Microsoft word - assessing drug development reforms.doc

Paper No. CAE 003
Assessing drug development reforms in
Angell’s The truth about the drug companies:
how they deceive us and what to do about it
W. Maurice Young
Centre for Applied Ethics
Assessing drug development reforms in Angell’s The truth
about the drug companies: how they deceive us and what to do
Abstract
In articles and the book, The truth about the drug companies: how they deceive us and what to do about it, Marcia Angell, a former editor of the New England Journal of Medicine, identifies a host of problems within the contemporary biopharmaceutical industry and proposes reforms that would, in her view, resolve seven of the most critical ones. In this (partial) review article I confine my discussion to an analysis and critique of two of Angell’s main reforms, namely, the ones aimed at drug development: her suggestion to create an institute for prescription drugs, which would take over the conduct and interpretation of all clinical trials; and her proposal to require that new drugs be compared to existing drugs before marketing approval. My overall position is that being convinced of serious shortcomings in some areas of the industry – such as mammoth marketing budgets and offensive advertising – does not justify a blank check to potential reformers. Although I agree that many of Angell’s issues need to be addressed, I show that both of her proposed reforms are fundamentally flawed. Along the way I examine a point that influences many of Angell’s views about biopharmaceutical R&D, namely her answer to the question, who innovates? I contend that Angell underestimates the contribution that industry makes in the R & D process. I conclude with some suggestions about what should be the general thrust of reforms in drug development. The author would like to give special thanks to Alain Curaudeau, Josesph DiMasi, Patrick Lauzon, Julia Levy, and Emily Marden for extensive comments on earlier drafts. In addition, J. Anslow, M. Battersby, N. Cagic, P. McKelvey, B. Newell, C. Perry and J. Tansey made helpful suggestions. The author takes sole responsibility for the text. Please observe standard academic conventions for citation. For example: Levy E. (2005) “Assessing drug development reforms in Angell’s The truth about the drug companies: how they deceive us and what to do about it.” Electronic Working Papers Series. W. Maurice Young Centre for Applied Ethics, University of British Columbia at www.ethics.ubc.ca. 1 Adjunct Professor, Ethics and Science, the W. Maurice Young Centre for Applied Ethics, UBC. Former VP Corporate Development, QLT Inc. email: [email protected] Assessing drug development reforms in Angell’s The truth
about the drug companies: how they deceive us and what to do
1 Introduction
Recently there have been trenchant criticisms of the biopharmaceutical industry and of the drug development process. Marcia Angell’s book, The truth about the drug companies: how they deceive us and what to do about it, is one of the most critical and is important not only for its content, but also for its provenance, coming as it does from a former editor of one of the most respected journals in the U.S., the New England Journal of Medicine (NEJM) (2004b). Indeed this volume, plus related articles by the same author that appeared in NEJM, The New Republic and The New York Review are significant contributors to the ground swell of calls to reform the US Food and Drug Administration (FDA) or the U.S. health care system (Angell, 2000, 2004a; Relman and Angell, 2002).1 This movement has become so strong that neither a formidable lobby nor a sympathetic administration in the U.S. likely will be able to shield the pharmaceutical sector from changes that could have profound effects on the industry and on its customers, namely all of us. In her last chapters Angell elaborates on the seven most important problems that she has discussed in the book, and then she proposes reforms. Here are the problems, mostly in Angell’s own words, but abbreviated in some cases. 1. Too many me-too drugs, and too few innovations. 2. FDA is too much in the thrall of the industry it regulates. 3. Drug companies have too much control over clinical research of their own 4. Patents and other exclusive marketing rights are too long and elastic. 5. Companies have too much influence over medical education about their products. 6. Important information about R &D, marketing, and pricing is secret. 7. Prices are too high and too variable. In this paper I focus on Angell’s proposals to reform drug development, that is, to provide solutions mainly for the first three problems. I first examine a matter that influences all of Angell’s comments about biopharmaceutical R&D, namely her answer to the question, who innovates? Then I examine two of Angell’s explicit proposals for reform in the area of drug development: her suggestion to establish, within the U.S. National Institutes of Health, an Institute for Prescription Drug Trials (IPDT), which would carry out all the clinical trials currently conducted by biopharmaceutical companies; and her proposal to require that new drugs be compared to existing drugs. My overall position is that being convinced of serious shortcomings in some areas of the industry – such as mammoth marketing budgets and offensive advertising – does not justify a blank check to potential reformers. Although I agree that Angell raises issues in drug development that need to be addressed, I will show that both of Angell’s reforms are Levy: Assessing drug development reforms…. fundamentally flawed. I conclude with some suggestions of my own about what should be the general thrust of reforms in drug development. 2 Who innovates?
Angell’s view is that the treatment of the pharmaceutical industry by government is based
on the erroneous assumption that pharma is the major source of discovery, development,
and innovation. Angell raises this point in several places including in her discussion of
reforms in general.
