Neuroscience Trials of the Future: Proceedings of a Workshop (2016)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

5 The Regulatory Landscape: International Opportunities and Challenges Highlights • Regulators desire good proof-of-concept studies, leading to rep- resentative clinical trials that can then, in turn, result in appropri- ate clinical practice guidelines and performance measures (Califf). • Regulators are also increasingly focused on sustainability, which may be achievable by simplifying trials and enrolling repre- sentative populations (Califf), and through the use of real-world data or drug-product monitoring registries (Pani). Another way to help sustainability is early health technology assessment (HTA) evaluation (Pani). • Integrating clinical research networks is one way to ensure more efficient trials (Califf). • The patient’s voice in particular has become increasingly im- portant to regulators in defining the value of a medicine, yet in- corporating patients’ viewpoints into clinical trials may further increase the complexity of those trials (Romano). • Regulators have been pushing for increased data sharing to pro- mote efficiency and ensure continued progress (Califf and Pani). • There is a need for new regulatory approaches for combining passive and experimental data that would meet the needs of the 21st century (de Vries, Kieburtz, and Reitz). • Drug-product monitoring registries enable tracking of longitudi- nal outcomes and adverse effects (Laughren and Pani). • Randomization with a registry enables real-world testing of hy- potheses (Califf). NOTE: These points were made by the individual speakers identified above; they are not intended to reflect a consensus among workshop participants. 43 

44 NEUROSCIENCE TRIALS OF THE FUTURE Many of the challenges raised in the previous chapters with regard to integrating novel tools, trial designs, and statistical approaches into neu- roscience trials have regulatory implications. Regulatory agencies around the world have developed strategies to keep pace with the evolving prod- uct development in the context of each country’s particular political, le- gal, economic, social, cultural climate. Yet despite their geopolitical differences, the goals of these countries remain essentially the same. As articulated by Carlos Peña, director of the Division of Neurological and Physical Medicine Devices in the Office of Device Evaluation at the FDA’s Center for Devices and Radiological Health (CDRH), one central goal is to optimize trial design approaches to get products to patients who desperately need them while ensuring that these products undergo appro- priate evaluation for safety and effectiveness. REGULATORY PERSPECTIVES FROM THE UNITED STATES Robert Califf, Commissioner of Food and Drugs at the FDA, said that within the FDA, there is a shared view that the best medical out- comes occur when doctors and other health care providers and patients are armed with high- quality evidence to support what they do, and this is most likely to happen when the clinical trials and observational studies are actually done in practice. The FDA is tasked with providing instruc- tions in the label on how to use the product in practice, not in theory, said Califf, so making an extrapolation from a rarified clinical trial to real practice does not make much sense. He acknowledged, however, the dif- ficulty of collecting data from real-world situations, which has led to a “parallel universe” of data collected specifically for clinical trials. In ad- dition, there is a disconnect between the instructions for use provided in the product label and actual use in clinical practice, such that, according to a recent study in Canada, more than 25 percent of drugs prescribed for CNS conditions were for “off-label” uses, and 21 percent were off label with no credible evidence of the drug’s effectiveness for that condition (Eguale et al., 2016). For some CNS drugs, including clonazepam1 and amitriptyline,2 more than 70 percent of use was off-label. Interestingly, the adverse event rate for drugs used off label was twice as high as for drugs used on label, but when considering only drugs used off label but with good evidence of effectiveness, the rate was the same as the on-label rate. 1 Clonazepam is used to treat seizures and panic disorder. 2 Amitriptyline is a tricyclic antidepressant used to treat symptoms of depression. 

THE RE EGULATORY LA ANDSCAPE 45 Caaliff offered a glimpse at what w he thinkss the system m might look likke. It begiins with good d proof-of-co oncept studiess, which leadd to representta- tive cllinical trials. The best knoowledge from m those trials is then used to craft clinical c practice guideliness and performmance measurres. When whhat happen ns in practicee deviates fro om what was expected, revvisions may be needed d in the clinical trials proocess. Underlyying all of thhese steps, saaid Califf,, is measurem ment and educcation both inn the clinical ttrials arena annd in clin nical practice,, which he deescribed as a learning heaalth care systeem (see Fiigure 5-1). Caaliff also add dressed conceerns about suustainability, citing a receent study showing that the cost of clinical c trials is rising twicce as fast as tthe rest off the Americaan economy (Berndt and Cockburn, 20014). He notted that th he driver of thhis unsustainab ble growth inn the cost of cconducting cliin- ical triials is compleexity, which also drives a reduction inn the number of peoplee who enroll in trials—a trend that iis worsening in the Unitted FIGUR RE 5-1 Learnin ng health care system. SOURC CES: Presented d by Robert Califf at the Woorkshop on Neuuroscience Trials of the Future, F March 4, 2016. From m Annals of Inteernal Medicinee, Greene, S. M M., R. J. Reid, R and E. B.B Larson. Imp plementing the learning heallth system: Froom conceppt to action, 1557, 3, 207–210 0. Copyright © [2012] Ameerican College of Physiciians. All Rights Reserved. Reprinted R withh permission of American Col- lege off Physicians, In nc. 

