Proceedings of a Workshop

WORKSHOP OVERVIEW¹

Public–private partnerships (PPPs) are collaborations among public and private entities (e.g., government agencies, research institutions, nonprofit organizations, and industry organizations) that have the goal of bringing together diverse expertise, resources, and perspectives to solve complex challenges. To examine opportunities to improve the care and outcomes for patients with cancer through PPPs for clinical cancer research, the National Cancer Policy Forum collaborated with the Forum on Drug Discovery, Development, and Translation to host a public workshop on October 17–18, 2023. The workshop convened speakers from government agencies, industry, academia, health systems, foundations, and patient advocacy to explore opportunities to advance cancer treatments through PPPs, discuss the challenges in establishing effective PPPs and factors that contribute to successful partnerships, and consider the ways in which PPPs can enhance clinical cancer research and advance broader health equity goals including clinical trial diversity.

___________________

¹ This workshop was organized by an independent planning committee whose role was limited to identification of topics and speakers. This Proceedings of a Workshop was prepared by the rapporteurs as a factual summary of the presentations and discussions that took place at the workshop. Statements, recommendations, and opinions expressed are those of individual presenters and participants and are not endorsed or verified by the National Academies of Sciences, Engineering, and Medicine, and they should not be construed as reflecting any group consensus.

This Proceedings of a Workshop summarizes the presentations and discussions at the workshop. Observations and suggestions from individual participants are discussed throughout the proceedings, and highlights are presented in Boxes 1 and 2. (Box 1 includes observations on PPPs for clinical cancer research, and Box 2 outlines potential strategies for advancing the development and execution of PPPs in clinical cancer research.) Appendixes A and B provide the workshop Statement of Task and agenda, respectively. Presentations and the workshop webcast have been archived online.²

___________________

² See https://www.nationalacademies.org/event/40283_10-2023_optimizing-public-private-partnerships-for-clinical-cancer-research-a-workshop (accessed December 4, 2023).

OVERVIEW OF PUBLIC–PRIVATE PARTNERSHIPS TO ADVANCE CLINICAL CANCER RESEARCH

Julie Gerberding, president and chief executive officer of the Foundation for the National Institutes of Health (FNIH), described the role of FNIH as a neutral convener of broad partnerships and offered lessons learned from experiences in three main areas: building collaborative science, supporting scientists, and earning trust in science. One of the most important dimensions in successful partnerships at FNIH is rigorous criteria for what creates a partnership opportunity, she said. FNIH focuses on tackling large-scale problems that require a diverse array of experts from multiple sectors.

Gerberding shared some examples of FNIH partnership platforms that have advanced cancer research. The Biomarkers Consortium³ has enabled the development of standards for assessing alternative endpoints, such as minimal residual disease, which could serve as biomarkers in clinical trials. In addition to funding through FNIH, the Biomarkers Consortium is an example of a membership partnership in which participants pay to participate in the shared value of the partnership. While there are different models of fundraising, she said, the scale and scope of investment are significant. She outlined five key phases for developing a successful partnership (see Figure 1), beginning with effective alignment and agreement on the objectives and roles within the partnership.

FNIH has formalized a pipeline approach, similar to what is used in the pharmaceutical industry, which starts with an exploratory idea, advances to concept development and project launch, and concludes with an “after action review”⁴ and impact assessment. The average time from idea to launch has historically been more than 2 years, but efforts are underway to shorten that, Gerberding explained. She highlighted the significance of the after action review, a stage that FNIH is dedicated to enhancing and is crucial in any partnership. Several innovations were initiated in 2023, including master agreements with industry partners that will eliminate the need to engage lawyers and seek approval every time a new phase of a project is starting. The goal is to better articulate the true value created by a partnership and encourage broader participation in the shared efforts. She also emphasized equity and the need to be inclusive in the scientific process to ensure that all will benefit from advances. None of the successes and achievements through partnerships matter if they do not get to the people most in need, she concluded.

___________________

³ For information on the FNIH Biomarkers Consortium, see https://fnih.org/our-programs/biomarkers-consortium (accessed March 29, 2024).

⁴ An after action review is a structured analysis of actions taken to complete the set objectives. See https://www.who.int/europe/activities/highlighting-the-benefits-of-an-after-action-review (accessed January 25, 2024).

Dinah Singer, deputy director at the National Cancer Institute (NCI), said that PPPs are increasingly being used to support medical research, and NCI engages in a wide range of PPPs through collaborations with both commercial and noncommercial entities. Outlining the five guiding principles that NCI adheres to when entering into any form of partnership, Singer said that partnerships should:

• Be undertaken for the public good.
• Address important scientific questions of high priority to the NCI mission with clearly articulated goals.
• Be scientifically meritorious and address unmet needs.
• Leverage both NCI and partner resources to accomplish the scientific goals that would not be possible by either party alone.
• Promote the generation of knowledge for public use and be fair, transparent, and inclusive.

Richard Pazdur, director of the Food and Drug Administration (FDA) Oncology Center of Excellence, explained that oncology drug development has evolved, with approximately 40 percent of all pharmaceutical development now focused on oncology.⁵ Additionally, there are often multiple companies working on similar products. He emphasized that the pharmaceutical industry is not monolithic, and small biotech companies are quite different from big pharmaceutical companies. These differences need to be carefully considered in developing PPPs to leverage the different contributions of each partner, he said.

Pazdur described four major issues that need to be negotiated to establish PPPs for drug development: risk reduction in developing a compound, control in drug development, speed, and data quality. To be successful, he said that

___________________

⁵ See https://www.mckinsey.com/industries/life-sciences/our-insights/pursuing-breakthroughs-in-cancer-drug-development (accessed May 28, 2024).

give-and-take is necessary on both sides, but understanding what motivates each collaborating partner can foster success.

Developing Successful Public–Private Partnerships

Several speakers shared their experiences with PPPs and the factors that contributed to their success. Stacey Adam, associate vice president of FNIH, underscored the crucial role of building strong trust and communication among entities. She also emphasized the importance of level-setting expectations from the start and identifying champions among partner organizations. Janet Dancey, director of the Canadian Cancer Trials Group, added that large trials necessitate international partnerships that require engagement and planning from the very beginning, including who should be at the table, who the key decision makers are, and how governance will be optimized across jurisdictions and partners.

Charles Blanke, chair of the SWOG Cancer Research Network⁶ and professor at Oregon Health & Science University’s Knight Cancer Institute, said that foundations tend to offer more agile funding support and can also support elements such as education, new types of clinical trials, additional use of data (e.g., secondary analysis), diversity and inclusion, and novel patient advocacy. Foundation-funded trials are often also better able to engage private and industry partners in PPPs, he said.

Meg Mooney, associate director of the Cancer Therapy Evaluation Program at NCI, discussed cooperative research and development agreements (CRADAs)⁷ and their evolution. She highlighted the need to continually respond to what is happening in science and noted that data sharing and intellectual property (IP) have become more important in PPPs. She said that flexibility, a key characteristic in long-standing PPPs, was essential for modifying agreements for the Pragmatica-Lung trial, which was informed by lessons learned from Lung Cancer Master Protocol (Lung-MAP) (see Box 3). Jeff Allen, president and chief executive officer of Friends of Cancer Research, also highlighted the potential for smaller PPPs that may be easier to get off the ground. Partnerships are very diverse, he said, and not every one needs its own data use agreement or legal contract.

Singer emphasized trust as the most important aspect of any relationship, as well as a clear understanding of the interests of all partners and where those interests diverge. Ellen Sigal, chairperson and founder of Friends of Cancer Research, added that aligning incentives and the mission of a partnership, as well as navigating cultural differences across partners are critical. William

___________________

⁶ See https://www.swog.org (accessed January 26, 2024).

⁷ See https://www.techtransfer.nih.gov/partnerships/cradas (accessed January 29, 2024).

Dahut, chief scientific officer at the American Cancer Society (ACS) added that the core aspect of a partnership is the perceived value by all members. Developing focused and unified goals is important, he continued. For example, diversity in clinical trials cannot easily be achieved by one group, because the issues are systemic.

Examples of Public–Private Partnerships in Clinical Cancer Research

Singer shared some examples of NCI’s PPPs to demonstrate the breadth across the research continuum. On the noncommercial side, NCI has partnered with Cancer Research UK to establish the Cancer Grand Challenges initiative, which identifies difficult problems and challenges the scientific community to assemble large international teams of researchers to address them. It is mainly focused on basic research but includes components of translational and clinical research as well. As a commercial example, she said Phase 2b of the Small Business Innovation Research⁸ program is specially designed to accelerate the commercialization of innovative technologies and help small businesses cross through the “valley of death” in product development. Through this program, small companies can commercialize their technologies with NCI grant support on the condition that they raise matching funds themselves, Singer explained. She said that CRADAs are used for many extramural clinical trials conducted through the NCI’s National Clinical Trials Network (NCTN) and several collaborations focused on basic and translational cancer research. Lastly, she described the Liquid Biopsy Consortium,⁹ which aims to develop, validate, and qualify biomarkers and seek regulatory approval for clinical use.

