We should not distinguish between symptomatic and disease-modifying treatments in Alzheimer s disease drug development

Transcription

1 Alzheimer s & Dementia 4 (2008) S21 S25 We should not distinguish between symptomatic and disease-modifying treatments in Alzheimer s disease drug development Rachelle S. Doody* Baylor College of Medicine, Houston, TX, USA Abstract Keywords: The terms symptomatic and disease-modifying have become standard in discussions of Alzheimer s disease therapeutics, yet there is little justification for their use. Currently marketed drugs are presumed to be symptomatic because they lead to some degree of mean improvement over baseline and because of the widespread belief that their mechanisms are limited and their effects are completely reversible. Current trial methodologies cannot distinguish between symptomatic and disease-modifying effects. Furthermore, it is highly likely that many trials will demonstrate a combination of such effects at the level of the trial or at the level of the individual. The forces that drive this distinction are largely social, as opposed to scientific. It would be preferable for drugs to first seek an antidementia claim, preferably on the background of conventional therapies when possible, in trials that are as small and as short in duration as possible. Further refinements would come by demonstration of how substantial and how enduring the antidementia benefits are The Alzheimer s Association. All rights reserved. Alzheimer s disease; Clinical trials; Symptomatic treatment; Disease-modifying treatment 1. Definitions of symptomatic and disease-modifying therapies are unclear Commonsense notions of symptomatic relief mean that a patient should feel or function better after taking a therapy. Medical practice is largely predicated on this expectation. Over time, symptomatic treatments have also come to be contrasted with curative or disease-modifying treatments, with the connotation that symptomatic treatments make the patient feel better without really modifying the process that causes the symptoms and that they are therefore in some way illusory. In self-limiting conditions such as viral syndromes, a symptomatic effect might be all that is required. But in chronic or progressive conditions we have grown to expect more powerful interventions that alter the long-term outcome or natural history of the disease. Ironically, therapies that improve symptoms are sometimes perceived negatively because of the assumption that if their effects are clearly visible in the short-term, they must have a symptomatic, and therefore limited, overall impact on the disease. *Corresponding author. Tel.: ; fax: address: rdoody@bcm.tmc.edu Disease-modifying treatments are highly desirable but difficult to develop. In a phasic illness such as some forms of multiple sclerosis, the initiation of a disease-modifying therapy during an episode might improve the long-term outcome without alleviating symptoms. If initiated between episodes, long-term observation that controls for phasic alterations (variability) but still measures cumulative longterm gains is essential to prove a treatment effect, but it is logistically challenging. The situation in chronic progressive multiple sclerosis or Alzheimer s disease (AD) is even more challenging because patients, both treated and untreated, naturally accumulate more symptoms over time, and the symptoms become more severe over time. Therefore, a disease-modifying therapy could conceivably neither relieve current symptoms nor prevent the emergence of new symptoms, while still altering the long-term outcome of the disease. In other words, a disease-modifying therapy might have no discernable effects on the patient s current symptoms and might not prevent the patient from getting worse, but it would still be valuable. The concept of disease modification also encompasses therapies that do not alleviate symptoms but do prevent future symptoms from developing /08/$ see front matter 2008 The Alzheimer s Association. All rights reserved. doi: /j.jalz

2 S22 R.S. Doody / Alzheimer s & Dementia 4 (2008) S21 S25 This discussion of symptomatic versus disease-modifying therapies assumes that for each therapy, all of its mechanisms of action are clearly understood, and we can a priori classify each as symptomatic or disease-modifying. A further assumption is that any successful clinical trial can be interpreted as showing either a symptomatic or disease-modifying effect. These assumptions are clearly not the case for current AD therapies or for those under development. 