The meetings convened by the National Institutes of Health (NIH) in 2017 to gather input about how to respond to the opioid epidemic (see Chapter 1) concluded that public–private partnerships (PPPs) were needed in two different domains, said Nora Volkow. One domain would address the need to develop more effective non-opioid pain medications or nonaddictive opioid medications, and the other is to better develop treatments for opioid use disorders and for reversing overdose, she said. She envisioned that the pain partnerships would be a more classic PPP with investment by academia, government, and industry in precompetitive space through the development of resources to accelerate drug development, while the opioid use disorder partnership would be targeted more toward interactions in competitive space by engaging industry in the development of medications. There is strong support at the highest levels of government to prioritize the treatment of addiction and pain, she said, noting that the Food and Drug Administration (FDA) will play a key role and is highly motivated to help facilitate easier regulatory pathways for novel products.
Volkow said that in discussions with pharmaceutical and biotech companies working in the pain therapeutics space, a major roadblock emerged regarding the sharing of data that pertains to efforts on successful and failed drug developments for pain. A PPP could address this problem by establishing a shared database, although Volkow acknowledged that legal issues will need to be addressed in order for this to become a reality. Other points endorsed at the stakeholder meetings convened by NIH included the creation of a research trial network, the development of biomarkers to demonstrate target engagement and stratify patients for trials, the development of objective measures of pain sensitivity, reengineering preclinical platforms to have better predictive efficacy, and applying new technologies to improve pain drug discovery, said Volkow. Robert Gereau pointed to the example of the Brain Research through Advancing Innovative Neurotechnologies (BRAIN) Initiative, which was launched in 2003 and has already produced an abundance of new technologies that enable the identification of cells and circuits that change in association with the development of chronic pain. The BRAIN Initiative has achieved this not only
through the infusion of funds, but also by bringing diverse groups of investigators together to work on big problems, which Gereau said is the type of change needed in the pain field.
Some partnerships have already been established to advance the strategic goals of the National Pain Strategy, said Sean Mackey. For example, his group is involved in the multicenter consortium for pelvic pain called the Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP). Through 11 sites, the MAPP consortium enabled collection and analysis of imaging data that led to the discovery of an imaging-based pain signature for pelvic pain (see Chapter 4). Mackey also has collected large amounts of data through the Collaborative Health Outcomes Information Registry (CHOIR), which was established in response to the Institute of Medicine’s report Relieving Pain in America: A Blueprint for Transforming Prevention, Care, Education, and Research (IOM, 2011). CHOIR uses an open-source learning health care platform to develop a deep signature of individual patients across multiple dimensions of physical, psychosocial, and social functioning. Mackey said it has been “incredibly informative” as to what drives pain and leads to persistence of pain. For example, social isolation and social satisfaction appear to be key drivers for patients with pain, as well as those with addiction, he said.
These partnerships are important, said Mackey, because they enable the collection of large aggregated amounts of data, which need to be integrated into clinical trials in order to understand the characteristics of heterogeneous populations. However, he noted that for every patient brought into a clinical study, nine are typically excluded, often leading to homogeneous populations of “simpler” patients who may not represent the broader population.
William Maixner suggested that the primary goal of a pain consortium might be to collect and integrate data from large human cohort studies, develop subpopulations, and do molecular profiling to begin validating targets and developing new chemical entities or repurposing existing compounds. A second goal, he said, would be to develop new bioinformatics tools for target discovery. Third, Maixner cited the need to conduct mechanistic studies and cellular studies for both reverse and forward translation.
Walter Koroshetz said that a problem companies have made in the past was trying to solve big problems with a single drug, only to shut down the whole program when the drug failed to reach its primary endpoint. While “multiple shots on goal” are clearly needed, he advocated what he called “developing a beachhead” by focusing first on the target most likely to yield success, such as calcitonin gene-related peptide (CGRP) in migraine,
before generalizing to a vast array of pain conditions, that is, moving step by step rather than trying to address everything at the same time.