___________________

⁸ See https://www.sbir.gov/about (accessed January 29, 2024).

⁹ See https://prevention.cancer.gov/major-programs/liquid-biopsy-consortium (accessed January 29, 2024).

It includes members from across federal agencies and industry but is managed by FNIH as a neutral broker.

James Doroshow, director of the Division of Cancer Treatment and Diagnosis at NCI, gave an overview of the NCI Experimental Therapeutics Pipeline, also known as the NExT Pipeline,¹⁰ a PPP for drug discovery. The NExT Pipeline has two major arms, the Chemical Biology Consortium, which acts as the drug discovery group, and the Frederick National Lab, which handles development and good manufacturing practice production. Every step in trial development requires agreements, said Doroshow, with the CRADA as one example. He said that it used to take more than a year to negotiate a CRADA with companies, but after implementing a 6-month deadline, agreements were finalized more quickly. However, CRADAs are typically not flexible, and this sometimes leads to renegotiation of settled issues, which extends the timeline. He cautioned that business decisions will override scientific rationale in partnerships, so clarifying roles and responsibilities is critical to realizing success.

Gail Eckhardt, associate dean of experimental therapeutics at Baylor College of Medicine, described the Experimental Therapeutics Clinical Trials Network (ETCTN), which launched in the 1990s and focused in part on providing clinicians and academics access to clinical trials and mentorship for junior investigators. ETCTN aims to evaluate experimental therapies in early-phase clinical trials using a “coordinated, collaborative, and inclusive team-based approach.”¹¹ One challenge has been obtaining drugs from industry partners for studies, but overall, she said it has been a very effective program. She offered several suggestions for strengthening PPPs in the next generation of the ETCTN, such as adding trial sites in community clinics, developing joint clinical research training opportunities, providing more access to mentorship outside home institutions, and creating academic training rotations in FDA and industry.

Matthew Cooney, vice president of Therapeutic Development for Oncology at Tempus Labs, Inc., described how Tempus uses integrated electronic health record (EHR) and genomics data to conduct efficient collaborative trials. He said the company has developed a clinical trials network with more than 85 sites and more than 100 pharma partners, and it can rapidly open hundreds of trials because of standardized contracts and budgets, a Central Institutional Review Board,¹² and other efficiencies of scale. He added that

___________________

¹⁰ See https://next.cancer.gov/about/default.htm (accessed January 29, 2024).

¹¹ See https://ctep.cancer.gov/initiativesprograms/etctn.htm (accessed January 29, 2024).

¹² See https://www.fda.gov/regulatory-information/search-fda-guidance-documents/using-centralized-irb-review-process-multicenter-clinical-trials (accessed April 3, 2024).

the company’s EHR integration enables rapid identification of patients who are eligible for trials.

Box 3 outlines three examples of trials conducted via PPPs in oncology.

REALIZING THE POTENTIAL OF PUBLIC–PRIVATE PARTNERSHIPS IN CLINICAL CANCER RESEARCH

Many speakers cited the potential of PPPs to achieve scientific goals that would not be possible for any entity to do alone. For example, Gopalakrishna said that clinical sites and nonprofit organizations can gain access to expert project management and greater access to novel drugs and technologies through partnerships. PPPs also provide opportunities to diversify clinical research and may have a greater impact on outcomes that are meaningful to patients with cancer.

Diversifying Clinical Trials and Reducing Health Disparities

Several speakers said evidence suggests that PPPs can help address the lack of diversity in clinical trials. Aida Habtezion, chief medical officer and head of Worldwide Medical & Safety at Pfizer, Inc., described Pfizer’s analysis of U.S. clinical trial demographics between 2011 and 2020. Over the decade, Pfizer found that its oncology trials specifically had underrepresentation of African American and Hispanic or Latino populations (Rottas et al., 2021). Through partnerships, Pfizer is developing ways to engage communities and train investigators to improve inclusion and diversity among trial participants, staff, and investigators.¹³

Habtezion highlighted Pfizer’s pursuit of a more equitable and inclusive end-to-end research and development process, launching the Institute of Translational Equitable Medicine in 2021.¹⁴ The goal is to embed equity broadly, beyond clinical trial diversity. For example, she said that many of the biobanks that the industry uses for genome discovery studies are heavily skewed toward Caucasian populations. To address this limitation, one of the pilot projects the Institute of Translational Equitable Medicine is leading is an African-Caribbean Cancer Consortium,¹⁵ a collaboration that leverages a transcontinental network of scientists, oncologists, and health professionals to

___________________

¹³ See https://giving.columbia.edu/columbia-university-and-pfizer-establish-clinical-trials-diversity-initiative (accessed February 1, 2024).

¹⁴ See https://www.pfizer.com/news/articles/pfizer_s_institute_for_translational_equitable_medicine_addresses_equity_in_health_research (accessed February 2, 2024).

¹⁵ See https://ac3online.org/about-ac3 (accessed February 2, 2024).

develop a cancer genome registry of treatment outcomes for people of African ancestry. She said that this is critical to understanding disparities in outcomes. For example, breast cancer is not more common among women of African descent, but their breast cancer mortality rate is 40 percent higher compared to White women (Giaquinto et al., 2022). Additionally, Habtezion referenced Pfizer’s ongoing efforts to return data to trial participants, empowering them to make more informed decisions about their care in consultation with their clinicians.

Lawrence Shulman, director of the Center for Global Cancer Medicine at the University of Pennsylvania Abramson Cancer Center, also highlighted opportunities for PPPs to improve access and reduce disparities. Tremendous efforts are ongoing to facilitate access to drugs and clinical trials in more rural areas in the United States, said Shulman, but there is still work to do. Worldwide, the majority of patients with cancer live in resource-constrained settings and often lack access to clinical trials. Additionally, most U.S. patients with cancer are cared for in community settings, many of which are rural. Patients treated at these smaller community hospitals have worse outcomes, with survival for breast and lung cancer inferior to that of patients treated at academic and NCI-designated cancer centers (Shulman et al., 2018). Patients everywhere should expect accessible, affordable, high-quality care for treatable and preventable cancers, Shulman argued.

Cancer mortality is not evenly distributed throughout the world, he said. For example, sub-Saharan Africa has very high breast cancer mortality, and access to therapies that have proven effective in the United States is limited in this region (Anyigba et al., 2021; Barrios et al., 2023). To address some of these challenges and disparities, he shared an example of a global partnership to treat a large number of patients with chronic myeloid leukemia in Rwanda and elsewhere. Novartis supplied the drug; Cepheid provided diagnostics; the Max Foundation facilitated the delivery of drugs; and Partners In Health, together with the Rwandan Ministry of Health, delivered care (Morgan et al., 2022). He emphasized this model of PPP as an effective strategy to bring life-saving care to people who do not have the ability to pay for high-cost care.

Potential for Impact on Cancer Outcomes That Matter to Patients

Stacie Lindsey, founder and chief executive officer of the Cholangiocarcinoma Foundation, highlighted what is important to patients in clinical trials. Patients want to maximize overall survival with new therapies but also want a better quality of life, including fewer hospital admissions, fewer side effects so that they can remain on an effective therapy, the ability to return to work, and quality time with family and friends. They also want access to novel treatments without having to first try an ineffective standard of care. Lastly, she

said patients want some sense of control and hope, along with support and validation of their decisions.

Lindsey added that patients support meaningful use of their personal health data and want information reported to them so that they can understand how it is used. She also said that patients value the opportunity to partner in research and called for patient-centered research that seeks patient input early and often, from concept to final protocol. Additionally, patients want broader inclusion and exclusion criteria, acceptable trial designs with crossover arms (Selig et al., 2023), and 2:1 randomization.¹⁶ She shared a recent example of a trial that had a negative result, but researchers wanted access to the tissue samples from the enrolled patients (more than 600) for biomarker validation studies. Upon approaching the sponsoring company to request samples, they learned that everything had been destroyed, which Lindsey called a missed opportunity. Including patient goals and desires in trial and protocol designs, especially within a PPP, can result in more participation and potentially generate results that are more impactful to the patient.

Gwen Nichols, executive vice president and chief medical officer at the Leukemia and Lymphoma Society (LLS), highlighted the role of the patient advocate and shared a story from her experience during the pandemic. During this time, mortality rates for patients with blood cancer were quite high among those with COVID-19 (Mato et al., 2020). In addition, vaccine safety and efficacy were unclear in the specific blood cancer population, many of whom were over 65 years old. LLS wanted to quickly put together a registry to be able to generate data and systematically share information with patients, but academic collaborators estimated that it would take up to 12 months to begin trials. This was too long, so LLS worked with the biotech company Invitae’s Ciitizen platform¹⁷ to enroll 12,000 patients in just 6 months. Six thousand patients also provided EHR access if needed. Within 4 months, safety data were available on more than 2,500 patients,¹⁸ and it was possible to determine which patients were not generating a sufficient antibody response (Greenberger et al., 2022). “We published as rapidly as we could” so patients could have access to these results, said Nichols, and nurses and social workers were available to answer questions.