2. Currently approved AD drugs and treatments under development for AD cannot be definitively classified as symptomatic or disease-modifying 2.1. Currently approved AD drugs Approved AD therapies include the cholinesterase inhibitors and memantine. The Food and Drug Administration (FDA) approved package inserts for the cholinesterase inhibitors all mention their postulated mechanism(s) of action, acknowledge that their mechanisms of action are unknown, and mention that there is no evidence that [drug x] alters the course of the underlying dementing process. For memantine, the FDA-approved language mentions memantine s postulated mechanism of action and that there is no evidence that memantine prevents or slows neurodegeneration in patients with AD, although one is hard-pressed to imagine how it could be effective through an antiglutamatergic mechanism while not protecting cells. According to the Physicians Desk Reference (2007), these drugs are indicated for the treatment of mild to moderate (cholinesterase inhibitors) or moderate to severe (memantine) dementia of the Alzheimer s type, although donepezil has gained subsequent FDA approval for severe AD. When a report was published indicating that tacrine could improve the manifestations of AD [1], it was met with widespread skepticism. Eventually after multiple studies, some scientists and the public came to accept the idea that tacrine could alleviate the symptoms of AD in at least some subjects. Subsequent studies designed to test donepezil were specifically designed to demonstrate a reversible, symptomatic effect by including a washout period, in part to counter any concerns regarding overclaiming of benefits such as those that arose during the development of tacrine [2,3]. The end of study washout periods planned into the donepezil studies failed to show a washout of donepezil s effects on multiple measures at 3 weeks [2] and showed inconsistent results at 6 weeks because the cognitive portion of the Alzheimer s Disease Assessment Scale (ADAS-cog) effects were fully reversed by the end of the washout period, whereas Clinician Interview Based Impression of Severity (CIBIC) plus and Clinical Dementia Rating Sum of the Boxes (CDR-SB) effects were not fully reversed [3]. Yet these washout data, obtained from a single blind washout design in only two studies, are widely cited as evidence that donepezil is a symptomatic and not a disease-modifying drug, and this view is often extended to the whole class of cholinesterase inhibitors. In an attempt to guide future demonstration of diseasemodifying effects, one author suggested using randomized start and/or withdrawal maneuvers in parallel group efficacy studies [4]. Although these study designs were not used for any of the currently approved therapies, crossover observations regarding placebo patients in pivotal trials who went on treatment in open-label extension studies consistently showed that those with a delayed start did not catch up, although catching up would be expected in a merely symptomatic treatment [5 7]. None of the authors of these crossover studies claimed that the studies supported disease modification because concerns about differential attrition and incomplete enrollment into open-label studies make the data too unreliable, but the consistency of this finding across studies and across drugs bears notice. The preclinical models for AD treatments in use during the 1970s and 1980s differ from those used today. Naturally on the basis of their presumed mechanisms and diseaserelated models that were available at the time, cholinesterase inhibitors were studied in cholinergic lesion and aged animals, as opposed to amyloid overproducing transgenics or model systems that simulate amyloid toxicity or tau protein hyperphosphorylation.. Subsequent basic science research has pointed to an intriguing array of effects for cholinesterase inhibitors and memantine in models more relevant to our current understanding of the pathophysiology of AD [8 10]. Although these experiments and hypotheses do not prove new mechanisms of action for current antidementia drugs, they raise possibilities that these drugs have additional effects, blurring the distinction between strictly symptomatic versus disease-modifying effects Treatments under development for AD Treatments currently under development are even more likely than approved drugs to have unclear mechanisms of action, because they tend to be studied extensively in multiple disease-related preclinical models. The approaches in Table 1 are divided into groups loosely based on presumed mechanism of action and derive from basic experiments as well as epidemiologic observations. The presumed mechanisms listed are derived solely from information in the public domain and are included to represent the scope of mechanisms under study to treat AD. The table does not attempt to compare and contrast these agents, and some drugs might claim more than one mechanism. Although most of the companies developing these agents will emphasize those mechanisms thought to alter disease progression, none of these treatments have been tested well enough in humans to rule out an improvement in symptoms, ie, a symptomatic effect. Arguments can certainly be made to support the possibility that some of these agents will slowly but surely