46 NEUROSCIENCE TRIALS OF THE FUTURE States and is leading to increased inefficiency. His FDA colleagues in the CNS space reiterated many suggestions raised throughout the workshop, to include the need to simplify and enroll relevant, not rarified, popula- tions; make trials more inclusive; stop the collection of non-serious ad- verse events for every patient; reduce the number of clinic visits required for research studies; and invest in technologies that provide the clearest answers to critical questions. The FDA continues to optimize trial design approaches to achieve its goals, according to Peña. CDRH uses a risk-based approach, requiring increased oversight for products deemed to present a greater possible risk. The drug and biologic divisions of the FDA, CDER, and the Center for Biologics Evaluation and Research (CBER) use similar approaches. The FDA has published numerous guidance documents to clarify the considerations that should be taken into account by sponsors through the approval process, beginning with presubmissions. They also strongly encourage sponsors to initiate dialog with them at the earliest stages, so that the agency can provide feedback and suggestions on the anticipated designs before a study is initiated and data collection has begun. REGULATORY PERSPECTIVES FROM ITALY Luca Pani, director general of the Italian Medicines Agency (AIFA), also addressed sustainability. Because AIFA functions as a regulatory, payer, and HTA institution, it must grapple with all its roles to achieve better outcomes and controlling costs, he said. Sustainability and real- world effectiveness are thus central tenets of the Italian regulatory sys- tem. To meet these challenges, AIFA has invested more than €22 million (about $25 million) in information technology and the development of drug-product monitoring registries in the past 5 years. Given the im- portance of the investments made in this space, the National Health Ser- vice Information Technology Law was passed in Italy in 2012. It mandated the implementation of Web-based registries after marketing authorization to measure drug safety and effectiveness for approved ther- apeutic indications and some selected off-label uses (Montilla et al., 2015). As of February 2016, data from 850,000 patients have been cap- tured by the registries. A second key aspect of the AIFA system that addresses sustainability is pricing and reimbursement, said Pani. He described a range of possible reimbursement outcomes that await Market Authorization Holder who 

THE REGULATORY LANDSCAPE 47 seek registration by AIFA, ranging from a refusal to reimburse, reim- bursement without particular conditions, reimbursement with a control on prescription (adherence to an optimized therapeutic plan) to a Man- aged Entry Agreement (MEA). MEAs are a heterogeneous group of in- struments that are being increasingly implemented to guarantee sustainability of innovative and expensive medicines. MEAs can be pure- ly financial based (price/volume agreements) or health-outcome based (Ferrario and Kanavos, 2015). Most frequently a combination of the two may be applied. For example, AIFA has signed contracts with pharma- ceutical companies that set payment based on treatment effectiveness (performance-based risk sharing agreements), with companies refunding costs if the medication fails, said Pani. Pani described how the AIFA strategy was applied to the approval of new treatments for hepatitis C (HCV), which are highly effective, but extremely expensive. Like the United States, Italy has a high incidence of HCV, and the cost of treating them all would be prohibitive. Thus, AIFA created a permanent national working group to develop a strategy for providing HCV drugs. After developing seven prioritization criteria that would provide the drug to patients with the greatest clinical need, AIFA used data from their registries to calculate the total number of treatments needed, and thus to negotiate a price/volume discount with the manufacturer. AIFA also used earlier versions of the registries to determine how best to use new diabetes therapies (incretins) most effectively. In 2008, they approved the reimbursement of three drugs—exenatide, sitagliptin, and vildagliptin—in which patients were subject to enrollment in the real- world data. Data from the post-marketing registry revealed substantial off-label use, little adherence, as well as inconsistent effectiveness. How- ever, the data also showed that when used appropriately in combination with exercise, the effectiveness of the drugs was consistent with the re- sults seen in the registration trials (Montilla et al., 2014). In the area of psychiatry, AIFA used the national drug utilization da- tabase (Osmed Health-DB) to study treatment-resistant depression. They developed an antidepressant usage index, which revealed that the higher the degree of “resistance” along a continuum, the higher the cost of both depression-related and depression-unrelated resources for the National Health System, said Pani. 