In thinking about reform, it is useful to consider the industry in terms of its functions. Which does it do well, which badly, and which shouldn’t it be performing at all? The industry supposedly discovers, develops, tests, manufactures, distributes, and promotes drugs. We have seen that it contributes much less to discovery and early development than it claims but instead feeds off the NIH, and universities and smaller companies in the United States and abroad. Maybe we should simply accept that fact. But then it makes no sense to continue to reward big pharma as though it were the major source of innovation. (2004b: 258) Angell is probably correct in her view that pharma claims to be more innovative than is justified and that the true origins of basic discoveries are located in universities, research institutes, and smaller biotech companies.2 On the other hand I believe that the biopharmaceutical sector contributes far more than Angell estimates. Part of the reason that Angell sees minimal contribution to innovation by pharma is that she tends to conflate the concepts of innovation and basic research, and to minimize the importance and magnitude of applied research that takes place in drug development. It may seem odd to say that initial discoveries can be overemphasized because quite clearly, without the beginning, there can be no subsequent development. However, frequently the process goes like this. A university researcher discovers a mode of action or a compound or a class of compounds that affects the action in question. The compound(s) will have been shown to have some effect perhaps in rodents, but the compounds have not come close to having been tested in humans. In fact, very little work will have been done on toxicity or getting the material in a form suitable for administration to humans (the formulation), much less in the area of drug optimization, the highly challenging process of tuning potential drugs and their delivery systems. At this stage, most often the discoveries, or more accurately the science plus applications to the U.S. Patent and Trademark Office covering the science, either are spun out into small biotech companies or are licensed by larger biopharmaceutical companies. Thus the correct interface on which to focus is between the public sector, such as universities or research institutes, and the private sector in the form of corporations. That is, for the purposes at hand, biotech companies should be regarded as part of the pharmaceutical industry.3 What happens once the discovery is licensed by the private sector and begins a preclinical development program aimed at convincing the FDA that the drug is sufficiently safe to be administered to humans? I cannot in a short article attempt to summarize the extent of toxicological, pharmacokinetic, pharmacodynamic, chemical, manufacturing, and other Levy: Assessing drug development reforms…. data that have to be gathered and submitted. One could try to convey the magnitude of the effort required by citing the resources, in terms of persondays and/or dollars that are committed. However, the financial estimates are questioned for a number of reasons, including the fact that some of the raw information is only available to certain researchers such as the group at Tufts, which negotiated with companies to assemble and provide the data (DiMasi et al., 2003). At the time of Angell’s writing, they calculate the average R & D costs required to develop a drug that is a “new molecular entity” to be in the range of $800 million. That figure includes not only what is spent on successful drugs, but also on failed drugs. In addition, and it is a big addition, the figure includes estimates of opportunity costs which essentially match the out-of-pocket expenses of around $400 million. Angell and others such as Goozner argue, for example, that the figures cited are inflated in a variety of ways (Goozner, 2004). Angell says, “We don’t know what activities are included under the heading ‘R & D’. Much of it may really be marketing …” (2004b: 39). I prefer to bypass this battle of numbers except for two comments. First, although Goozner entitles his book The $800 Million Pill: The Truth Behind the Cost of New Drugs, it is not until the last chapter that he seriously addresses estimates of development costs. And his main evidence is a study conducted by the TB Foundation that claims to use a methodology similar to the Tufts group and estimates the total costs to develop a TB vaccine to be $240 million (Goozner, 2004). However, a number of differences between the TB and other cases deserve further examination: one phase 3 trial instead of two; discount rates; and chronic versus acute conditions, to name a few.4 Second, the fact that biotech companies that have no drugs in the market and yet have huge expenses supports the view of two researchers in the US Federal Trade Commission who also replicate the DiMasi/Tufts study and conclude that their own … results support DiMasi et al's claim that the average cost to develop a new drug is over $800m while also suggesting that for some drugs the costs can be much higher or much lower. The results suggest that drug development costs can be influenced by numerous factors …The results further suggest that a great deal of care must be taken before using these numbers in public policy debates or interpreting them as a measure of actual drug development costs. (Adams and Brantner, 2004) I agree with this conclusion especially if it is emphasized that the estimates of drug development costs do not justify drug prices; the estimates provide an order of magnitude of what it costs to meet the current standards of FDA. Having said something about the quantity of effort that goes into drug development, I want to address some aspects of its quality. It is all too easy for academic researchers and commentators to dismiss the drug development process as being standardized applications of routine tests. Although standardized experiments are certainly involved, creative, resourceful research is often done and new knowledge is generated. The most convincing examples of such research are cases in which a biopharma takes up what is basically a striking idea accompanied by a modicum