Partnership for a Preclinical Testing Platform to Catalyze Discovery of Novel Drugs
John Kehne of the National Institute of Neurological Disorders and Stroke (NINDS) proposed a PPP to develop a preclinical testing platform that would catalyze the discovery and characterization of non-addictive modalities—small molecules, biologics, and devices—for the treatment of pain. Such a partnership would enable the design of a rigorous, centralized, and flexible screening engine to generate robust data, provide a means of evaluating agnostically early-stage targets and leads, and promote translation of discoveries into new drugs, said Kehne. Chris Flores of Johnson & Johnson suggested that this platform also promote standardization and the use of informatics approaches, and consider many themes mentioned throughout the workshop, including the importance of spontaneous versus evoked pain behaviors. He also advocated focusing on validated targets, including cannabinoids, the opioid receptor, and various ion channels. John Dunlop added that the platform should be accessible to academics and both small and large companies, so that all are getting the same high-quality data. Incorporation of genome-wide association studies (GWASs), more systematic use of human cell–based models, and development of large panels of cell lines could also lead to the identification of new targets and improve translation, he said. Dunlop also supported extending the preclinical testing platform to include companion animals.
As partners in this collaboration, NIH would provide developers free-of-charge access to resources for testing and characterizing promising leads, including medicinal chemistry and drug metabolism and pharmacokinetics, said Kehne. He noted that the similar NINDS-funded Anticonvulsant Screening Program (ASP), currently known as the Epilepsy Therapy Screening Program or ETSP, has already proven successful. Since ASP’s creation in 1975, the program contributed substantially to advancing nine new epilepsy therapeutics to the market, said Kehne. He maintained that a similar preclinical testing program for non-addictive pain medications would incentivize researchers to pursue discovery and development of novel therapies by providing valuable tools, resources, and expertise while reducing the cost burden associated with drug development. Tony Yaksh noted that this could be especially valuable for smaller pharmaceutical companies that may not have, or may have cut
back, preclinical pain modeling resources. Existing NIH infrastructure for sample handling, testing, and data management would be made available to the research community, and the proposed partnership would add value beyond the ETSP model by accessing collective capabilities, scale, and resources from other partners, said Kehne. Equally important, the program would generate high-quality data for various models to support new business partnerships and applications for additional funding, he said. Finally, he suggested that harnessing the collective expertise and resources of partnering organizations would enable development of testing funnels, protocols, and reporting structures.
To create this partnership, Kehne said it will be important to find the right balance of robust assays, configure these assays into streamlined flowcharts with identified milestones, and build capacity for model development and refinement. A flexible decision-making process, qualified staff, and a commitment from partners to share all data generated by the partnership are other critical elements, said Kehne. He proposed a 5-year plan to launch the partnership, test targets and compounds, develop models, and evaluate progress. Maixner added that the partnership should include a prioritization process whereby an efficacy risk assessment is conducted for all the medications of interest to determine what will be advanced relative to what is available now.
Robert Dworkin, director of the Analgesic, Anesthetic, and Addiction Clinical Trial Translations, Innovations, Opportunities, and Networks (ACTTION),1 noted that preclinical studies may fail to predict efficacy in Phase II studies for reasons other than a failure of the model. For example, the Phase II study may have been underpowered. In such cases, a company may simply drop the program without exploring the reasons for the Phase II failure, he said. An additional goal of the preclinical testing platform could be to explore why the preclinical model was not predictive of the Phase II results and apply the learnings from those studies to improve future preclinical testing programs, said Koroshetz.
Accelerating Translation with a Network for Clinical Pain Research
One important goal of a PPP would be to develop a clinical network for pain that goes beyond enabling more efficient clinical trials, but also enables discovery research, said Volkow. Key components of discovery
research that were addressed at the workshop include identification of biomarkers, including biomarkers that predict toxicity, said Volkow. Koroshetz suggested that a clinical research network could accelerate the process of moving from discovery to therapy by providing the tools industry needs to make investments with higher predictability that they will pay off, for example, biomarkers and assays for target validation.
Putting together a clinical research network to conduct biomarker studies, deep phenotyping, and clinical trials will require streamlined operations and platforms that improve quality and efficiency, said Petra Kaufmann, director of Office of Rare Diseases Research at the National Center for Advancing Treatment Sciences. A single Institutional Review Board review, streamlined contracting, and improved data management systems, for example, can shorten the time to launch and complete a clinical study, she said. Shared or master protocols implemented with both traditional and novel paradigms for defining populations, such as those discussed in Chapter 4, and novel trial designs such as adaptive designs should also maximize the efficiency of studies and promote a deeper understanding of mechanisms, she said. Observational studies and biomarker validation studies that are conducted in parallel with clinical trials would also maximize the impact of these studies, said Kaufmann. Finally, she suggested integrating electronic medical records available at academic and health care centers into clinical data research warehouses to promote learning and help generate hypotheses.