48 NEUROSCIENCE TRIALS OF THE FUTURE POTENTIAL REGULATORY IMPLICATIONS FOR CLINICAL RESEARCH INNOVATIONS Recognizing the inefficiencies of the historical model of clinical re- search, where a single coordinating center manages a trial with top-down decision making with independently operated sites, Califf proposed a different model of interoperable networks that share sites and data. Inte- grating clinical research networks not only ensures more efficient trials, but also enables patients, physicians, and scientists to form true “com- munities of research.” Indeed, such systems are already being built. The FDA’s Sentinel In- itiative,3 launched in 2008, is a national electronic system that will enable postmarket safety monitoring of FDA-approved drugs by providing ac- cess to claims data from more than 100 million people. Linked to Senti- nel is the National Institutes of Health (NIH) Health Care Systems Research Collaboratory,4 which has initiated 10 demonstration projects spanning 12 NIH institutes and centers to reduce the cost of clinical trials by capturing electronic health record data. Another network, the National Patient-Centered Clinical Research Network (PCORnet),5 brings patients into the process as full partners. With the support of user fees from the Prescription Drug User Fee Act (PDUFA), a national evidence genera- tion system is being built to combine data from different networks through collaborations with industry, academia, and integrated health systems. The culmination of all of this, said Califf, is the Precision Medicine Initiative.6 The FDA is integrally involved in this NIH-led effort to get volunteers to participate in research as a normal part of their patient care. He noted that this initiative was fueled by a recognition by the U.S. Pres- ident and Vice President that what is holding the country back from ex- ploiting the computational power currently available is culture, not technology, which has resulted in people hoarding rather than sharing data. 3 For more information, go to http://www.fda.gov/Safety/FDAsSentine/Initiative/ ucm2007250.htm (accessed June 3, 2016). 4 For more information, go to https://www.nihcollaboratory.org/about-us /Pages/default.aspx (accessed June 3, 2016). 5 For more information, go to http://www.pcornet.org (accessed June 3, 2016). 6 For more information, go to https://www.whitehouse.gov/precision-medicine (ac- cessed June 3, 2016). 

THE REGULATORY LANDSCAPE 49 Encouraging Data Sharing A potential roadblock to these efforts is the privacy protections on data, particularly mental health data, noted Robert Bilder. Califf said the expert community and patients would need to push for solutions that en- able data sharing with appropriate protections. In Italy, said Pani, there has been a push for sharing data from control arms (which is often the standard-of-care treatment) in clinical trials; in the United States, the FDA is close to publishing the Final Rule for www.ClinicalTrials.gov, which will require sponsors to share results from clinical trials or poten- tially face fines of up to $10,000 per day, according to Califf. While sev- eral potential barriers exist to sharing clinical trial data, several participants acknowledged the many benefits. For example, sharing clin- ical trial data • “has great potential to accelerate scientific progress and ultimately improve public health by generating better evidence on the safety and effectiveness of therapies for patients; • increases patients’ contributions to generalizable knowledge about human health by potentially facilitating additional findings beyond the original, prespecified clinical trial outcomes; • could provide a more comprehensive picture of the benefits and risks of an intervention and allow health care professionals and patients to make more informed decisions about clinical care; and • could potentially improve public health and patient outcomes, reduce the incidence of adverse effects from therapies, and de- crease expenditures for medical interventions that are ineffective or less effective than alternatives” (IOM, 2015, pp. 31–32). No company has the power (or desire) to implement universal data sharing by itself; it will more likely come from legislation, said Pani. Other possible solutions include having the FDA serve as an honest bro- ker to manage shared data, or forming a consortium that hosts a clinical data repository. A model for this in Europe could be the Innovative Med- icines Initiative (IMI) called Novel Methods leading to NEWMEDS,7 an international public‒private partnership that encourages data sharing 7 For more information, go to http://www.newmeds-europe.com (accessed June 3, 2016). 