of data; then, in nonclinical investigations of modes of action and other features of compounds and by engaging in drug optimization, the researchers discover new effects or compounds. For example a key step in the development of the photoactive drug, Visudyne®, was the discovery by Levy: Assessing drug development reforms…. company scientists that there were new ways in which the drug is taken up by (abnormal) cells; without this discovery, it is unlikely that Visudyne would have been developed as a treatment for macular degeneration. However, I prefer to examine a more familiar case, which Angell also discusses, namely that of Gleevec, a breakthrough drug for the treatment of Chronic Mylogenous Leukemia (CML). From 1960 to 1992 a series of academic researchers showed that many patients suffering from CML have a (mutant) gene, Bcr-Abl, which expresses a tyrosine kinase, an enzyme involved in cell signaling that regulates cell growth and division. In 1992, Brian Druker, then a scientist at the Dana-Farber Cancer Institute in Boston but soon to move to the Oregon Health & Science University, was investigating just what the enzyme was doing that resulted in the massive overproduction of white blood cells that is the hallmark of CML. Previously, Druker had discussions with scientists at CIBA-Geigy about their drug discovery program focused on tyrosine kinases, so he reconnected with them. According to Druker, “’They had the idea to make compounds to inhibit tyrosine kinases, and we had the experience in reagents for detecting activity of tyrosine kinases. It was the perfect marriage’" (Vastag, 2000:7). It was this collaboration that led directly to the drug Gleevec. At the end of 1993 CIBA-Geigy scientists sent a half-dozen compounds to Druker’s lab to be put through his screening tests. Gleevec was the clear winner. For the next 4 ½ years CIBA-Geigy “jumped through all the hoops needed to start clinical trials: drug formulation, pharmacokinetic studies, initial toxicology screening” (Vastag, 2000:8). One of the key chemists at CIBA-Geigy, Jürg Zimmermann, was asked about his efforts to make the compound “more selective and drugable”: “Do you, as a medicinal chemist, use a toolbox of side chains of which you know the effects?” Zimmermann responded, In a way it would be nice to have such a toolbox, but then it would also take away a lot of the fun and fascination of my work. [W]hen we added a methyl group at a certain position, selectivity would go up but potency down… With every compound that we handed over to the biologists and with every result that we received back, we learned something. And this helped us design the next molecule. (Ziegler, 2004:22) Angell cites Gleevec as an example in which the drug company, Novartis, “had more than a little help” from academic researchers (2004b:63). That is undoubtedly true, as is the point, also dwelt upon by Angell, that Druker as well as some within Novartis had to lobby very hard to convince senior managers at Novartis to pursue the CML work because the small number of CML patients would not justify a full-scale development program. But neither fact denies either the quantity or the quality of Novartis’s contributions. Both of these points raise some questions about the function of profit-driven drug development, a topic to which I shall return in my concluding section.5 There is real science in the drug development process. A lot of knowledge about the particular compounds and their modes of action is generated as the development program is pursued interactively and iteratively among a half-dozen specialized labs.6 However, not nearly enough knowledge to avoid failure, which, as I discuss below, is the usual outcome of drug development. Levy: Assessing drug development reforms…. 3 Institute for Prescription Drug Trials (IPDT)
Because in her view pharmaceutical companies have too much control over clinical research, Angell proposes that … an Institute for Prescription Drug Trials be established within the National Institutes of Health (NIH) to administer clinical trials of prescription drugs. (2004b: 245) Based on their revenues, drug companies would be required to make contributions that would (help) underwrite IPDT, but the funds would go into a pool, thereby severing the direct link between a company and its drugs. The institute [IPDT] would then contract with independent researchers in academic medical centers to conduct drug trials. The researchers would design the trials, analyze the data, write the papers, and decide about publications. (Angell, 2004b: 245) Thus, IPDT, an independent public agency, would “administer all clinical trials to ensure that they are properly conducted – both scientifically and ethically,“ and “there would be no need for a private research industry” (Angell, 2004b:246, emphasis added). One motivation for IPDT: abuse of the Clinical Trial system There are several factors that motivate Angell to propose IPDT: avoid payoffs to physicians; improve trial design; achieve unbiased interpretation of data, etc. I will take a closer look at the allegation of there being large payments to physicians for conducting trials, and, in particular, a “bounty” system in which physicians are not only paid per patient, but also get a bonus for meeting or exceeding a quota: “Obviously, if the wrong patients are enrolled, the results of a trial are unreliable, and that is often the case” (Angell, 2004b:31). Obviously. But “often the case”? This may be one end of the spectrum of what occurs, but it is instructive to hear what can, and has, happened when enrollment is compromised. Consider the drug development example that has received the most press in the last 10 years, and perhaps ever: the Imclone/Martha Stewart case. The problem arose when the company received a letter from FDA about Imclone’s cancer treatment drug, ErbituxTM. Imclone’s president was eventually convicted of insider trading because he contacted several people, allegedly including Martha Stewart, to warn them of pending, as yet undisclosed news that would have a major negative effect on the company’s stock. And what was the main problem cited in the FDA letter? It was enrollment in clinical trials: FDA asked whether Imclone could actually