primarily performed by imaging the carotid artery, and in this regard no statistically significant difference was observed between patients treated with Vytorin (statin + Zetia) and statin alone. Also, there was no statistically significant difference in the number of patients who died from cardiac events or strokes, or suffered non-fatal infarctions in the study. While vytorin lowered cholesterol levels by 56% after 24 months vs. the 41% lowering seen in the statin group, this did not translate into improved outcomes by any measure.

So, is this a case of evil pharmaceutical companies trying to gouge consumers over worthless medications? Not exactly. Like every campaign promoting prescription drugs, commercials for Zetia and Vytorin oversold the benefits and did not sufficiently emphasize the risks of these medications. Their failure to improve outcomes in this study is troubling, but because adverse outcomes were so rare overall, it would be difficult to establish an effect one way or another. Also, this study was not geared towards measuring plaques and clots directly; rather it (indirectly) measured the effect of these drugs on atherosclerosis. The effect of uptake blockers on infarction rates was an incidental measurement.

One important consideration when interpreting these results is that the study group was not constructed to resemble the general population. Rather, the study was performed on individuals who had familial hypercholesteremia, a genetic condition that causes greatly elevated levels of LDL in the bloodstream. Without the data in hand I of course cannot make a solid judgment, but one possibility that immediately suggests itself is that in this population even a significant diminution of LDL levels is not sufficient to improve outcomes. There's just too much cholesterol for these uptake blockers to affect the outcome. In a normal individual with significantly lower levels of LDL Zetia and Vytorin might have a much greater effect.

Moreover, all the individuals in this study were already taking statins—which contributed to the low rate of adverse outcomes. It may be that the marginal improvement from adding an LDL blocker on top of a statin simply isn't that great. However, for individuals who cannot take a statin due to side effects, taking an uptake blocker might be a significant improvement over doing nothing. The study does not, as far as I can tell, speak to this possibility.

There is always the possibility that this study points to completely unsuspected aspects of arterial disease. For instance, these results might indicate that once lesions form they attract cholesterol very strongly, and therefore only extreme reductions in circulating LDL can affect their growth. Alternately, this outcome may indicate that mechanisms unrelated to circulating lipids play a more significant role in determining plaque thickness than previously suspected. In light of the highly unusual population used as study subjects, any grand pronouncements in this regard are premature. If these findings are replicated in upcoming studies of greater duration on more representative samples, however, a substantial re-examination may be in order.

The findings of the ENHANCE study don't really indicate that you should burn your Vytorin prescription and go back to just a generic statin, but it does create doubt as to whether this approach will prove efficacious in the general population. Further study, promptly published, is called for, and it would be wise for Merck and Schering-Plough to pull their advertising campaigns for the time being. Doctors should also be less eager to prescribe Vytorin for patients who are responding well to statins alone, but for patients who are not improving greatly with statins (or cannot take them at all), then a prescription for Vytorin (or Zetia) would still seem to be justified. As always, the best approach is to exercise and eat healthy foods, though that's much easier said than done.

Read the rest...

Pharma's Funding Lie

With medical coverage emerging as a major domestic issue for this election cycle, the cost of drugs is likely to become a hot topic, especially given the array of pills many aging baby boomers must take. The industry line is best articulated by a GlaxoSmithKline campaign implying that high drug prices really fund R&D for future drugs ("Today's medicines finance tomorrow's miracles"). Activists counter that the pharmaceutical companies are actually marketing-focused, a view popularly reinforced by incrementalism and the highly-visible promotional campaigns for conditions that are primarily cosmetic, such as hair loss and erectile dysfunction. A new study in PLoS Medicine by Marc-André Gagnon, and Joel Lexchin (click the link, PLoS is free) suggests that the latter view is closer to reality: they find that companies spend in excess of $57 billion annually on marketing to doctors and consumers. This amounts to nearly twice as much being spent on promotion as on R&D (which receives ~$30 billion). While by no means the final say in the dispute, this paper and distortions of its findings (on both sides) may figure significantly in future debates.

The easiest criticism of this paper, and one that may often be repeated, is that the authors did no primary research. The paper is essentially a synthesis of research performed by two independent firms: IMS (its data is widely cited by industry groups such as PhRMA), and CAM. The article contains a single table, and a quick glance at it will inform the reader that whenever two numbers were available, the authors always took the larger. The authors explain this well in the case of detailing, and I agree with their decision there. I also agree with the principle, though I am not convinced of the magnitude, of the "other" category. When it comes to the cost of free samples, however, I disagree with the choice to use the retail price in the assessment.

The authors' justification on this point is rather feeble, and comes across as almost petulant in tone:

Using the wholesale value for samples, the CAM figure would be appropriate if we were arguing that the money spent on samples should go to another activity such as R&D. However, we have used the retail value of samples because this is consistent with companies' reporting of drugs they donate [19]. As these are both categories of products that are being distributed without a charge to the user, it is inconsistent for donations to be reported in terms of retail value and samples in terms of wholesale value.