___________________

¹⁶ 2:1 randomization increases the chance that a patient receives the experimental treatment from 50% (1:1) to 67% (2:1) (Wu et al., 2014).

¹⁷ Invitae divested from Ciitizen Health in December 2023. The platform is now run by an independent company, Ciitizen Health. See https://ir.invitae.com/news-and-events/press-releases/press-release-details/2023/Invitae-Divests-Ciitizen-Health-Data-Platform-and-Implements-Further-Cost-Cuts/default.aspx (accessed May 28, 2024).

¹⁸ See https://www.lls.org/news/covid-19-vaccine-safety-among-blood-cancer-patients (accessed February 7, 2024).

Several speakers also said that another consideration for what matters to patients is the type of endpoints used in trials. Many PPPs have been involved in validating surrogate or correlative endpoints. Michael Morris, prostate cancer section head at the Memorial Sloan Kettering Cancer Center, said that developing interim endpoints is critical to get answers earlier in trials, especially for diseases with slower progression and longer survival times. He added that public education is important so that patients understand how endpoints are implemented. Richard Schilsky, professor emeritus at the University of Chicago, agreed, saying that in addition to survival, other endpoints important to patients should be measured, including assessing how patients feel. Allen suggested using a uniform measurement strategy in cancer trials so that results can be aggregated and compared.

NAVIGATING CHALLENGES IN PUBLIC–PRIVATE PARTNERSHIPS FOR CLINICAL CANCER RESEARCH

“The challenge is really forging more efficient, effective relationships,” said Eckhardt. Frustrations emerge without an honest broker to coordinate among partners in academia, pharma, and government, she said. John Byrd, the Gordon and Helen Hughes Taylor professor and chair of internal medicine at the University of Cincinnati College of Medicine, agreed, adding that it is necessary to improve communication at all levels. Allen said that the field could find a better way of collectively identifying and prioritizing evidentiary gaps at the outset and then creating the partnerships needed to solve them. Several speakers highlighted challenges in general operations, regulatory hurdles and policy changes, data sharing, and difficulties embedding research within cancer care delivery.

Overcoming Operations Challenges to Accelerate Partnership Success

Bringing several partners together across public and private sectors with varying priorities and streamlining multiple processes to optimize timelines and outcomes has inherent difficulties. Key challenges in general operations of partnerships and collaborations include aligning different interests among partners and maintaining quality control and agreed-upon timelines.

Aligning Differing Interests and Objectives

In any collaboration, bringing partners together can have challenges, said Habtezion, especially when there is misalignment in strategies, goals, or timelines. To minimize challenges, she reiterated Adam’s suggestion to

level-set shared expectations and clear roles at the beginning of a partnership. Bidirectional communication throughout the process is also critical, Habtezion added. Sigal said that partners needed to learn to work together and understand the objectives while developing Lung-MAP, adding that despite excitement for the collaboration, no one wanted to change their culture and challenge the status quo.

Vassilis Golfinopoulos, headquarters director of the European Organisation for Research and Treatment of Cancer (EORTC), said EORTC strives to have one set of systems (e.g., software or standard operating procedures) to avoid complexity in partnerships, but he added that a program’s outcome is ultimately determined not by the system, but by the drive to collaborate. He also described a recent attempt to reduce trial complexity: EORTC put forth a rationale to justify the limited collection of adverse events in a clinical trial submission because the toxicity profile of the drug was well known. Although the ethics committees reviewing the submission were open to this approach, questions were still raised about how to do a benefit–risk assessment without collecting these data in the trial. Morris highlighted the gap between what partners think FDA wants in terms of complexity and levels of data and what FDA actually needs, but Harpreet Singh, associate director of the FDA Oncology Center of Excellence, cautioned that a balanced approach is needed because FDA is responsible for finding safety signals, and there is still a need for risk mitigation.

Herbst noted an operational challenge in international trials due to policies within companies related to drug shipments across borders. Dancey agreed that this was a key challenge in her experience in Canada, which is why she emphasized having decision makers and operational leaders involved who could problem solve. She said that many company partners were multinational, and sometimes the distribution to Canada was not well delineated in the contracts, resulting in several months’ delay in getting a drug.

Maintaining Quality Control and Adhering to Timelines

Accelerating lengthy clinical trial timelines has also been an ongoing challenge because of the extensive discussions to finalize agreements before trials can begin. Sigal said that this was one of the biggest challenges with Lung-MAP, which took 18 months to implement. Trial leaders realized that they needed to work closely with FDA and NCI to accelerate processes where possible. Contracting was one of the biggest issues because finalizing the contract with each company took a very long time. Pragmatica-Lung had focused and streamlined development and was simplified to just ask one question about overall survival, but it became difficult to adapt the traditional culture of clinical trials to simplify processes for just one endpoint.

Heidi Smith, vice president of the Center of Operations and Research Excellence at Novartis Pharmaceuticals, suggested establishing standard operating procedures and best practices. Eckhardt added that establishing a 2-week turnaround on contracts as a best practice can raise the bar and get novel drugs to patients more quickly. Gopalakrishna said that one of the biggest challenges is ensuring that data collected in trials are adequate to submit for registration purposes. He highlighted the differences in collecting and formatting data for industry-sponsored trials versus academic research collaborations. Because reformatting the data can add to the timeline, efforts are underway to harmonize collection methods so that the database can be used for both academic and regulatory purposes.

Gerberding agreed on the importance of speed in trials and partnerships but said that once a group has agreed on objectives and a governance structure, execution gets easier. She reiterated that many of the challenges are bureaucratic processes based on cultures. Pharmaceutical developers are risk averse, she said, so companies have lawyers thoroughly review all parts of agreements. She suggested the use of a master agreement with task orders to speed up the process for partners that have already been involved in collaborations. Habtezion agreed and said that Pfizer is starting to look at master agreements. Pazdur agreed that speed is one of the biggest challenges but added that partners often do not want to relinquish control. As another potential solution, Gerberding said that FNIH has initiated a PPP of lawyers to support developing master agreements with private-sector partners.

Regulatory and Policy Hurdles

Singh discussed opportunities and challenges from a regulatory perspective when working with PPPs. “We cannot force companies to do trials they don’t want to do,” she noted, or “force entities to work together even when we do find that there is a common interest for patients.” Cost is also a challenge because FDA cannot take costs into account in decision making. FDA does have some flexibility in reviewing submitted data quality and can be more flexible with drugs that are more well known. She added that less data does not mean inferior data, and a smaller, streamlined dataset can still provide high-quality data. The perception is that less data implies problems with data integrity, which could be clarified. Morris suggested updating the framework in which FDA operates. He said that currently, FDA can only respond to trial design proposals and cannot suggest trial designs up front, but its representatives are very innovative and could contribute more to the process with fewer constraints. Singh replied that FDA has a broad view of the work of all competitors, and FDA representatives sometimes need to recuse themselves from discussions to avoid inappropriately influencing decisions. Doroshow said

that tradition is a hindrance to changing policy and cautioned that the whole system could collapse because of excess data requirements.

Singh said it is important to think about registrational intent when designing a trial, noting that on several occasions data from a clinical trial that produced positive results but was not designed for product registration were submitted to FDA for review. However, she said FDA had no advance knowledge of the trials or had not seen the design or protocol beforehand, which introduces challenges that slow the process. If the trial was not designed with registrational intent at the outset, explained Singh, it is not possible to go back and retroactively collect data that are fit for the purpose of FDA labeling. She said that the takeaway is that even if investigators do not know up front what part of a study may have registrational intent or the primary goal is just to publish the results, researchers should still consider that the trial may potentially inform a product label, and there is no harm in approaching FDA early to discuss that possibility.

Potential Impact of Federal Policy Changes on Research and Development

Josh Bilenker, chief executive officer and cofounder of Treeline Biosciences, provided an investor perspective on PPPs, saying that compared to the public sector, private capital devotes approximately three times as much funding for biomedical research and development. However, he explained that on the private side, the expected return has to be commensurate with the risk premium¹⁹ of the investment. When thinking about investing in biomedical technology development, Bilenker noted that money spent today will not result in revenue for 5–10 years. Adding to the challenge, dollars earned in the future are worth less when interest rates are higher, he said.

Bilenker said that this is important to consider when thinking about the Inflation Reduction Act (IRA) of 2022.²⁰ Provisions of the IRA aim to lower drug costs and reduce patient out-of-pocket spending (e.g., a yearly cap on patient out-of-pocket spending for Medicare Part D prescription drugs), and have the potential to improve health care cost containment.²¹ Bilenker described case studies from ZS Analysis, which supports the commercial launches of most companies, including more than 90 percent of the recent oncology launches. He said that the company’s analysis suggests that companies providing a small-molecule drug to Medicare recipients could potentially

___________________

¹⁹ Risk premium is the greater expected return when investing in riskier ventures. See https://www.forbes.com/advisor/investing/risk-premium (accessed March 29, 2024).