Dworkin said the value of an analgesic clinical trials network is unequivocal, pointing to the success of similar networks established in other disease areas, such as the Huntington’s and Parkinson’s study group, oncology cooperative groups, and NeuroNEXT,2 a clinical trials innovation network established by NINDS. To develop an analgesic clinical trial network, about 30 clinical trial sites could be selected, with both a junior and senior investigator at each site, and including pediatric sites. These investigators would be brought together for a training program led by biostatisticians and clinical trialists, prior to road-testing a clinical trial in the network by testing an analgesic medication with established efficacy versus placebo, said Dworkin.
One of the difficulties in establishing a clinical trial network for chronic pain is the fact that pain conditions are frequently managed by either primary care physicians or specialists from different areas (e.g., gy-
necology, oncology, rheumatology, and urology) rather than by pain specialists in neurology or anesthesiology, said Christin Veasley. Kaufmann agreed that the network would have to be multidisciplinary and suggested that this may be possible by actively involving patients who could drive this. Scott Powers added that demonstrating success in one area can attract other specialties to a network, similar to what was seen in building the cystic fibrosis network. Identifying champions for the network concept in various disciplines may also breed success, said Powers. Koroshetz commented that recognition of an opioid crisis might also incentivize the establishment of new structures such as a network for pain research that would lead to improved pain care and reduced reliance on opioids.
Jonathon Jarow said the FDA is supportive of platform trials, master protocols, and networks. Real-world data can be helpful in identifying clinical trial sites that have been successful in enrolling patients for various types of studies. Jarow suggested that because industry may be reluctant to submit their assets to platform studies, a network for pain research might consider a demonstration project with vitamin C or some other off-the-shelf compound. He added that the 21st Century Cures Act3 created a space called innovative trials and will begin funding pilot projects in early 2018, with statistical support.
Koroshetz suggested that another strategy might be to challenge existing heavily resourced programs with diverse expertise such as NIH’s Clinical and Translational Science Awards (CTSA) to respond to the opioid crisis. Story Landis also mentioned the “hub and spoke” strategy that NINDS used for its stroke network. For pain, the hubs could be specialty clinics, with spokes going out to primary care. As with stroke, this could help build bridges between acute and rehabilitative care for patients with pain, she said.
Several workshop participants mentioned the importance of data sharing, noting that capitalizing on the increasing willingness of companies to share data, compounds, and expertise may enhance opportunities to build precompetitive partnerships. Volkow added that data sharing efforts on both successful and failed drug development efforts would help researchers learn from both their successes and failures.
Volkow said the workshop made clear the value of a clinical network for pain. Now the key challenge is to translate this recognition into a program that is sustainable and that has the resources and expertise to back it up, she said. A second key challenge is to ensure that outcomes reflect
3 For more information, go to https://www.congress.gov/bill/114th-congress/house-bill/34 (accessed on January 2, 2018).
what is most important to patients and at the same time are quantifiable, reproducible, and standardized to enable creation of large datasets that will provide information on long-term outcomes.
Pediatric Considerations for a Clinical Trial Network
The inclusion of pediatric sites in a clinical trial network is essential, said Powers, because clinicians are taking care of patients every day with imprecise knowledge about pain management. Children are not little adults, but they are willing and able to participate in clinical trials, said Powers. Moreover, pediatric specialists across disciplines have demonstrated the ability to collaborate and execute clinical trials, for example, through the Cystic Fibrosis Therapeutic Development Network.4 Powers cited an additional advantage of networks beyond the efficiencies mentioned by Kaufman. Networks create friendships that help investigators navigate the ups and downs of clinical studies as a team. They promote a sense of purpose and common goals.