3 R.S. Doody / Alzheimer s & Dementia 4 (2008) S21 S25 S23 Table 1 Possible disease-modifying classes of drugs Class of drug Presumed mechanisms Possible claim to disease modification? Possible improvement of symptoms Active monoclonal vaccines Induce clearance of beta amyloid by activating, likely microglia and shifting amyloid out of the brain Passive immunization Induce clearance of beta amyloid by binding and, likely clearance Secretase inhibitors Reduce the production of beta amyloid If effect is complete and rapid enough Anti-fibrillization or Reduce the polymerization and 3-D If effect is complete and rapid enough deposition agents conformational changes Glial cell modulators Induce clearance of beta amyloid and/or reduce its production and/or inflammatory response Neurotransmitter Counteract transmitter imbalances while creating, likely modulators anti-amyloid or neuroprotective effects Neuroprotective or Introduce protective peptides or metabolic Unlikely neurotrophic agents regulators of energy metabolism, amyloid production, or tau hyperphosphorylation Anti-tau drugs Reduce the likelihood of tangle formation, likely Drugs based on risk factors Shift sex hormonal balance, lipid metabolism, glucose metabolism, or energy metabolism Neutraceuticals Alter oxidative mechanisms, block amyloid fibril formation improve the brain while not improving the patient s symptoms. Presuming that some of these therapies will indeed alter the natural history of AD without improving symptoms, the question remains how we will know, in the context of a clinical trial, that this effect is present. Proposals for picking up disease-modifying effects abound [11 13], and companies with clinical trials underway are struggling to produce a convincing portfolio of evidence to support such claims. Even if global regulatory authorities had decided on and published clear targets to prove disease modification, the status of our understanding of how AD progresses is too incomplete to predict success. 3. Symptomatic and disease-modifying effects are not mutually exclusive The natural history of AD is poorly understood, although some observational studies suggest that there are cohort effects over time [14], that progression rates differ between patients [15], and that treatment with conventional antidementia drugs alters group progression rates [16 18]. Because many ongoing studies and most future pivotal trials will be testing drugs on the background of conventional treatments, time since diagnosis, month of enrollment, initial rate of decline, baseline duration, and adjustments during the study of marketed treatments could all introduce real effects into the studies that must either be controlled for or could potentially alter the outcome. These effects would be expected to impact both slopes of decline and mixed effects models. Therefore, an effective disease-modifying drug might appear to improve patients, hold them stable, or slow their decline compared with placebo, depending in part on these other factors. In this example a combination of symptomatic and disease-modifying effects would be due to the effects of different drugs in the same trial. There is also no a priori reason why putative diseasemodifying therapies should not improve patients symptoms, even in monotherapy trials. Most experts agree that disease-modifying therapies need to be introduced early to deliver the most desirable lifelong benefits [19]. But some have even argued that disease-modifying therapies, particularly anti-amyloid therapies, will only be effective early in the course of AD through a subtle effect of slowing progression, without causing any improvement over baseline. This assumption seemed to weigh heavily in the Phase 3 development programs for Alzhemed (tramiprosate) and Flurizan (tarenflurbil). In contrast, on the basis of the rapid onset of effects in vitro and in animal models for many of these agents, symptomatic effects are likely if the agents are effective to the degree that preclinical studies would predict. An agent that quickly and effectively blocks beta amyloid metabolism from the precursor protein, for example, could pervasively alter brain amyloid levels, shift intracellular and extracellular pools of beta amyloid, deter fibril formation, and alter the balance between deposition and clearance mechanisms. A strategy that induces rapid and substantial clearance of amyloid, eg, immunization, would be even more likely to show symptomatic benefits. Why would we not expect some symptomatic improvement under these conditions? By designing and powering clinical trials with putative treatments on the assumption that there will be no clinical improvement, we advantage the drugs under study