²⁰ See https://www.congress.gov/bill/117th-congress/house-bill/5376/text (accessed April 2, 2024).

²¹ See https://www.cms.gov/inflation-reduction-act-and-medicare (accessed April 2, 2024).

lose as much as 50 percent of that product’s cumulative revenue due to the implementation of the IRA.²² Bilenker argued that while the IRA elements do not begin until year nine, the first 3–5 years of a product are often spent just creating clinician awareness, so he said this effect will come into play just as the product starts to generate a profit. When the regulatory environment gets tough, the smaller companies pay the biggest price, he added.

Bilenker also commented on the march-in rights provision within the Bayh-Dole Act of 1980.²³ It is a safeguard that gives the federal government the authority to require relicensing a patent that resulted from any amount of federal support, if good faith efforts are not being made to commercialize the research.²⁴ The goal is to incentivize invention from federally funded IP so that it can be made available to the public on reasonable terms. The National Institutes of Health (NIH) have never utilized the march-in provision, but Bilenker cautioned that if relicensing is continually requested, fungible capital may flow away from biotech into other industries.

Use of Personal Health Information

Geoffrey Oxnard, vice president of clinical development of Loxo Oncology at Eli Lilly, raised the question of whether patients would prefer to have specimens from clinical trials discarded when the study is completed or have more opportunities for them to be used in subsequent studies. Lindsey said that the eight biggest clinical trials for cholangiocarcinoma²⁵ all lacked patient consent for using the diagnostic images collected for other studies. To advance care, she argued that those images are needed for research, and described language developed by the Cholangiocarcinoma Foundation’s industry council that allows patients to authorize access to their data for further research. Because of the small patient populations with rare diseases, she said that it is wasteful to discard data after just one application. However, Barbara Bierer, professor of medicine at Harvard Medical School, highlighted the potential downsides to reusing patient data. As a community, it will be important to find better ways to educate patients about research and obtain informed consent, ensuring that people understand their choices, she said.

___________________

²² See https://www.zs.com/insights/navigating-the-inflation-reduction-act-next-moves-for-biopharma (accessed February 13, 2024).

²³ See https://grants.nih.gov/grants/bayh-dole.htm (accessed March 28, 2024).

²⁴ For more information on march-in rights, see U.S. Code, Title 35, Part II, Chapter 18, Section 203 https://uscode.house.gov/view.xhtml?path=/prelim@title35/part2/chapter18&edition=prelim (accessed April 22, 2024).

²⁵ See https://www.cancer.gov/types/liver/bile-duct-cancer (accessed April 2, 2024).

Overcoming Data Sharing Challenges Within Public–Private Partnerships

Singer highlighted data sharing as a key challenge when developing PPPs, noting the difficulty of balancing the need to make data publicly accessible while also recognizing the proprietary needs of the private sector. Several aspects of data sharing challenges were highlighted throughout the workshop, including accessibility of the knowledge generated from PPPs, information blocking, misaligned academic incentives, lack of standards to enable data sharing through interoperable systems, a need for greater patient engagement and education, and the complexity of emerging data sharing laws.

Accessibility of Knowledge Generated from Public–Private Partnerships

Penny Burgoon, director of the Office of Policy Communications and Education at NIH, provided a perspective on the role of government in sharing data. She cited the NIH Scientific Data Sharing platform²⁶ as a central source of guidance related to NIH data sharing policies and offering access to the repositories for scientific and genomic data. To highlight the benefit of making data transparent and available, she described the National COVID Cohort Collaborative (N3C)²⁷ as a controlled access database of EHR information to study COVID-19. N3C consists of more than 75 institutions building a centralized data resource with an intent to carefully balance access to essential data for COVID-19 research with protecting the privacy of individuals represented by that data. The platform was launched and active by September 2020, and it has nearly 100 sites contributing data to the repository, with 370 sites signing data user agreements. N3C contains deidentified EHR data for more than 20 million people, including more than 8 million COVID-19 positive cases, said Burgoon. It is a data enclave,²⁸ so the users cannot remove the data but can retain the results of their analyses. It has also enabled research on long COVID. With this wealth of data, she said investigators were able to analyze the “Paxlovid rebound” that occurs in some infected individuals; for more than 50,000 patients receiving Paxlovid, researchers found that less than 1 percent of the patients receiving the drug experienced a rebound of symptoms, she said. Gerberding also mentioned that data generated through FNIH programs

___________________

²⁶ See https://sharing.nih.gov (accessed February 13, 2024).

²⁷ See https://ncats.nih.gov/research/research-activities/n3c/overview (accessed November 28, 2023).

²⁸ A data enclave is “a secure network through which confidential data, such as identifiable information from census data, can be stored and disseminated.” See https://www.nnlm.gov/guides/data-thesaurus/data-enclave (accessed March 29, 2024).

are available to all scientists, regardless of partnership status, which is a critical element of their approach.

Information Blocking

Deven McGraw, lead for data stewardship and data sharing at Invitae/Ciitizen, said that a data policy that could advance research is the prohibition of information blocking, which was included in the 21st Century Cures Act.²⁹ Rules such as these can unlock sharing EHR information in major ways, she said, because they create an expectation that data will be shared. Information blocking rules cover numerous sources of data, including those from clinicians, EHRs, and health information exchanges (HIEs). Penalties have only recently been established for certified EHR vendors and HIEs, up to $1 million per violation, but without a single enforcement case yet, she said. McGraw said that information blocking will almost always be implicated when it interferes with the ability of patients to access their own information or exchange and use it without special effort. She posited that information blocking rules could also be leveraged for other types of research models that do not rely on direct patient involvement (Bitterman et al., 2022). Kristen Rosati, partner at Coppersmith Brockelman PLC, said that by creating the assumption that data will be shared, the information blocking rule could be a paradigm shift for research.

Aligning Academic Incentives

Bierer provided a perspective on data sharing from an academic research institution, noting the continuum of interest in PPPs across academia, ranging from modest enthusiasm to absolute resistance. She said that institutions need established policies and processes for data access, compliance, and oversight, as well as supportive systems for investigators, including for data anonymization. Bierer described Vivli, a global clinical research data sharing platform,³⁰ as one such system.

The return on investment for PPPs can be uncertain, and potential revenue from licensing may be overvalued, Bierer said. Furthermore, a significant cost is associated with compliance and formatting data made available for additional use. Risks also differ depending on the type of data (e.g., derived

___________________

²⁹ Information blocking is intentional or unintentional interference with electronic health information. See https://www.healthit.gov/topic/information-blocking and https://www.fda.gov/regulatory-information/selected-amendments-fdc-act/21st-century-cures-act (accessed February 13, 2024).

³⁰ For more on Vivli, see https://vivli.org (accessed November 30, 2023).

from basic science, translational, or clinical research). On the investigator side, reputational issues also need to be considered. Within the current structure, she said that investigators benefit from using shared data, but academic incentives are lacking for the trialist to make that data available. Research institutions also fear that others will not be able to reproduce results if the data are shared. Another challenge is that many institutions are developing policies that prohibit investigators from sharing data without prior institutional clearance.

Bierer described a positive example of an institution that is doing this well, noting that the Center for Open Science³¹ has an open data badge, which is the only incentive shown to facilitate sharing (Kidwell et al., 2016). Data sharing and collaboration should be valued in academic promotion criteria and as a contribution to grants and other funding mechanisms, Bierer argued. If a traceable system existed, then research could be evaluated by not only primary publication, but all downstream publications made possible by that original data. Supportive institutional leadership and culture are also needed to incentivize data sharing—including clarifying needs to enable Findable, Accessible, Interoperable, and Reusable (FAIR) data sharing.³²

Patient Engagement, Privacy Protection, and Standardization in Data Sharing

Across various types of PPPs, several speakers called for greater patient engagement and education, as well as considerations for protecting patient privacy. Nichols emphasized the importance of patient awareness and education about how data are used.

Rosati highlighted concerns regarding the use of real-world clinical data in research without patient consent, especially the risk of reidentification given the lack of U.S. federal laws prohibiting the reidentification of individuals present in deidentified datasets. She said that many institutions are realizing that they need to go beyond the current legal requirements to build patient trust and protect the collective ability to use deidentified data for research. Burgoon noted a concern about reidentification in N3C, so they developed a model of shared responsibility to ensure that regulations, policies, and a code of conduct were followed. For example, she said that NIH sought a formal consultation with tribal nations to ensure that the N3C exerted respectful and responsible data stewardship over American Indian/Alaskan Native data. This includes the deidentification of tribal affiliation.

Morris said that people know that much of their personal data outside of the medical world is already being shared. For example, their cars and phones are sending location data, but they think they get something for it in return.

___________________

³¹ See https://www.cos.io (accessed February 16, 2024).

³² For more on FAIR, see https://www.go-fair.org (accessed November 30, 2023).