Powers also recognized the vulnerability of children and adolescents, and suggested that some of the regulatory policies concerning safety and protection may need to be reconsidered for studies involving pediatric populations. However, he also suggested that waiting until adult studies are completed before enrolling children has slowed the development of effective treatments for children. Children and families interested in participating in studies have to wait far too long, and that needs to change, he said. Sharon Hertz said there is no regulatory barrier for starting pediatric clinical trials concurrently with adult studies; however, industry may be reluctant to take on pediatric studies because of perceived risks of studying a drug in children that is not known to have a clear safety profile.
Collaboration is needed in many areas of drug development because the expertise in industry is not matched in academia, said Sharon Walsh. However, she said that in the area of opioid use disorders, most expertise
4 For more information, go to https://www.cff.org/Research/Researcher-Resources/Therapeutics-Development-Network (accessed January 2, 2018).
resides in academia, where many researchers are eager to work with companies to develop new therapeutics. Academic labs, for example, have the expertise to conduct basic research on abuse liability, she said.
Walsh noted that there are many joint funding opportunities through the National Institute on Drug Abuse (NIDA) that are specifically designed to bring people together across academia and industry. Another organization that promotes collaboration in this area is the College on Problems of Drug Dependence, the longest standing scientific organization in the United States, which has been dedicated to strengthening science on drug use disorders to improve therapeutic development, said Walsh.
One of the lowest hanging fruits for safer opioids are abuse-deterrent formulations, said Volkow, yet these approaches are utilized in less than 3 percent of all prescriptions, presumably because they are much more expensive. She added that if payers are not incentivized to cover these medications, pharmaceutical companies may be reluctant to invest in developing them. This is an area where patient advocates have played and can continue to play a critical role in trying to change the conversation about reimbursement and parity across Medicaid, Medicare, and private insurers, she said.
Alternate formulations and development of novel approaches to addressing opioid addiction are in development as described in Chapter 5, but the urgency of the problem demands more attention and more resources, noted several workshop participants. Volkow noted that repurposing existing medications and combining these medications to improve outcomes may also prove beneficial, but have not been prioritized by industry. Christian Heidbreder suggested that PPPs might also focus attention on studying the effectiveness of extended-release buprenorphine and/or other treatments in patient subpopulation programs; exploring the pharmacogenetics of addiction; and developing programs to deliver extended-release formulations of buprenorphine in emergency department settings as a way of preventing opioid overdose and subsequent relapse to drug-seeking and drug-taking behaviors.
There has also been interest in developing biomarkers for vulnerability of addiction, which could increase understanding of why some people but not others become addicted, and identify new therapeutic targets that are less likely to trigger addiction, said Volkow, adding that NIDA has been working for many years in this area. Heidbreder added that biomarkers are also needed for patient stratification and endpoints in clinical trials. New patient-outcome measures that address the impact of treatment on quality
of life also need to be developed and validated, along with tools to understand the mechanisms of and assess craving, he said. Heidbreder also argued that a PPP should scrutinize research using preclinical addiction models with a focus on reliability and reproducibility of published data. Finally, to make meaningful changes in the opioid use disorder landscape, he suggested a need for an interdisciplinary approach that involves the engagement of several disciplines, including engineering, chemistry, physics, and mathematics.
William Potter, senior advisor at the National Institute of Mental Health (NIMH), wondered if multiple partnerships are needed to address different paths individuals take on the road to addiction, such as physical or psychological pain. Indeed, Volkow noted that only a small minority of patients who are prescribed opioid analgesics for pain management ever develop an opioid use disorder (estimated around 10 percent) and about 20 percent of individuals with opioid use disorder started out with pain. The overprescription of opioid medications for pain, however, led to their diversion and facilitated the illicit use of opioid medications for their rewarding effects, she stated. Andrey Ostrovsky, then chief medical officer for the Center for Medicaid and CHIP (Children’s Health Insurance Program) Services, added that components of the heroin epidemic differ markedly from city to city. For example, in one city the predominant heroin product may be black tar from Mexico, while in another city it might be carfentanil sent through the mail from China. Joblessness, despair, stigma, and other psychological and sociological components that contribute to addiction also vary across cities, said Ostrovsky. David Shurtleff added that self-medication for disorders such as depression, anxiety, posttraumatic stress disorder, and trauma also contribute to the huge increase in the incidence of opioid use disorders. The complicated nature of the problem argues strongly for a PPP, with one pillar focusing on developing medications and strategies to treat opioid use disorder, and another on new medications for pain, said Shurtleff.