It seems to me that there is no point to this assessment unless our ultimate intention is to compare money spent on promotions that could be spent on R&D to the actual money spent on research. Certainly that is the only comparison that actually speaks to the issues the authors raise in the introduction. Including money that could not be spent on research or anything else, because it is fictional money, is nonsense. If the aim is to define the actual costs of the promotion to the drug company then the wholesale price is more appropriate. If the authors have an objection to using the wholesale cost for one kind of handout and retail costs for another, then the correct response would have been to use the proper, intellectually honest number (wholesale cost) in their own analysis and argue that pharmaceutical companies ought to do likewise when reporting charitable contributions. Just because Eli Lilly misrepresents its costs doesn't mean you can, too.

However, there are other factors that make the CAM estimate questionable on this point, especially that any number of samples was reported as one. Most likely the actual cost of samples to the company lies somewhere between the CAM and IMS estimates; promotional costs therefore lie somewhere between $48 and $57 billion, or 160% to 190% of research expenditures.

Another key feature to note from the table is that direct-to-consumer advertising makes up a relatively small fraction of the total expenditure. This reflects a truth that anyone even tangentially related to the healthcare system has known for a long time: most of the actual marketing that pharmaceutical companies do is lobbying your doctor. Although they are pervasive, fundamentally uninformative, and frankly annoying, television and print advertisements for drugs constitute such a small percentage of the actual promotional costs that banning them again would not lead to a noticeable reduction in drug prices.

The findings of this paper should also be put in the context of a transformation underway in the pharmaceutical industry. The so-called "blockbuster" drugs that provided substantial profits over the past decade or so will soon lose (if they have not already lost) their patent protection, and some needed to be withdrawn due to failures of the clinical trial system. In order to adapt, many companies are shedding most or all of their R&D operations. The emerging model in the industry is to allow "small pharma"—tiny companies started by academics or entrepreneurs using venture capital—to do most of the legwork and then buy up or enter marketing partnerships with those companies once they have promising products that have passed phase I or II clinical trials. This model was promising and robust up to about two years ago, because plenty of capital was available. In the present economic climate the availability of capital is far less certain, however. In addition, the same issue that induced big pharma to shed R&D—poor ROI—will eventually act to inhibit the venture capital investments that small pharma requires.

Despite this increasing aura of uncertainty, the fact remains that many pharmaceutical companies appear to be undergoing a transition from being primarily research entities to being primarily production and marketing entities. In that light, the new estimate of the research-to-marketing ratio is hardly surprising, though it will doubtless be embarrassing to PhRMA and feed the rhetoric of populists such as John Edwards. Yet despite the vitriol that will surely be spewed, in reality there is little that can be done. As mentioned, the most visible marketing efforts of pharmaceutical companies constitute only a minor portion of actual promotional costs. While it would be wise and probably beneficial to public health to restrict these advertisements once again, it is unlikely that any reduction in medical costs or increases in R&D budgets would be achieved by such regulations.

Congress, if it desired, could take steps to reverse the current trend and strengthen FDA power to restrict off-label marketing of existing drugs, and of course a new President could make this an enforcement priority at FDA. However, enforcement of any such provision would be extremely difficult and subject to legal challenge on First Amendment grounds. Moreover, the FDA (and not coincidentally the USDA) are in need of a major overhaul and possible restructuring in order to achieve their existing missions; stapling on another major enforcement problem will not serve anyone. Regardless, off-label marketing does not constitute a majority of the promotional budget.

The First Amendment clearly protects on-label marketing, and at any rate promotions of proven drugs actually serve the public interest, up to a point, by making doctors aware of improved approaches for dealing with illness. The truth of this statement, however, is inversely related to the degree of incrementalism in drug discovery. "It's a bigger pill" is generally not a compelling rationale for new prescriptions or enormous marketing outlays. Nonetheless, the presence of a definite public interest in allowing marketing to doctors makes unclear what steps Congress should (or even can) take to regulate or diminish promotional spending.

The only tool readily available for public use against on-label marketing is shame. Either the companies themselves can be pressured to reduce their promotional budgets (unlikely), or activists can put pressure on professional societies and medical boards to implement ethical restrictions on what kinds of promotions their members can engage in (possible). Regulations against accepting expensive lunches and dinners or attending marketing "seminars" in exotic locales may be able to push back some spending. In this regard, the Gagnon and Lexchin study may prove a useful tool; the grandstanding of politicians most likely will not.

Fundamentally, however, this trend cannot be stopped, because marketing will always be a better—or at least more predictable—investment than research. Although this attitude is ultimately self-defeating, the safer course to higher profits in the near term is to aggressively market existing drugs and secure longer periods of exclusivity by lobbying for longer patent protection or incrementally improving medicines and delivery systems. The release of combination drugs such as Caduet reflects this sensibility. Most money spent on research never produces so much as a Phase I trial, and the discovery of a revolutionary medicine, though extremely profitable, is also extremely rare. For that reason, a conservative mind will always prefer promotion and production to research and development. This is the attitude that underlies the ongoing strategic shift in big Pharma's approach, as well as the findings, debatable though they may be, of Gagnon and Lexchin.

Gagnon MA, Lexchin J (2008) The Cost of Pushing Pills: A New Estimate of Pharmaceutical Promotion Expenditures in the United States. PLoS Med 5(1): e1 doi:10.1371/journal.pmed.0050001

Read the rest...