document that the patients enrolled in the trial met the inclusion criteria; if they couldn’t, then positive results might not have been due to Erbitux. Here are the points that this story would illustrate if filled out in full detail: clinical trial protocols are vetted by regulatory authorities and by review boards at each site; the trials are conducted under Good Clinical Practices (GCP), which involve strict monitoring and documentation; all the trial records are subject to audit by regulatory authorities, etc. Of course these measures are not enough to ensure that all trials are properly conducted. But let’s look at what the new Institute brings to the table. Levy: Assessing drug development reforms…. IPDT takes over all clinical trials, so that Institute members and independent researchers serving as contractees design, implement, and interpret the trials. The first question is, will FDA have the same responsibilities of oversight, regulation, and audit that it does now? I assume that it will, because I see absolutely no reason to cut out FDA; most of its functions must be performed no matter who conducts the trials. So the new system includes FDA and to a large extent the contractees are the same: independent physicians in medical centers and private practice. Angell’s approach differs from the current system by leaving out the companies, and, in particular, the large amount of funds that she alleges corrupts the system.7 One could argue against Angell’s approach by pointing out that although company money would not be available to the contractees who are currently the investigators in trials, there are other kinds of benefits that already exist and would become even more valuable currency. Or one could point out that even well intentioned government agencies can be culpably unresponsive, inefficient, or worse. But I will approach Angell’s version of IPDT by asking what may appear to be an irrelevant question: Why do so many drugs fail even after they have begun clinical trials? Years of work and millions of dollars go into the development of a drug before it is actually tested in humans in clinical trials. And yet the overwhelming majority of drug candidates that begin clinical trials will fail to make it all the way through to approval. According to FDA statistics roughly one in four or five drugs that begin clinical testing actually end up on the market. Why do more than 75% of the drugs fail at the clinical stage of drug development?8 Sometimes trials are poorly designed or poorly administered and sometimes there are economic or market reasons for terminating a program, but the main factor is that not enough is known about the drug or the disease or how the drug, its breakdown products and metabolites behave within the enormously complex system that is the human body (Kola and Landis, 2004). That is, drugs are largely defeated by Nature, or rather our lack of knowledge about Nature.9 If the current clinical trial system were as subject to corporate manipulation as Angell suggests, then to the failures of the companies one should add gross incompetence at cheating. 4 Reducing the numbers of me-too drugs: require comparison
with standard therapy
The reform that Angell regards as the singularly most important one is aimed at the first problem, namely, that there are too many me-too drugs and too few innovative ones. Angell proposes “[FDA] regulations should require that new drugs be compared not just with placebos but with old drugs for the same conditions” (2004b: 240).10 In discussing the proposal Angell includes the requirement that FDA “not approve drugs that on balance offer trivial or no advantages over drugs already available” (2004b: 240-1). In Angell’s view, raising FDA’s bar would not only lower the number of me-too products, but also free up drug development and regulatory review capacity so that more resources can be devoted to more innovative prospects. In addition, with fewer me-too products on the market, sales and marketing budgets could be reduced and more could be Levy: Assessing drug development reforms…. devoted to R & D. Or so Angell suggests, but provides no evidence for. Still, the proposal has intuitive appeal. The Me-too issue: one problem, many, or none? Although the notion of a me-too drug may seem obvious, actually there is no generally accepted definition of what counts as a me-too drug and the different versions of the concept(s) can cause significant confusion. Angell initially characterizes me-toos this way: The stream of new drugs has slowed to a trickle, and few of them are innovative in any sense of that word. Instead, the great majority are variations of oldies but goodies – “me-too” drugs. (2004b: 16) The basic ideas underlying the me-toos are drugs that are non-generic and clinically substitutable. The members of a class of me-toos fall under different patents, so they are engaged in what has been called “between-patent” competition in contrast to the “within-patent” rivalry of a generic and its original (Lichtenberg and Philipson, 2002). Besides patent coverage, members of a class can differ with respect to non-active ingredients, dosage, route of administration, etc. Although these remarks clarify the concepts somewhat there still can be room for disagreement about what counts as me-toos. Angell and others make a distinction between two kinds of me-toos, though they don’t use the terminology I employ: (a) first, there are minimally competing successors, which are drugs where the original and the successor come from the same company; and (b) the second kind of me-toos are developed by different companies and are thus competing me-toos. (a) Angell presents what she considers to be odious recent examples of minimally competing me-toos, i.e., cases in which companies have threaded the needle by creating a product close to their existing one, but far enough away to obtain a new patent. These include the pairs Prilosec®-Nexium®, used for the treatment of heartburn, and Claratin®- Clarinex®, which are anti-allergy medications. In the first case, Prilosec is a mixture of two forms or isomers of the same molecule, whereas its successor, Nexium, consists of only one of the isomers. And the makers of Claritin discovered