However, a clear and positive return for sharing medical data is often still unclear to people, he noted. These data are collected by commercial entities every day, but the medical community has been absent from the discussion and could leverage the tools already in place, he said.

Understanding the Complexity of Data Sharing Laws

Rosati provided an overview of the legal landscape and the increasing complexity of data sharing laws. Regulations promulgated under the Health Insurance Portability and Accountability Act,³³ federal regulations regarding information about substance use, and regulations governing research, such as the Common Rule,³⁴ all need to be considered, she said, adding that state laws are also becoming more influential and can include laws that regulate consumer data privacy, health information confidentiality, and genetic privacy. For those involved with international data sharing collaborations, she said that it is imperative to consider other countries’ laws as well, including Europe’s data privacy and security law, the General Data Protection Regulation.³⁵ The European Union also has new legislation, the European Health Data Space,³⁶ which will create an ecosystem composed of rules and common standards aimed at empowering individuals to take control of their own health data and providing consistent guidance for the use of health data across research and policy making.

Rosati discussed the challenges associated with evolving standards for deidentifying data, which is essential for many data sharing agreements. She explained that state laws are increasingly influential in determining whether data can be shared. For example, many states are revising their genetic privacy laws. State-level consumer data privacy laws are also increasing, beginning with the California Consumer Privacy Act of 2018,³⁷ on which many other state laws are being modeled. These state privacy laws pose tough challenges for research collaborations across states, she explained, because the state laws have so much variation. In closing, she emphasized the increasing complexity

___________________

³³ See https://www.cdc.gov/phlp/publications/topic/hipaa.html (accessed February 16, 2024).

³⁴ See https://www.hhs.gov/ohrp/regulations-and-policy/regulations/45-cfr-46/index.html (accessed March 29, 2024).

³⁵ For more on the General Data Protection Regulation, see https://gdpr.eu (accessed November 28, 2023).

³⁶ See https://health.ec.europa.eu/ehealth-digital-health-and-care/european-health-data-space_en (accessed December 1, 2023).

³⁷ See https://leginfo.legislature.ca.gov/faces/codes_displayText.xhtml?division=3.&part=4.&lawCode=CIV&title=1.81.5 (accessed May 9, 2024).

of the entire data sharing landscape and recommended involving legal experts early in the process to help determine the appropriate path forward.

EMBEDDING RESEARCH IN CARE DELIVERY

A number of speakers spoke on the challenge of successfully executing partnerships is embedding clinical research into care delivery, so patients can enroll in trials while still receiving high-quality routine clinical care as needed. Several speakers discussed how disparities in access for certain populations have made it difficult for some PPPs to achieve adequate patient enrollment in trials, as well as the logistics and workforce challenges that prevent a seamless and sustainable execution of research partnerships.

Historical Lack of Access to High-Quality Care and Clinical Trials

Multiple speakers discussed the goal of diversifying clinical trials in light of the historical lack of access to high-quality cancer care and clinical trials among some patient populations. Dahut and Pazdur recommended implementing more decentralized trials closer to where patients live to increase trial enrollment. Eckhardt agreed that access for people in rural communities should be prioritized, especially for early-stage clinical trials. Sigal and Habtezion focused on engaging patients, ensuring the trial is meaningful to them, and embedding their perspective in the entire process. Singer noted that NCI has started several initiatives focusing on direct patient engagement based on a primary recommendation of the Cancer Moonshot.

Logistics and Workforce Challenges

Successful PPP implementation also depends on developing and maintaining an appropriate workforce and the logistics spanning different sectors and locations, said Eckhardt, adding that in the past, there have been mechanisms to support physician-scientists, but the pipeline for clinical investigators is increasingly at risk. This affects not only academic investigators but the entire field. Clinical revenues are based on the relative value unit (RVU)³⁸ of care services, so clinicians are under financial pressure to meet RVU targets, she said. In oncology, chemotherapy revenue is driving the clinical activity of clinicians, which leads to decreased clinical research productivity. In addition, the path to promotion and tenure for clinical investigators is unclear. Lastly,

___________________

³⁸ RVUs are units used by Medicare and Medicaid Services to calculate the work done by clinicians and resulting reimbursement (Baadh et al., 2016).

Eckhardt noted the lack of mentorship for early-career physician-scientists because mentors are under the same fiscal pressures and often do not have the time to spend with trainees. For the next generation, she said that a collaborative partnership is necessary to develop and retain the workforce needed to design and run clinical trials.

With this pressure to maximize RVUs and clinical productivity, getting community oncologists to participate in clinical trials can also be challenging. Carolyn Muller, professor and chief of the Division of Gynecology Oncology at the University of New Mexico Comprehensive Cancer Center, and Melissa Johnson, medical oncologist at the Sarah Cannon Research Institute, discussed approaches to streamline the process and make participation easier for busy oncologists, such as using EHR tools to simplify trial enrollment tasks and delegating informed consent procedures to research nurses. Muller and Johnson also called attention to the shortage of nurses and research coordinators to support clinical trial participation. Muller noted the lack of coordinated efforts to create career pathways and engage schools for the health professions to promote clinical research roles.

HIGHLIGHTING OPPORTUNITIES TO IMPROVE PUBLIC–PRIVATE PARTNERSHIPS FOR CLINICAL CANCER RESEARCH

Many speakers highlighted key opportunities to advance PPPs and improve outcomes for patients in clinical cancer research. These included building trust and alignment among partners, improving patient engagement and diversity, optimizing technology and data sharing, creating sustainable funding models, and developing a robust workforce.

Building Trust and Alignment Among Partners

Herbst and Schilsky highlighted the mutual benefit that is necessary for successful PPPs, saying that all participants must see some benefit in leveraging the resources to enable scientific achievements that would otherwise be impossible. Although partnerships can mitigate some risks in trials, collaborations can also heighten risks in other ways, so it is important to outline and balance the risks early in the process, considering the perspectives of all partners. They emphasized that trust and communication, agreed-upon approaches, leadership, and accountability are critical, and reiterated the importance of enabling fair, transparent public access to the insights generated by these partnerships.

Pragmatism and Flexibility in Trial Design and Operations

Edith Perez, chief medical officer at Bolt Biotherapeutics, and Smith highlighted the importance of team building and flexibility, keeping patients at the center of initiatives, and reducing data collection requirements and burdens wherever possible. Morris emphasized the need to improve cooperation among regulatory agencies, academia, and clinicians to alleviate the burden on patients with cancer, particularly the financial strain and inequitable access to care and clinical trials. He said that clinical trials often suffer from overly complex designs, excessive data collection, and restrictive eligibility criteria, resulting in drug approval delays and limited trial access (see Figure 2).

Morris described a successful example of collaboration among the Department of Defense, an academic consortium supported by a prostate cancer foundation, and FDA that led to the establishment of new surrogate endpoints for clinical trials. This initiative enabled the development of validated endpoints for FDA, which led to faster drug approvals for industry, successful research for academia, and, importantly, earlier patient access to novel drugs (Scher et al., 2016). Yet the consortium lacked agreement on IP or commercialization, raising questions on how such a collaborative effort could be systematically replicated. To address these challenges and foster innovation, Morris described the establishment of the NCI Clinical Trials Innovation Unit (CTIU),³⁹ which aims to streamline processes, enhance efficiency and equity, and reduce regulatory hurdles. He advocated for strategic collaborations that leverage the strengths of multidisciplinary groups like the CTIU to modernize clinical trial designs and improve patient outcomes.

Singh mentioned the flexibility around data collection that was implemented during the COVID-19 pandemic, noting that FDA developed guidance with the fastest timeline she has ever seen. Yet, much of what was in the guidance did not change any policies but rather outlined how to proceed in a more decentralized way to bring drugs to patients faster within the existing regulatory framework. She noted that FDA will likely develop guidance informed by the COVID-19 era.

As another example of simplifying trial design and streamlining processes, Byrd described the BEAT acute myeloid leukemia (AML) trial⁴⁰ that brought together diverse partners to conduct a platform trial⁴¹ in which patients with AML were assigned to treatments based on a genetic analysis. LLS held the

___________________

³⁹ See https://www.cancer.gov/research/infrastructure/clinical-trials/ctiu (accessed February 16, 2024).

⁴⁰ See https://www.lls.org/bringing-precision-medicine-aml-patients (accessed February 16, 2024).

⁴¹ A platform trial combines characteristics of umbrella and basket trials in which multiple treatments and/or multiple treatment groups can be studied (Lu et al., 2021).

investigational new drug application and used contract research organization (CRO) monitoring to ensure generation of rigorous data that could be submitted to FDA. As a central rule, he said they did not attribute individual credit and tried to stay nimble. The first success was realizing that this could be done safely, said Byrd. Second, the trial demonstrated that patients who were assigned to genetically guided precision therapy did better than patients receiving standard therapy (Burd et al., 2020). Data from this study changed clinical practice through numerous publications that provided input for the National Comprehensive Cancer Network guidelines and were used to validate futility metrics for terminating studies early when a therapy is not working. As takeaway lessons, Byrd said that the trial demonstrated that a nonprofit can successfully run disease-specific platform trials for patients with a rare disease by staying nimble at the top and enabling rapid decisions without an individual need for credit.