4 S24 R.S. Doody / Alzheimer s & Dementia 4 (2008) S21 S25 because we power the studies for a smaller drug-placebo difference. 4. The forces driving a firm distinction between symptomatic and disease-modifying treatments are neither clinical nor scientific Several interest groups have, knowingly or unknowingly, advocated for the distinction between symptomatic and disease-modifying therapies in AD. Some policy makers and others entrusted with ensuring truth in advertising want to make the public aware of the limitations of AD treatment. In addition to making sure that potential side effects are disclosed, these groups and individuals have emphasized the symptomatic nature of treatments to warn the public that treatments do not arrest or reverse AD. Although these disclosures should be part of the public discussion, this approach has occasionally been carried so far in media accounts as to discourage patients from being treated at all. This strategy will surely backfire if diseasemodifying treatments are developed that slow the course of disease without altering symptoms, because the public has come to expect that disease-modifying treatments will be more obvious and more dramatic than symptomatic ones. Pharmaceutical companies, particularly their marketing divisions, have seized on disease modification as a holy grail that will ensure market dominance for the treatment that attains it. As such, they encourage their colleagues in research and development to discover mechanisms that are plausibly disease-modifying and to design clinical trials that demonstrate this promise. These priorities have led to longer, costlier clinical trials with no guarantees that regulators will grant a claim for disease modification in AD. More likely, the indication language will stay as close to a conservative interpretation of trial data as possible. Although one can imagine a claim for alters some of the underlying processes thought to account for worsening in AD patients, one cannot imagine that a treatment would win provides disease modification that alters the natural history of AD. Finally, researchers, working in both academics and industry, often reify the distinction between symptomatic and disease-modifying as a rhetorical gesture to support the need for new research. Conventionally, grant applications (or discussions with upper management) begin with a description of the devastating nature of AD (to the individual, to society) and often transition to the request for funding by critiquing currently available therapies to demonstrate the magnitude of unmet need. Statements that have been made about the marketed antidementia drugs in these contexts have emphasized their merely symptomatic benefits or emphasized that benefits are small or modest or shortterm, leading us to assume that the next generation of therapies will provide large, easily observable benefits of long duration. 5. Conclusion There are several reasons to abandon or at least to de-emphasize the distinction between symptomatic and disease-modifying treatments in AD. First, we do not really know all of the mechanisms for most drugs. Second, our expectations shape clinical trial designs and our interpretations of these studies when, in fact, we cannot tell whether a drug is symptomatic, disease-modifying, or both from our current methodologies. Third, we might see both symptomatic and disease-modifying effects in the same trials, either at the level of the study (manifested by group changes) or at the level of the individual when the treatment under study has both types of effects. U.S. and European regulators have engaged the business and scientific communities in an attempt to guide the field toward disease-modifying drug development. Some have put forward model-based longitudinal data analysis as a way to discern between protective (disease-modifying) and symptomatic drug benefits [20]. These approaches are severely limited by our lack of models for the natural progression of AD, especially in treated patients. Others have emphasized a two-stage approach in which a delay in natural progression should be demonstrated on the basis of clinical signs and symptoms (for a claim of delay of disability) followed by a convincing package of biologic and/or neuroimaging data for a full claim of disease modification [21]. This approach is currently severely limited by our lack of understanding of biologic markers and their relationship to disease state over time. Regulatory discussions of AD treatments have raised the concepts of enduring benefit and of added benefit over currently available treatments. These concepts are easily understood, clearly desirable for patients and clinicians, and blind to the symptomatic versus disease-modifying distinction. For example, a reversible, purely symptomatic treatment that maintains its benefits over the lifespan of the patient with continued use would be valuable, regardless of mechanism of action. An agent that substantially improved the treated group over