that the therapeutic effect of the drug could be achieved by using a metabolite of Claritin, which metabolite is now marketed as Clarinex. In each case since both the original and the successor are owned by the same company, they can take steps to minimize, if not eliminate competition between the two. Here I am not going to assess either Angell’s claims that the advantages of the newer drugs are vanishingly small or the companies’ counterclaims. However, it is important to point out that there are examples where making what may appear to be insignificant changes such as eliminating an isomer or switching to a metabolite of a drug makes a significant difference. As an illustration of how significant an isomer difference may be consider the case of thalidomide. As is well known thalidomide was responsible for a large number of birth defects in the early 1960s. It is a mixture of two isomers; after the fact it was discovered that one isomer is responsible for the beneficial effects and the other, the horrendous side-effects. Separating the isomers does not work, however, because when injected, the remaining good isomer is quickly metabolized into a mixture of the two.11 There are, however, many examples of drugs that are in the same class and Levy: Assessing drug development reforms…. yet have significantly different properties. I mention some of those in the following discussion of competing me-toos. (b) Competing me-toos are the more numerous. One of the most detailed analyses of these is a recent work by DiMasi and Paquette (2004). They examine all drugs approved by FDA between 1960 and 1998, identify 72 drugs that in their view qualify as breakthrough or first-in-class, and then the ones that are second, and subsequent entrants to the respective classes. The authors make these claims about these follow-on scenarios, where “follow-on” is the more value-neutral term the authors prefer: … the period of marketing exclusivity that the breakthrough drug enjoys in a new class has fallen dramatically over time (a median of 10.2 years in the 1970s to 1.2 years for the late 1990). Approximately one-third of follow-on new drugs received a priority rating from the US FDA. The vast majority of the follow-on drugs for drug classes that were created in the last decade were in clinical development prior to the approval of the class breakthrough drug. (DiMasi and Paquette, 2004: 1) In addition, the authors (2004) found that “in a substantial number of cases in recent periods, the first drug in class to reach the US marketplace was not the first to enter clinical testing either in the US or anywhere in the world” (DiMasi and Paquette, 2004: 8). These results lead the authors to conclude, “… development races better characterise new drug development than does a model of post hoc imitation. Thus the usual distinctions drawn between breakthrough and ‘me-too’ drugs may not be very meaningful” (DiMasi and Paquette, 2004: 1). As mentioned above, Angell, Goozner and others have complained that the data underlying analyses of costs of drug development were not available to critics; in that case significant amounts of information were provided to the authors by companies on a confidential basis. That charge does not apply to the analysis of follow-ons, since the major results are drawn from data available in public sources or in commercial databases.12 However, even if these results are borne out – and I suspect they will be13 – they do not settle the entire question whether me-toos are a serious problem for the current drug development system. As the authors themselves say, “A full assessment of the social rate of return to [competitive] follow-on drug development must account for any clinical and economic benefits [or losses] it engenders” (DiMasi and Paquette, 2004:11).14 That is, a full social evaluation would have to take into account not only minimally competing follow-ons, but also such factors as the relative clinical contributions of all follow-ons.15 This latter consideration could end up going either way: some follow-ons will no doubt show only minor or no differences from the breakthrough or from each other, but there are others that will differ significantly. For example, the first class of drugs approved for the treatment of HIV/AIDS was nucleoside reverse transcriptase inhibitors. Today individual patients may be treated with several of these in combination precisely because they have slightly different, complementary characteristics. Levy: Assessing drug development reforms…. My comments about “ideal” drug development are of course open to a general challenge: one could maintain that the market, though imperfect, is still a better arbiter of needs than the clumsy, visible hand of government; and that follow-on drugs are simply an inevitable by-product of the profit-driven drug development system, just as multiple, similar products occur in every other industry. Any rebuttal would point to the fact that biopharmaceutical products are a very special case because they can often mean the difference between life and death, or between great suffering and much less. In fact, drug development is regarded as so important that government already acts as a major arbiter with respect to safety, efficacy, and promotion, so why shouldn’t stiffer rules be put in place to improve the character of the pharmacopoeia? Even if the rebuttal wins the day, those on the side of Angell do not get carte blanche to make changes to the system. I believe that the Tufts group and others show that the problems may not be exactly as depicted and, most importantly, they show the complexity that any proposed solutions would have to face. Assessing Angell’s proposal: comparators, indications, and timing FDA regulations already require that wherever possible new drugs should undergo comparative testing, which usually means that at least the last stage of clinical trials, phase 3, are conducted with one or more control groups. There are various options for what the control groups can be, including different doses of the new drug itself, but usually new drugs are compared either to the current standard of care or to a placebo. As Angell points out, …drug trials for serious diseases, like cancer or