50 NEUROSCIENCE TRIALS OF THE FUTURE from companies and investigators with the goal of identifying new meth- ods for drug development for depression and schizophrenia. Califf said ECG data were shared in this way years ago, after overcoming legal hur- dles. This resulted in a much better understanding of cardiac events asso- ciated with prolonged QT interval8 issues. Califf added that part of the solution has to be a firewall and an audit trail to protect data that are con- tributed to such initiatives. A New Regulatory Framework for Combining Passive and Experiential Data In 1962, the U.S. Congress passed the Kefauver-Harris Amendments to the Federal Food, Drug, and Cosmetics Act, for the first time requiring trials to demonstrate substantial evidence of effectiveness prior to market approval. With the passage of the Orphan Drug Act in 1983, the FDA recognized the need for a different regulatory framework around rare disease entities. Since then, the identification of specific genetic muta- tions that could predict prognosis or response to the therapy, as well as the emergence of the Internet and the digital revolution, have highlighted the need for a new regulatory framework appropriate for the 21st centu- ry, according to Karl Kieburtz. Thomas Laughren cited some of the 21st century approaches that present challenges to be addressed in this new framework: • Incorporating novel designs and methods into clinical studies, and how sponsors can gain regulatory guidance on the adoption of new methodologies in a drug development program; • Using the Internet and social media for patient recruitment and assessment; • Conducting trials at geographically dispersed sites; and • Incorporating more clinically meaningful endpoints into trials. Pani added that the shifting classification boundaries for therapeutics present a challenge for regulators with respect to considerations of both efficacy and safety. A drug may first be developed for the treatment of psoriasis, for example, and then be used to treat arthritis, Crohn’s dis- 8 The QT interval is a measure of the time between the start of the Q wave and the end of the T wave in the heart’s electrical cycle. A long QT interval is a risk factor for sudden death, while a short QT interval indicates a genetic condition (http://www.tga.gov.au/ file/1140/download, accessed June 23, 2016). 

THE REGULATORY LANDSCAPE 51 ease, or even cancer, said Pani. In addition, particularly for precision medicine but for other drugs in development as well, variability in re- sponse needs to be better understood. Thus, rather than striving for ho- mogeneity in clinical trials to reduce variability, it may be more advantageous to enroll very large heterogeneous populations and then use analytics to discover commonalities and differences. However, the enrollment of large heterogeneous populations in the RCTs is the oppo- site approach of the adaptive pathway, said Pani. He added that enrolling large population from an academic point of view is acceptable, but on the other hand, is very expensive and it may delay the access of new drugs. As the range of treatment targets expands, there will be an increasing need for collaboration among regulators, other governmental agencies, and the research community, added Tiffany Farchione. Drug Product Registries Registries are an enormously useful source of information that could generate many hypotheses that might lead to randomized trials, accord- ing to Laughren. For example, a registry was required for the clozapine9 registration trials, which enabled generating data regarding the hazard curve for the potentially fatal adverse event agranulocytosis, in which the level of white blood cells called neutrophils drops to dangerous levels, causing suppression of the immune system (Alvir et al., 1993). It also suggested that clozapine reduced suicides in patients with schizophrenia. Italy also has established registries for many medications, including the treatment-resistant depression registry mentioned earlier. These regis- tries provide information about the effectiveness and safety of a treat- ment. In addition, because registries collect fairly rigorous data, they are an ideal place to embed randomization, said Califf. For example, Sweden is doing a series of registry trials where everyone gets randomized as part of routine care. As a result, they were able to conduct a trial of a throm- bectomy device, which removes a clot from a blood vessel, for a tiny fraction of what it would have cost to run such a trial in the United States. Randomization with real-world data is done every day in the business world, but has encountered roadblocks in medicine, said Pani. However, Laughren cautioned that many registries that have been set up are failing to generate useful data because they were poorly designed. Italy has addressed this by requiring that registry databases meet Interna- tional Organization for Standardization standards and by continuously 9 Clozapine is an atypical antipsychotic agent. 

52 NEUROSCIENCE TRIALS OF THE FUTURE and randomly checking the data, according to Pani. One challenge is the quality of data; the data must be well structured but the quality may de- pend by other factors, he added. Califf commented that there are excel- lent models in the United States, such as one developed by the Society of Thoracic Surgeons, where data are checked and audited. They were de- signed that way because surgeons realized they would not be reimbursed unless they produced reliable outcome data. Several health systems are investing heavily in integrated data warehouses; there are also efforts to enable these data warehouses to communicate with each other while re- taining local control.