Improving Patient Engagement, Trust, and Diversity

Summarizing discussions related to patient engagement and trust, Neal Meropol, vice president of research oncology at Flatiron Health, and Shulman emphasized that inequitable access to quality cancer care and clinical trials remains problematic, and existing community-based models often lack scale. To achieve the goals of improved patient engagement, trust, and diversity within cancer research, they highlighted the value of providing patient education and navigation, building community-based research, designing for equity early in the process, and conducting postmarket research.

Improving Patient Trust

Many speakers said that successful PPPs depend on patient trust in the system and those conducting the trials. They commented on the challenges of building trust for data sharing in communities that historically had reasons to mistrust the medical establishment. Otis Brawley, Bloomberg Distinguished professor of oncology and epidemiology at Johns Hopkins University, referenced the NIH Revitalization Act of 1993,⁴² mandating increased minority participation. Some institutions have a long history of providing services in the community, he said, such as the University Hospitals of Cleveland and Henry Ford Hospital in Michigan—both of which have over accrual of minorities in clinical trials.

___________________

⁴² See https://www.congress.gov/bill/103rd-congress/senate-bill/1 (accessed February 16, 2024).

McGraw outlined how Invitae/Ciitizen leveraged information technology to implement models in which patients can contribute data for research purposes. Patients already contribute patient-reported outcomes, she said, but only more recently have people thought about ways that patients can contribute their medical data in EHRs. Patient organizations are increasingly involved in research and connecting directly with life sciences companies. The 21st Century Cures Act contained several provisions poised to disrupt the way data are gathered for health research and put the patient in the driver’s seat, said McGraw. As one example, EHR platforms used by clinicians in the United States are now required to have open standard Fast Healthcare Interoperability Resources⁴³ application programming interfaces that patients can connect to.

Multiple speakers said that another method of building trust between researchers and participants is returning research results to patients, an ethical practice that can enable them to see the direct benefits of their involvement in research. However, it is also a complex issue because not all results may be meaningful or useful to participants, and sharing data may have regulatory or logistical challenges. Bierer noted that Pfizer’s practice of allowing patients to view their clinically valid results is a step in the right direction.

Byrd added that it is important to establish patient trust at the outset, considering the socioeconomic factors that may limit a patient’s ability to participate in a trial. He also said that it was important to expand the diversity of the clinical workforce to reflect the patients they are serving and recruiting for trials. Adam commented that no “one-size fits all” approach exists; what works in Black communities may not work for Hispanic or Asian communities. Gwen Darien, executive vice president of patient advocacy at the National Patient Advocate Foundation, said that many patients do not participate in trials because they are not asked. She said that patients often feel that they are not trusted by clinicians who make assumptions about what they want and can afford. It is important to talk about trust bidirectionally, she said, because patients need to trust the system, but it is also necessary for the system to trust patients.

Building on Community-Based Research Consortia Models

As another approach to improving trial diversity and access, several speakers shared examples of community-based consortia focused on oncology research. Johnson described the Sarah Cannon Research Institute, a community-based clinical research partnership between oncology care clinicians and research sponsors (nonprofits, pharma, and biotech), with 250 sites across 24 states. She

___________________

⁴³ This is a platform that shares disparate EHR data between clinicians and patients. See https://ecqi.healthit.gov/fhir (accessed February 16, 2024).

reiterated that partnerships need to be based on aligned interests and incentives, noting that this partnership enables patients to receive cutting-edge therapies close to home by staying in the local health care system. Some key elements of Sarah Cannon services, such as patient navigation, are not reimbursable by health care insurers but rather funded by research partners. She described the Institute’s impact on cancer drug development, with more than 1,000 trials actively enrolling patients and more than 4,500 registered patients participating in trials each year.⁴⁴ She said that the Institute also participated in developing 29 of 37 cancer therapies approved for adults by FDA in 2022. She emphasized the benefit of shared efficiencies across its network—one budget, one contract, and centralized education materials can all be shared across oncology practices, which is appealing to practices interested in joining the consortium. As another advantage, Johnson said that Sarah Cannon is the only organization in the industry that has enabled shared data among the site management organization, the oncology-focused CRO, and personalized medicine powered by Genospace.⁴⁵ The Institute is also piloting an electronic data capture system to automate transferring routinely collected clinical EHR data that are relevant to a clinical trial.

Muller discussed the New Mexico Cancer Research Alliance,⁴⁶ one of the 46 community sites supported by the NCI Community Oncology Research Program.⁴⁷ Unlike much of the United States, New Mexico is a majority minority state, said Muller, with 21 tribal communities, but very few people are spread across an extremely large rural area. Initially created in 2003 as the New Mexico Cancer Care Alliance, engaging five community care practices, it evolved to focus on clinical research and was renamed. However, during the COVID-19 pandemic, conflicts emerged between prioritizing research or care delivery in the community. To encourage sustainability, she said that the Alliance has annual scientific retreats, a community oncology working group, and centralized services, such as regulatory processes and quality assurance within the CRO at the cancer center. Muller suggested ways to optimize partnerships, including mission-driven common goals, flexible incentives, centralization of education and staff training, and inclusion of community advocates and consumers.

Elizabeth Garrett-Mayer, vice president of the Center for Research and Analytics at the American Society of Clinical Oncology (ASCO), shared some examples of studies conducted through ASCO consortia. To bring knowledge and precision medicine to community settings where it was not easily avail-

___________________

⁴⁴ For more information on the Sarah Cannon Research Institute, see https://www.scri.com (accessed February 20, 2024).

⁴⁵ See https://www.genospace.com (accessed April 2, 2024).

⁴⁶ See https://nmcca.org (accessed February 26, 2024).

⁴⁷ See https://ncorp.cancer.gov (accessed February 26, 2024).

able, ASCO launched the Targeted Agent and Profiling Utilization Registry (TAPUR) study,⁴⁸ a precision medicine Phase 2 basket⁴⁹ trial, providing access to targeted agents via trial protocol. TAPUR provides a mechanism for collecting evidence on off-label prescriptions⁵⁰ for targeted cancer therapies. As of September 2023, she reported, they had 267 sites across 28 states, with more than 80 percent of sites based in the community. This led to a more representative patient population than is typically seen in trials, providing access to patients who otherwise might not participate in trials. Garrett-Mayer said that 25 different treatments from numerous private pharma partners have been tested, but she emphasized the need for an incentive for partner retention. Overall, TAPUR focused on good trial design elements that can be used in all sites, including those in the community. She concluded that prioritizing pragmatic elements in a trial allows more sites to participate, makes trials more accessible, and can lead to better patient inclusion and equity.

Focusing on Diversity and Equity Throughout the Process

Shulman and Brawley emphasized that it is critical to deliver high-quality care in alignment with clinical practice guidelines, equitably across populations, beyond clinical trials. Shulman described the high levels of variation in cancer therapies being offered to patients across different demographics and insurance types (Roberts et al., 2023). Brawley agreed that many patients are not receiving the existing treatments they need and that simply increasing participation in clinical trials will not overcome this disparity in treatment access.

With restrictive eligibility criteria as a persistent challenge to diversity in clinical trials, Reckamp mentioned the many discussions about broadening the criteria and emphasized the need to be more creative and think more about how to care for patients in real-world settings. Singh added that Pragmatica-Lung used a more expansive range of performance status⁵¹ in inclusion criteria.

___________________

⁴⁸ For more information on TAPUR, see https://www.tapur.org (accessed March 7, 2024).

⁴⁹ This is a type of clinical trial that tests how well a new drug or other substance works in patients who have different types of cancer that all have the same mutation or biomarker. In basket trials, patients all receive the same treatment that targets the specific mutation or biomarker found in their cancer. See https://www.cancer.gov/publications/dictionaries/cancer-terms/def/basket-trial (accessed April 2, 2024).

⁵⁰ An off-label prescription is the use of “a drug that the U.S. Food and Drug Administration (FDA) has approved to treat a condition different than your condition.” See https://www.ahrq.gov/patients-consumers/patient-involvement/off-label-drug-usage.html (accessed March 29, 2024).

⁵¹ Performance status evaluates how disease impacts a patient’s daily function (Magnuson et al., 2021).

She stressed that this approach is not advisable for first-in-human trials, but after data have been collected on enough patients to assess safety signals, pragmatic trial designs with more selective safety measures and more flexible inclusion criteria can be used. She said that everyone benefits when trials can document the experiences of a broader patient population.