HIV/AIDS, almost never have placebo groups. Instead the new drug is compared with whatever is currently being used. But most new drugs are not for serious diseases. (2004b: 241) So the problem as Angell sees it is mainly to do with drugs for non-serious conditions. However, according to the current regulations the choice of control group does not hinge on whether the condition is serious or not; if there is an extant standard of therapy for the (exact) condition in question, the new drug is supposed to be compared to it. What challenges would the drafters and implementers of Angell’s regulation face? First, there is the way that FDA and all other regulatory bodies view the universe of disease. That universe is composed of a multitude of quite narrow strips called “indications”. Drugs receive FDA approval to treat a particular indication, and it is only approved indications that are permitted on the label and in promotional activity such as advertising. Indications divide the disease space both by type and stage, and, in addition, take into account the patient’s treatment history. For example, a drug will receive approval for a certain type and stage of bladder cancer such as stage 2 papillary bladder cancer, and its approved use will usually specify whether other treatments should take place first. However, FDA does not regulate the practice of medicine; FDA does not have the mandate to restrict prescribing physicians or consumers to use the drug to treat only the approved indication.16 The significance of narrow indications is that drug sponsors can, in some circumstances, seek an approval for a new drug for a somewhat different indication than the breakthrough drug, and later expand the indication or go with off-label use. Thus, a great deal hinges on what counts as “the same condition” in Angell’s proposal. Levy: Assessing drug development reforms…. Second, the previous point can be combined with the matter of timing. Certainly, companies with follows-ons that are already in the clinic – especially those that are in comparative (phase 3) trials – should not be required retroactively to redesign and restart trials when a breakthrough product in its class is approved. But then companies would have even more incentive to accelerate drug trials as a means of avoiding Angell’s condition. Third, there are myriad challenges associated with unpacking Angell’s enjoinder not to approve new drugs that offer no or only trivial advantages. Imagine that a breakthrough and a follow-on drug were in a randomized, comparative trial and the results show that the nature, timing and frequency of both safety and efficacy results are the same. This ideal case would not conclusively establish that the follow-on should be rejected. In addition to having the possibility of price competition and the security of two different manufacturing processes (recall the recent vaccine shortages), there is also the possibility that even the relatively slight physical differences between the two drugs actually have treatment differences. For example, even though patients are randomized with respect to known factors, this does not guarantee that the characteristics of patients benefiting from the breakthrough drug are the same as those from the follow-on; the two could actually serve different sub-populations, or, as is the case with nucleoside reverse transcriptase inhibitors, have added benefits when used in combination. In pointing to these challenges, I am not saying that they are insurmountable. In fact in the current regulatory system such judgments have to be made comparing drugs, e.g. whether a particular candidate is non-inferior to another drug. I think Angell’s view that we need more comparative trials is correct. The issue is where and how that should be accomplished. Macro-picture: Expanding Mandates Angell’s proposed reform is a member of a class of suggestions to broaden the types of scrutiny that drugs undergo. There are several elements to such proposals and they can be combined in various ways: • Timing: pre-approval or post-approval. • Agency: FDA or other. • If FDA, is this a major mandate extension or not? Angell’s proposal is a requirement that applies pre-approval, is to be carried out by FDA, but does not represent a major expansion of that agency’s mandate which is supposed to be centered on science-based determinations of safety and efficacy. To illustrate a contrasting approach, consider the “Prescription Drug Comparative Effectiveness Act of 2003,” (H. R. 2356) which was introduced in the U.S. Congress but not passed, and which would … authorize spending $75 million by the National Institutes of Health and the Agency for Healthcare Research and Quality [AHRQ] on studies of the "comparative effectiveness and cost-effectiveness of prescription drugs that account for high levels of expenditures or use by individuals in federally funded health programs." (Mulligan, 2003: 8) Levy: Assessing drug development reforms…. This Bill would mandate the study of economic factors and comparative effectiveness after a drug is already on the market, and the primary agencies involved would not be FDA. Congressional proposal 2356 aims to broaden NIH’s purview by linking activities with AHRQ and adding cost-effectiveness to areas investigated, thus in principle providing additional government-generated data to pricing decisions.17 I bring up this Bill not to endorse its specifics, but to indicate the breadth of options available. 5 Prolegomena to Alternative Reforms