its baseline, despite the fact that both treated and untreated patients were already maximally treated with marketed therapies, would be valuable, regardless of its impact on the underlying neuropathology. We should continue to strive for improved therapies to prevent and treat AD. Current approaches to the development of new AD treatments have led to expensive studies (because of inclusion of large numbers of subjects to power for a small effect and inclusion of unproven neuroimaging and biologic markers), long studies (to allow for placebo decline, especially in studies weighted toward milder patients), and to analysis plans that are being developed in the course of the efficacy trials themselves. We should return to the approach whereby treatments must first prove that they meet the standards of an antidementia drug, preferably with the added benefit of showing these effects on the back-

5 R.S. Doody / Alzheimer s & Dementia 4 (2008) S21 S25 S25 ground of already approved therapies. The next stage in development should be an attempt to show that the benefits are both substantial (perhaps through inclusion of biologic markers and/or novel individualized efficacy measures) and enduring throughout the longest reasonable duration that can be studied. References [1] Summers WK, Majovski LV, Marsh GM, Tachiki K, Kling A. Oral tetrahydroaminoacridine in long-term treatment of senile dementia, Alzheimer type. N Engl J Med 1986;315: [2] Rogers SL, Doody RS, Mohs RC, Friedhoff LT, the Donepexil Study Group. Donepezil improves cognition and global funtion in Alzheimer disease. Arch Intern Med 1998;158: [3] Rogers SL, Farlow MR, Doody RS, Mohs R, Friedhoff LT, the Donepezil Study Group. A 24-week, double-blind, placebocontrolled trial of donepezil in patients with Alzheimer s disease. Neurology 1998;50: [4] Leber P. Observations and suggestions on antidementia drug development. Alzheimer Dis Assoc Disord 1996;10(Suppl1):31 5. [5] Farlow M, Anand R, Messina J Jr, Hartman R, Veach J. A 52-week study of the efficacy of rivastigmine in patients with mild to moderately severe Alzheimer s disease. Eur Neurol 2000;44: [6] Doody RS, Geldmacher DS, Gordon B, Perdomo CA, Pratt RD, for the Donepezil Study Group. Open-label, multicenter, phase 3 extension study of the safety and efficacy of donepezil in patients with Alzheimer disease. Arch Neurol 2001;58: [7] Raskind MA, Peskind ER, Truyen L, Kershaw P, Damaraju CRV. The cognitive benefits of galantamine are sustained for at least 36 months. Arch Neurol 2004;61: [8] Beach TG, Walker DG, Roher AE, Potter PE. Anti-amyloidogenic activity of cholinergic agents. Drug Dev Res 2002;56: [9] Sabbagh MN, Farlow MR, Relkin N, Beach TG. Do cholinergic therapies have disease-modifying effects in Alzheimer s disease? Alzheimer s & Dementia 2006;2: [10] Gunnarsson MD, Kilander L, Basun H, Lannfelt L. Reduction of phosphorylated tau during memantine treatment of Alzheimer s disease. Dement Geriatr Cogn Disord 2007;24: [11] Leber P. Slowing the progression of Alzheimer disease: methodologic issues. Alzheimer Dis Assoc Disord 1997;11:S [12] Whitehouse PJ, Kittner B, Roessner M, Rossor M, Sano M, Thal L, et al. Clinical trial designs for demonstrating disease-course-altering effects in dementia. Alzheimer Dis Assoc Disord 1998;12: [13] Fox NC, Cousens S, Scahill R, Harvey RJ, Rosser MN. Using serialregistered brain magnetic resonance imaging to measure disease progression in Alzheimer disease: power calculations and estimates of sample size to detect treatment effects. Arch Neurol 2000;57: [14] Doody R, Pavlik V, Massman P, Kenan M, Yeh S, Powell S, et al. Changing patient characteristics and survival experience in an Alzheimer s center patient cohort. Dement Geriatr Cogn Disord 2005; 20: [15] Doody RS, Massman P, Dunn JK. A method for estimating progression rates in Alzheimer disease. Arch Neurol 2001;58: [16] Doody RS, Dunn JK, Clark CM, Farlow M, Foster NL, Liao T, et al. Chronic donepezil treatment is associated with slowed cognitive decline in Alzheimer s disease. Dement Geriatr Cogn Disord 2001; 12: [17] Lopez OL, Becker JT, Wisniewski S, Saxton J, Kaufer DI, DeKosky ST. Cholinesterase inhibitor treatment alters the natural history of Alzheimer s disease. J Neurol Neurosurg Psychiatry 2002;72: [18] Rountree S, Chan W, Doody R, Pavlik V, Darby E, Saddiqui S, et al. Exposure to antidementia drugs slows clinical progression of Alzheimer s disease (AD). Ann Neurol 2007;62(Suppl 11):S55. [19] Cummings J, Doody R, Clark C. Disease-modifying therapies for Alzheimer disease. challenges to early intervention. Neurology 2007; 69: [20] Jadhav PR, Gobburu JVS. Model-based longitudinal data analysis can lead to more efficient drug development. a case study. FDA Science Forum. EPub April 27, [21] European Medicines Agency. Committee for medicinal products for human use (CHMP) draft document doc. ref. CPMP/EWP/553/95 Rev.1. July 19, 2007.