As one example of overcoming challenges limiting clinical trial participation among underserved minority populations, Craig Tendler, vice president of oncology, clinical development, and global medical affairs at Johnson & Johnson Innovative Medicine, shared the findings from case studies of enrollment challenges. His team learned that many clinical trials are not open in community care settings affiliated with NCI-designated cancer centers, even if they are just a few blocks away. This informed a plan to expand access by bringing the trials to the patient instead of vice versa, mainly through supplemental study budgets for research support to optimize participation from these community sites. The plan also includes soliciting feedback from research staff about what support is required to ensure the study can be conducted at other sites and offering resources for community outreach and educational efforts, with occasional work also addressing health literacy gaps. He added that many cancer center networks have a component affiliated with the Veterans Health Administration (VHA), which offers opportunities to identify efficiencies within the large VHA network, such as implementing and executing master clinical trial agreements to reduce time spent negotiating contract terms.

Meropol stressed the need for a multipronged approach to promote equity, including interventions to improve access, enable clinical trial participation in community as well as academic settings, and promote high-quality care more broadly. He suggested the following potential actions to make trials more representative through PPPs:

• Utilizing public/nonprofit consortia infrastructure.
• Supporting training.
• Reducing regulatory hurdles for sites/investigators.
• Simplifying study designs.
• Enabling sites through technological innovations to identify and enroll patients and collect data

He also noted that FDA has released draft guidance⁵² for industry to improve enrollment of participants from underrepresented racial and ethnic

___________________

⁵² See https://www.fda.gov/regulatory-information/search-fda-guidance-documents/diversity-plans-improve-enrollment-participants-underrepresented-racial-and-ethnic-populations (accessed April 2, 2024).

populations in clinical trials, in addition to recent legislation (the Food and Drug Omnibus Reform Act of 2022).

Opportunities for Postmarket Research Public–Private Partnerships

Meropol shared examples of PPPs in postmarket research settings, highlighting the growing need for more evidence to guide routine care, given the large number of new cancer therapies that have been approved recently and the recognition that clinical trial participants generally are not representative of U.S. patient populations. He described the Clinical Trials Transformation Initiative (CTTI) Transforming Trials 2030 goals,⁵³ saying that these are all critical components of evidence generation (see Figure 3).

The postmarket setting is especially conducive to research in routine care settings, Meropol said, because clinicians have more experience with drugs and can design studies that fit into routine clinical workflows. Enabling access to clinical research in community practices and other settings is needed to address evidence gaps and ensure the participation of historically underrepresented populations. He said that shared goals for PPPs in this setting might include deriving treatment insights; understanding practice patterns; improving access; exploring predictive biomarkers; and providing an opportunity for precompetitive studies with multiple commercial sponsors.

Optimizing Data Sharing

Brawley and Rosati reviewed the misaligned incentives that undermine data sharing and collaboration and the lack of standards that create challenges for data aggregation. They noted several potential solutions, including leveraging journal and funder data sharing requirements that already exist and developing more standardized nomenclature.

Leveraging Funder and Journal Requirements for Data Sharing

Without an inherent benefit in sharing data from a trial, companies or institutions may be more likely to withhold them. However, both funders and journals have begun including requirements to make data available to ensure transparency and encourage reproducibility. Bierer highlighted the Gates Foundation as an example of a funder requiring data sharing as a contingency term of funding, with clear expectations and oversight. From a sponsor

___________________

⁵³ See https://ctti-clinicaltrials.org/who_we_are/transforming-trials-2030 (accessed March 7, 2024).

perspective, Oxnard suggested precompetitive PPPs as a solution to maximize flexibility because they can be a powerful mechanism for collaborative analysis of data from trials conducted by multiple sponsors. Tony Hickson, chief business officer at Cancer Research UK, described recent changes his funding organization made after realizing that although it collected large amounts of data, they were unable to utilize those data due to a lack of proper consent for reuse and anonymization. They developed a new data sharing policy similar to that of NIH, and although they do not mandate specific data formats or plans, they do have common data standards and require plans for how data will be accessed or requested. Commercial data partnering needs a controlled process that also embraces transparency and trust, said Hickson, and the new policy ensures that data and revenues are shared with host researchers and institutes. In addition to clarifying patient consent and data ownership, he said they are focused on ensuring that the research they support is translated into results that benefit the public.

From a journal perspective, Eric Rubin, editor in chief of the New England Journal of Medicine, recalled when genes were first being patented and many journals were encountering this type of data protection for the first time, with no set standards. As a result, researchers could publish an article on a gene sequence without necessarily revealing what the sequence was, he explained. He saw the enforcement role lying more with journals than with

funders because funders have difficulty in ensuring the data are disseminated post-award. For basic science data, journals have some standard rules already in place, which he said is a good paradigm to follow for other types of data. Kirsten Bibbins-Domingo, editor in chief of the Journal of the American Medical Association (JAMA), reviewed a study from JAMA Oncology that examined the utility and completeness of individual-participant data (IPD) and supporting documents provided from industry-sponsored cancer clinical trials through data sharing processes over a 1-year period (Hopkins et al., 2023). IPD packages were received from only 70 of 91 eligible trials (77 percent), with substantial variation in data completeness. She showed the key elements of full data sharing that are needed for secondary use in research (see Figure 4).

Bibbins-Domingo said that data sharing is almost nonexistent for academic trials, even though that requirement or expectation has been in effect for years. However, journals such as JAMA support data sharing to enable reproducibility in science, she said, adding that journals are trying to enable authors to comply with the NIH data management and sharing policy. Although people sometimes suggest that journals should mandate data sharing, she said that this is challenging because of the broad heterogeneity of study types published. Journals do have an important role, but it is difficult for journals to make decisions to move the field forward without being too far ahead of the field, she clarified.

Standardizing Nomenclature to Simplify Data Sharing

To improve and ease the burden of data sharing, many speakers also highlighted opportunities for more standardized nomenclature and formatting. Bierer said that data standards have not been well developed, so the data collected are often not interoperable. She said that the Vivli platform could help address this challenge and that expanding its use could enable more work across datasets and investigators. She also noted the lack of attention to attaching metadata to each data element, which obfuscates the level of permission obtained from informed consent. Remedying this would remove the need to go back to the original data or investigator to check each time the data are used. She noted the dearth of data scientists with the necessary skillset for this work. Rubin agreed that if deposited data are appropriately packaged and annotated, many more opportunities exist for access and analysis. Shulman highlighted the advantage of collecting structured data to facilitate extraction and analysis from data reservoirs such as EHRs. Schilsky also discussed mCODE (minimal Common Oncology Data Elements), which aims to promote EHR interoperability by creating a structured data standard for oncology.⁵⁴

Creating Sustainable Public–Private Partnership Models and Developing the Workforce

Several speakers discussed opportunities to improve PPPs by creating sustainable models through enhanced funding, infrastructure, and workforce. Many speakers called for forming an interagency task force coordinated by the U.S. Department of Health and Human Services to create a national clinical trials ecosystem to improve efficiency and effectiveness of U.S. government efforts to address public health needs. Esther Krofah, executive director of FasterCures at the Milken Institute, noted the disjointed and fragmented ecosystem in clinical trials, saying that it needs to be unified to address bigger challenges. Krofah and Schilsky both emphasized the need for this type of deliberative body to help articulate large-scale research policies and have oversight of the clinical trials enterprise in the United States. Krofah referred to a National Academies workshop proceedings, Envisioning a Transformed Clinical Trials Enterprise for 2030 (NASEM, 2022), in which a number of speakers also made this suggestion. She argued that the system has to change to solve all of these challenges, including diversity and equity, while also improving health outcomes.

___________________

⁵⁴ For more information on mCODE, see https://www.hl7.org/fhir/us/mcode (accessed June 21, 2024).

Patient Navigation Support

Herbst and Shulman stressed the need for a sustainable business model to fund patient navigation programs, which are critical to increase participation in clinical trials and provide equitable access to standard therapy. Tendler emphasized the nurse navigator role, pointing out that community sites are typically very underresourced but see many patients who need extra time and explanations that they are unable to get, creating another layer of disparity (Fashoyin-Aje et al., 2023). He shared that Johnson & Johnson works with ACS to support the Navigating Patients Across the Care and Treatment Continuum program,⁵⁵ a formal nurse navigation care program that is collaborating with 20 cancer health systems. It funds nurse navigators and the technology to help them work more efficiently while also focusing on out-of-pocket patient costs for transportation, lodging, and child care. Tendler said that his company has also partnered with Stand Up to Cancer to support the Diversity in Early Development Clinical Trials Research Grants Program,⁵⁶ which aims to increase the accessibility of clinical trials for minoritized and underserved communities. As another example, Nichols said that LLS has 11 nurse navigators who are clinical trial specialists in oncology and staff who go to communities in 31 regions around the country, including community centers and local health agencies, to promote the availability of clinical trials. Getting the word out, especially in rural and underserved communities, needs a boots-on-the-ground approach, she argued.