The two main avenues that should be pursued for drug development reform in the U.S.are (a) some systematic consideration of comparative economic and health impacts, and (b) a greatly increased role for needs-driven medicine to complement profit-driven medicine. I won’t address (a) here, except to say that even though the U.S. is the only developed country that does not employ price regulation, I don’t see broad-based regulation as achievable in the U.S. in the foreseeable future. But there could well be a point in the near term when there is irresistible impetus to introduce some strong, credible mechanisms for ascertaining the cost effectiveness and comparative health effectiveness of medications, and a way to translate those factors into policies in the delivery of health care. Under such circumstances there could be a place both for tightening pre-approval requirements with respect to comparators and for an Institute for Prescription Drug Trials, but one having a mandate and role similar to those proposed in Bill 2356 rather than the mission Angell proposes. Such an IPDT should conduct trials on any drugs used in practice and should develop standardized economic endpoints that would supplement safety and efficacy ones. Its effects would surely extend to pre-approval drug development strategies. I wish to spend my remaining space on (b), alternatives that complement and supplement profit -driven medicine. I think it useful to recognize that governments have enacted other measures whose goal is explicitly to fill a gap in market mechanisms. One of the clearest examples is the Orphan Drug program, which was first legislated in the U.S. in 1983. Although in aggregate there are estimated to be 20,000,000 people in the U.S. who suffer from what are called orphan diseases, the numbers who suffer from a particular condition is so small that the investment required to develop a medication could never be recovered (DHHS, 2001). An Orphan Drug program provides incentives to companies willing to develop specific medication to treat orphan diseases. Orphan Drug programs establish the principle that governments can effectively intervene in drug development to ameliorate market failures. Currently, needs-driven drug development is undertaken mainly by organizations aimed at the developing world such as the non-profit Institute for One World Health and the Drugs for Neglected Disease initiative (DDNi), an NGO created by Doctors without Borders and kindred spirits. What are needed are initiatives either to develop drugs de novo or to adapt existing drugs to treat underserved conditions. Even though these latter may be responding to incentives, in my view the needs-driven agenda is furthered, to varying degrees, by all instances in which some quantum of short-term profit is traded off in favor of meeting social needs.18 The drug development system has provided a fairly steady stream, certainly more than a trickle but less than a torrent, of safe and effective drugs.19 In addition to the expertise Levy: Assessing drug development reforms…. and experience in drug development that reside in the biopharmaceutical industry, the traditional strengths of for-profit enterprise – e.g. risk-taking, opportunism, and competitiveness – have something to offer, as long as they are well regulated and market failures are creatively addressed.20 Better to conduct the experiment of having for-profit and not-for-profit systems operating side-by-side, than creating one chimera that could fail and do untold damage in the process. As Angell points out and everyone knows, reforms will be difficult to effect: “I will propose reforms even though I know the obstacles to achieving them are formidable” (2004b: 238). Better not to dissipate energies on half-measures that industry can oppose not just because of its well-honed sense of self-interest, but also because it’s best to preserve and extend the positive contribution industry makes. Reforms are surely coming. Let’s ensure that they are the right ones. Levy: Assessing drug development reforms…. References
Adams, C., & Brantner, V. V. (2004). Estimating the Costs of New Drug
Development: Is it really $802m? URL: http://ssrn.com/abstract=640563. Angell, M. (2000). The pharmaceutical industry--to whom is it accountable? Angell, M. (2004a). The Truth About the Drug Companies. The New York Angell, M. (2004b). The truth about the drug companies: how they deceive us and what to do about it (1st ed.). New York: Random House. DHHS. (2001). The Orphan Drug Act: Implementation and Impact (No. OEI-09- DiMasi, J. A. (2005). Comment on “Me-too drugs: is there a problem?”. URL: http://www.who.int/intellectualproperty/forum/en/HollisResponse.pdf. DiMasi, J. A., Hansen, R. W., & Grabowski, H. G. (2003). The Price of Innovation: New Estimates of Drug Development Costs. Journal of Health Economics, 22, 151-185. DiMasi, J. A., Hansen, R. W., & Grabowski, H. G. (2004). Assessing claims about the cost of new drug development: a critique of the Public Citizen and TB Alliance reports. URL: http://csdd.tufts.edu/_documents/www/Doc_231_45_735.pdf. DiMasi, J. A., & Paquette, C. (2004). The Economics of Follow-on Drug Research and Development: Trends in Entry Rates and the Timing of Development. Pharmacoeconomics, 22(2), 1 - 14. Flieger, K. (2000). Testing Drugs in People. URL: http://www.fda.gov/cder/about/whatwedo/testtube-3a.pdf. Goozner, M. (2004). The $800 Million Pill: The Truth Behind the Cost of New Drugs. Berkeley, CA: U California Press. Hollis, A. (2004). Me-too drugs: is there a problem? URL: http://www.who.int/intellectualproperty/topics/ip/en/Me-tooDrugs_Hollis1.pdf. Kaitin, K., Bryant, N., & Lasagna, L. (1993). The role of the research-based pharmaceutical industry in medical progress in the United States. Journal of Clinical Pharmacology, 33(5), 412-417. Kola, I., & Landis, J. (2004). Can the pharmaceutical industry reduce attrition rates? Nature Rev. Drug Discov. 3, 711–715 (2004). 2 . 3(2), 711-715. Lichtenberg, F. R., & Philipson, T. J. (2002). The Dual Effects of Intellectual Property Regulations: Within and Between Patent Competition in the US Pharmaceutical Industry. URL: http://papers.nber.org/papers/w9303.pdf. McNeil, D. G. (2005, April 12, 2005). Clinton Starts AIDS Drug Plan. NY Times. Mulligan, K. (2003). Bill Calls for Multimillion-Dollar Study Of Medication Cost- Effectiveness. Psychiatric News, 38(18), 8. NIH. (2004). NIH Halts Use of COX-2 Inhibitor in Large Cancer Prevention Trial. URL: http://www.nih.gov/news/pr/dec2004/od-17.htm. Relman, A. S., & Angell, M. (2002). How the drug industry distorts medicine and politics. America's Other Drug Problem. The New Republic, 32(4), 47-61. Levy: Assessing drug development reforms…. Vastag, B. (2000). Leukemia Drug Heralds Molecularly Targeted Era. Journal of the National Cancer Institute, 92(1), 6-8. Ziegler, S. (2004). Jürg Zimmermann: Extraordinary Contributions to Oncology Research. NIBR Science, Winter 2004, 20-23.