Other crosscutting partnerships proposed at the workshop included a precompetitive consortium on target identification and validation, which was advocated by Flores. Potter also endorsed this idea, noting that when multiple companies are pursuing the same target, it would be useful to better understand what differentiates their compounds by using the same assays. Also proposed by Flores and Maixner was a PPP between pharmaceutical and biotech industries and academia, to put forward shelved compounds that could be tested in preclinical pain and addiction models for potential efficacy and repurposing.
UNDERSTANDING THE ROLE OF PAYERS IN SUPPORTING THERAPEUTIC DEVELOPMENT FOR PAIN AND OPIOID USE DISORDERS
Shari Ling, deputy chief medical officer for the Centers for Medicare & Medicaid Services (CMS), described how the output of clinical studies, including potential studies emerging from the PPPs previously described, informs coverage decisions made by CMS. Quality measurements are key, she said, and evidence must go beyond supporting clinical and analytical validity, which is required for FDA approval, to clinical utility. Ling said she was heartened to hear workshop participants focusing on co-morbidities and what matters most to patients and families. CMS requires patients to be at the center of studies, with determination of risks and benefits central to coverage decisions for treatments as well as diagnostic and prognostic tests, she said.
Because of the complexity of the Medicare population, co-morbidities are a particular challenge in the context of how Medicare views data and evidence, said Ling, noting that arthritis, cancer, and several other conditions associated with high pain levels are common in this population. One of the major reasons that evidence may fall short in Medicare determinations is that it was collected in non-Medicare populations with the expectation that findings would be generalized to the Medicare populations, she said.
Ling advised that it is never too soon for investigators to begin thinking about how data will be interpreted clinically and translated into information useful to beneficiaries. Phenotypes that are well described in a clinical setting and biomarkers of clinical pathways need to inform the utility and value of treatments and diagnostic tests, she added. As health care moves away from fee-for-service to alternative payment models in which there may be shared management across primary and specialty care, the focus must remain on outcomes that are meaningful for patients and families, said Ling. The CMS Innovation Center5 has the authority to test new care and payment models, and recently issued a request for information regarding possible demonstration projects in that area, including projects to lower the cost of prescription drugs and provide mental and behavioral health services for beneficiaries, including those affected by the opioid crisis.
Ostrovsky focused his remarks on how to get new innovations paid for through Medicaid, which is the largest insurer in the United States, covering about 75 million individuals per year at an annual cost of about $500 billion. Because Medicaid is a federal–state partnership, both the federal government and states participate in coverage decisions. For example, states submit plans for how they want to pay for pharmacological and non-pharmacological approaches to treating opioid use disorders, chronic pain, and acute pain. Most state plans are accepted by CMS. However, with state budgets constrained by declining tax revenues, even new therapeutics with strong supporting data may face state restrictions on coverage, said Ostrovsky. States are also exploring alternative payment programs such as bundled payments and Medicaid Health Homes, rather than fee-for-service for substance use disorders. Ostrovsky suggested that new PPPs consider approaching states early to learn about their concerns regarding coverage of novel therapeutics and diagnostics.
Koroshetz reflected on the workshop presentations, noting that the stories shared by Veasley and Jessica Hulsey Nickel set the stage for framing discussions around what is important for patients. The pain treatment and opioid crises have now surpassed the AIDS crisis in terms of the number of people who die each year, he said, yet the resources dedicated to finding effective, non-addictive treatments for pain and addiction are low in comparison to what has been allocated for AIDS. While the field is expanding rapidly, he noted that the time frames for moving a compound from discovery to an approved drug are unacceptably long and must be accelerated enormously. To accomplish this, he suggested emphasizing the development of tools that industry needs to make their investments with high predictability that they are going to pay off, including platforms such as clinical research networks that enable testing of multiple drugs in parallel rather than sequentially.
Shurtleff added that a long-term strategy is needed that incorporates new technologies and tools to move drugs through the development pipeline more quickly. Moreover, because many factors contribute to the pain syndrome, therapies will need to incorporate both medications and behavioral approaches, he said. Patients must be at the center of the treatment plan, Shurtleff continued, which involves both letting them take control of their own treatment and being part of the therapy-development process.
This page intentionally left blank.