Beyond nurse navigators, Meropol, Muller, and Shulman noted a need to more broadly build a sustainable clinical trial workforce. Susan Schneider, associate professor emerita of the oncology specialty at Duke University School of Nursing, emphasized the critical role that nurses and trial managers have in ensuring that patient consent is adequately informed and coordinating patient care, and raised concerns about the challenges that impede greater involvement of nurses. Muller suggested engaging professional organizations and schools to develop career paths for nurse navigators, research coordinators, engineers in artificial intelligence, and legal support, noting the need for a larger workforce and career development pathways and specialties in clinical trials. Johnson noted the need for professional ladders for nurses in clinical trials and detailed the steps of career progression, from eligibility assessment to managing complex trials. Robert Winn, director and Lipman Chair in Oncology at Virginia Commonwealth University Comprehensive Cancer Center, added that it will

___________________

⁵⁵ See https://pressroom.cancer.org/Janssen (accessed March 29, 2024).

⁵⁶ See https://standuptocancer.org/research/research-portfolio/research-teams/enhancing-diversity-in-early-phase-clinical-trials-in-an-urban-underserved-community (accessed March 29, 2024).

be necessary to think about the skills needed for those specific roles within PPPs and the training that will be required when investing in this new workforce. Shulman echoed the need for financial support and career incentives for these professionals.

Incentivizing Action and Leveraging Technology

Additional ideas for sustainable models of successful PPPs included incentivizing action and leveraging digital tools. For example, Herbst suggested that health systems could be incentivized to accelerate contract review times and trial launches, reducing negotiation over minor adjustments to language. Schilsky shared an example from his experience setting up the ASCO TAPUR trial, for which his team developed a contract template for all participating sites. He acknowledged that some level of negotiation always occurred, but they got less pushback than expected, and it was a quicker way to sign up institutions than starting from scratch each time. Bierer also referenced the potential for incentivizing data sharing among investigators through modalities such as the open data badge⁵⁷ or considering published datasets and citations based on secondary use of the data to support academic promotion. Lastly, Meropol and Schilsky called for incentives for collaborations to support postmarket studies, recognizing the value to society and public health.

Meropol said that technology can support PPPs by customizing data capture tools within routine workflows such as EHR documentation, increasing data accuracy, and providing the option of remote monitoring of source documents. It can also activate sites that might not otherwise participate in research because of staffing burden by automating data transfer directly from EHRs to study databases. Overall, the availability of digital tools offers a great opportunity to simplify operational burdens for sites and patients, he said.

Winn emphasized the need for integration and alignment among collaborating partners and encouraged all the workshop participants to think about “not just doing something different but doing a different thing.” He proposed training clinical trial “quarterbacks” who have a broader vision for the field and think about new models and new ways of doing things, noting that “we train in silos and then think that we are going to get different results.” As an example, he said “wouldn’t it be wonderful if you had someone who actually understood how to design and implement clinical trials but also was trained on how to use a community asset map?”

___________________

⁵⁷ The Open Science Framework awards an open data badge when datasets used in research are publicly available. For badge criteria, see https://osf.io/g6u5k (accessed March 26, 2024).

In closing, Schilsky reminded the workshop participants that “we all want better evidence to care for our patients, and we have a responsibility to try to generate that. That is the ultimate big picture of public–private partnerships.”

REFERENCES

Kidwell, M. C., L. B. Lazarevic, E. Baranski, T. E. Hardwicke, S. Piechowski, L. S. Falkenberg, C. Kennett, A. Slowik, C. Sonnleitner, C. Hess-Holden, T. M. Errington, S. Fiedler, and B. A. Nosek. 2016. Badges to acknowledge open practices: A simple, low-cost, effective method for increasing transparency. PLoS Biology 14(5):e1002456.

Lu, C. C., X. N. Li, K. Broglio, P. Bycott, Q. Jiang, X. Li, A. McGlothlin, H. Tian, and J. Ye. 2021. Practical considerations and recommendations for master protocol framework: Basket, umbrella and platform trials. Therapeutic Innovation and Regulatory Science 55(6):1145–1154.

Magnuson, A., S. S. Bruinooge, H. Singh, K. D. Wilner, S. Jalal, S. M. Lichtman, P. G. Kluetz, G. H. Lyman, H. D. Klepin, M. E. Fleury, B. Hirsch, A. Melemed, F. I. Arnaldez, U. Basu Roy, C. Schenkel, S. Sherwood, and E. Garrett-Mayer. 2021. Modernizing clinical trial eligibility criteria: Recommendations of the ASCO-Friends of Cancer Research Performance Status Work Group. Clinical Cancer Research 27(9):2424–2429.

Mato, A. R., L. E. Roeker, N. Lamanna, J. N. Allan, L. Leslie, J. M. Pagel, K. Patel, A. Osterborg, D. Wojenski, M. Kamdar, S. F. Huntington, M. S. Davids, J. R. Brown, D. Antic, R. Jacobs, I. E. Ahn, J. Pu, K. M. Isaac, P. M. Barr, C. S. Ujjani, M. B. Geyer, E. Berman, A. D. Zelenetz, N. Malakhov, R. R. Furman, M. Koropsak, N. Bailey, L. Hanson, G. F. Perini, S. Ma, C. E. Ryan, A. Wiestner, C. A. Portell, M. Shadman, E. A. Chong, D. M. Brander, S. Sundaram, A. N. Seddon, E. Seymour, M. Patel, N. Martinez-Calle, T. Munir, R. Walewska, A. Broom, H. Walter, D. El-Sharkawi, H. Parry, M. R. Wilson, P. E. M. Patten, J. A. Hernandez-Rivas, F. Miras, N. Fernandez Escalada, P. Ghione, C. Nabhan, S. Lebowitz, E. Bhavsar, J. Lopez-Jimenez, D. Naya, J. A. Garcia-Marco, S. S. Skanland, R. Cordoba, and T. A. Eyre. 2020. Outcomes of COVID-19 in patients with CLL: A multicenter international experience. Blood 136(10):1134–1143.

Morgan, J., R. J. DeBoer, J. B. Bigirimana, C. Nguyen, D. Ruhangaza, A. Paciorek, F. Mugabo, C. Villaverde, N. Nsabimana, P. Bihizimana, A. Umwizerwa, L. E. Lehmann, L. N. Shulman, and C. Shyirambere. 2022. A ten-year experience of treating chronic myeloid leukemia in rural Rwanda: Outcomes and insights for a changing landscape. JCO Global Oncology 8:e2200131.

NASEM (National Academies of Sciences, Engineering, and Medicine). 2022. Envisioning a transformed clinical trials enterprise for 2030: Proceedings of a workshop. Washington, DC: The National Academies Press. https://doi.org/10.17226/26349.

Patsopoulos, N. A. 2011. A pragmatic view on pragmatic trials. Dialogues in Clinical Neuroscience 13(2):217–224.

Roberts, T. J., A. S. Kesselheim, and J. Avorn. 2023. Variation in use of lung cancer targeted therapies across state Medicaid programs, 2020–2021. JAMA Network Open 6(1):e2252562.

Rottas, M., P. Thadeio, R. Simons, R. Houck, D. Gruben, D. Keller, D. Scholfield, K. Soma, B. Corrigan, A. Schettino, P. J. McCann, III, M. P. Hellio, K. Natarajan, R. Goodwin, J. Sewards, P. Honig, and R. MacKenzie. 2021. Demographic diversity of participants in Pfizer sponsored clinical trials in the United States. Contemporary Clinical Trials 106:106421.

Scher, H. I., M. J. Morris, W. M. Stadler, C. Higano, E. Basch, K. Fizazi, E. S. Antonarakis, T. M. Beer, M. A. Carducci, K. N. Chi, P. G. Corn, J. S. de Bono, R. Dreicer, D. J. George, E. I. Heath, M. Hussain, W. K. Kelly, G. Liu, C. Logothetis, D. Nanus, M. N. Stein, D. E. Rathkopf, S. F. Slovin, C. J. Ryan, O. Sartor, E. J. Small, M. R. Smith, C. N. Sternberg, M. E. Taplin, G. Wilding, P. S. Nelson, L. H. Schwartz, S. Halabi, P. W. Kantoff, A. J. Armstrong, and Prostate Cancer Clinical Trials Working Group. 2016. Trial design and objectives for castration-resistant prostate cancer: Updated recommendations from the Prostate Cancer Clinical Trials Working Group 3. Journal of Clinical Oncology 34(12):1402–1418.

Selig, W., S. Lindsey, M. Bachini, and R. Shroff. 2023. Cross-over arms in oncology clinical trials: An important consideration for patients. Herriman, UT: The Cholangiocarcinoma Foundation.

Shulman, L. N., B. E. Palis, R. McCabe, K. Mallin, A. Loomis, D. Winchester, and D. McKellar. 2018. Survival as a quality metric of cancer care: Use of the National Cancer Data Base to assess hospital performance. Journal of Oncology Practice 14(1):e59–e72.

Wu, D., E. A. Akl, G. H. Guyatt, P. J. Devereaux, R. Brignardello-Petersen, B. Prediger, K. Patel, N. Patel, T. Lu, Y. Zhang, M. Falavigna, N. Santesso, R. A. Mustafa, Q. Zhou, M. Briel, and H. J. Schunemann. 2014. Methodological survey of designed uneven randomization trials (DU-RANDOM): A protocol. Trials 15:33.