Endnotes
1 I follow Angell in using “FDA” even though in many instances it would be more accurate to refer to 2 See for example the frequent citations of (Kaitin et al., 1993), where the pharmaceutical industry is said to be the source of 92.3% of the new chemical entities approved by the FDA from 1981 to 1990. This is based on patent holders and may be accurate in some narrow sense. Companies may patent the exact drug approved, but that may often be a result of the industry optimizing a lead obtained from academia or government. (See note 5 below.) 3 This is why I use the term “biopharmaceutical” to include both. 4 For a detailed rejoinder, see (DiMasi et al., 2004). 5 In this case the source of the actual compound was big pharma, but it is misleading to count pharma as the sole source of this innovation. And the counting is also confounded when small biotechs are part of the process. 6 This sort of work could in principle be carried out in academic or government labs, though very few academic labs are set up to comply with the extremely heavy requirements of working according to Good Laboratory Practices (GLP) as defined by FDA regulations. 7 Angell says that “Clinical testing could continue to be the industry’s responsibility,” but this is a rather peculiar sense of responsibility – basically, the companies put dollars into a pool and the trials are done by IPDT (2004b:258). Previously Angell said, The FDA now assigns responsibility for the conduct of clinical trials to sponsors. That practice would end. Responsibility would lie exactly where it should – with independent researchers and their institutions. (2004b: 245) 8 Of 100 drugs for which investigational new drug applications are submitted to FDA, about 70 percent will successfully complete phase 1 and go on to phase 2; about 33 percent of the original 100 will complete phase 2 and go to phase 3; 25 to 30 of the original 100 will clear phase 3; and about 20 of the original 100 will ultimately be approved for marketing (Flieger, 2000). 9 It may appear that there is an inconsistency in my presentation. Earlier I argued that when a drug candidate is taken up by biopharma a great deal of knowledge is generated; and yet here I am saying drugs fail because of a lack of knowledge. But there is no inconsistency; there is just an enormous amount to know about drugs and how they operate. 10 I am using the term ‘drug’ to include what are, in technical regulatory terms, drugs, biologics, and medical devices. By and large the latter have less stringent approval requirements. 11 Clearly, the Thalidomide isomers are not examples of me-too drugs. The case is cited solely to illustrate what difference isomers can make. It’s also worth noting that although thalidomide was not approved in the US in the 1960s when the teratogenic effects were manifested, it was approved in the late 1990s in the US and elsewhere for the treatment of lesions associated with Hansen’s disease, which is commonly known as leprosy. The approvals were accompanied by extraordinary restrictions and requirements; other applications are under investigation. 12 The exceptions concern detailed information about the timelines for the development of specific drugs. These may be proprietary, but the major milestones are a matter of the public record. 13 Hollis has criticized mainly the economic analyses, but some other aspects as well (Hollis, 2004). DiMasi has responded, largely successfully (DiMasi, 2005) 14 Although the authors‘ comment about “the usual distinctions drawn between breakthrough and ‘me-too’ drugs may not be very meaningful” sounds like an attempt to dissolve the whole problem, it is clear from the rest of the paper that their conclusions are more circumscribed (DiMasi and Paquette, 2004:1) Levy: Assessing drug development reforms…. 15DiMasi and Paquette make some circumstantial comments on clinical importance. They point out that these are based on FDA’s rating system, but as they point out, that system is a tool for managing FDA’s workflow, rather than a definitive assessment of clinical safety, efficacy, and other factors (DiMasi and Paquette, 2004). 16 There are of course other factors that exert pressure to restrict use to approved indications, e.g., matters 17 Recent news has made it clear that NIH already has the power to conduct studies: on December 17, 2004 NIH announced that it was suspending a trial with VioxxTM that was being conducted by NCI (NIH, 2004). This has led to Merck’s withdrawing Vioxx from the market and to reviews of the entire class of COX-2 inhibitors by multiple agencies. 18 See the websites for Drugs for Neglected Diseases initiative (http://www.dndi.org/) and OneWorld Health (http://www.oneworldhealth.org/). Of course efforts by governments, agencies of the United Nations, companies, and foundations to subsidize drug distribution in the developing world indirectly encourage needs-driven drug development. In effect many of these initiatives expand markets for drugs. For example, the Clinton Foundation is partnered with a number of governments and agencies to purchase at greatly reduced prices medication for the treatment of HIV/Aids (McNeil, 2005). 19 This year, 2005, is the 25th anniversary of Terry Fox’s Marathon of Hope, which was an attempt by a one-legged man to hop/run across Canada from the Atlantic to the Pacific in an effort to raise money for cancer research. Fox had lost his leg to cancer, and subsequently, his life at the age of 26. His marathon was cut short because of a recurrence when he had gone 3300 km. Someone suffering from his condition today would have a much greater chance of keeping his leg and of surviving. No doubt that improvement owes a great deal to basic research, which Fox helped fund by raising hundreds million dollars, but the contemporary drug development and regulatory systems deserve some credit for such advances. 20 “Well regulated” includes both expanding some requirements and introducing new ones. Examples include the need for more complete and timely disclosure of information about ongoing clinical trials and of results. Levy: Assessing drug development reforms….

Source: http://www.ethics.ubc.ca/workingpapers/cae/cae003.pdf