7

Envisioning the Future of Pathogen Genomics

The sixth session of the workshop highlighted potential applications of future pathogen genomics systems to prevent and mitigate disease outbreaks. Larry Madoff, medical director of the Bureau of Infectious Disease and Laboratory Sciences for the Massachusetts Department of Public Health, moderated the panel discussion. Chikwe Ihekweazu, deputy executive director at the Health Emergencies Programme of the World Health Organization (WHO), discussed WHO initiatives to advance worldwide collaborative surveillance and pathogen genomics. Michael Worobey, professor and head of the Arizona State University Department of Ecology and Evolutionary Biology, contrasted the timelines for identifying SARS-CoV-2, mpox, and highly pathogenic avian influenza A (H5N1) to emphasize the speed that metagenomic sequencing brings to outbreak detection. Courtney Lias, acting director of the Office of In Vitro Diagnostic Devices at the Food and Drug Administration (FDA), outlined considerations for the development of effective diagnostics and efficient test deployment during crises.¹ Sharon Peacock, master of Churchill College and professor at the University of Cambridge, presented a study demonstrating the value of routine genomic surveillance in clinical practice. Barney Graham, founding director of the David Satcher Global Health Equity Institute and professor at the Morehouse School of Medicine, highlighted pandemic prevention needs and barriers.

___________________

¹ Starting in August 2024, Lias assumed the role of permanent director of the Office of In Vitro Diagnostic Devices at FDA.

ACCELERATING THE USE OF PATHOGEN GENOMICS AND METAGENOMICS IN PUBLIC HEALTH

Ihekweazu outlined WHO’s efforts to advance collaborative surveillance and build pathogen genomics capacity to bolster resilience against disease threats and minimize the effects of pandemic and epidemic threats. Established in 2021, WHO’s Hub for Pandemic and Epidemic Intelligence works to transform collaborative pathogen and disease surveillance by connecting, innovating, and strengthening capabilities to produce better data, analytics, and, ultimately, decisions (WHO Health Emergencies Programme, 2023). Ihekweazu clarified that this program is not a data collection hub but rather an effort to connect opportunities at the pathogen, individual, and population levels to enable improved collective decision making. This program pertains to policy decisions as well as determinations regarding acute emergency response, such as development of countermeasures. The hub works to link data sources, with a focus on increasing coverage and quality of pathogen and disease surveillance. Ihekweazu, who leads the technical advisory panel for the Pandemic Fund, a multilateral financing partnership that provides grants to increase pandemic preparedness in low- and middle-income countries, noted that grant proposals to the Pandemic Fund rarely include efforts to connect the range of existing pathogen and disease surveillance systems, despite the need for collaboration to strengthen these efforts.

Collaborative pathogen and disease surveillance operates across four primary dimensions: diseases threats, sectors (e.g., health, animal and agricultural, environmental), emergency cycles, and disease geographic levels. Ihekweazu explained that countries have different starting points in terms of capacity and collaboration, so supporting efforts to maximize both capacity and collaboration could help optimize decision making and coordinated action. He specified that this collaborative approach surpasses integrated disease surveillance by including capacity of and access to health care and veterinary services, environmental monitoring, and other insights such as community dynamics, behavior, and vulnerabilities (Figure 7-1). The hub approach examines each country’s existing systems, gaps, and opportunities using data sources from the health sector and beyond (e.g., education, industry, transport). Ihekweazu remarked that assessing how sectors and systems communicate informs better decision making and reinvigorates thinking around pathogen and disease surveillance. He added that pathogen genomics also plays a role in this collaborative approach; the hub works to connect genomics data and build requisite capacity across sectors.

Opportunities and Challenges in Advancing Pathogen Genomics

Although the COVID-19 pandemic spotlighted opportunities to apply pathogen genomics in outbreak preparedness, it is not yet optimally deployed, said Ihekweazu. He described the pandemic as providing an important test case that led to the rapid rollout of new pathogen genomics capacity in many regions of the world. Although genomic sequencing technology has developed quickly in recent years, achieving radical reductions in cost and greater volumes of pathogen analysis data, he explained that this amplification has not been consistent across geographies, and there are further opportunities to improve efficiency. Best practices have yet to be fully established in many parts of the world where capacities remain uneven and uncoordinated. Moreover, there are signs of decline in the capacities and knowledge sharing that the COVID-19 pandemic initially amplified. Despite continued innovation and efforts of nations such as South Africa and Nigeria to drive pathogen genomics forward, capacity and data sharing worldwide are still fragmented. Ihekweazu called for collaboration to address the technical, political, and economic challenges hampering both decision making and the ability to rapidly characterize new pathogens and variants, particularly for diseases without existing data-sharing agreements.

WHO Efforts to Advance Pathogen Genomics

Ihekweazu outlined WHO’s efforts to set norms and standards and convene stakeholders to collectively address current genomic surveillance gaps. WHO developed a global genomic surveillance strategy in 2022 with objectives to improve access to tools, strengthen the workforce to deliver speed, scale, and quality, enhance data sharing and utility, maintain readiness for emergencies, and maximize connectivity across countries to the public health surveillance architecture (WHO, 2022). Additionally, WHO convened a new International Pathogen Surveillance Network (IPSN) to work toward the vision of every country having “equitable access to sustained capacity for genomic sequencing and analytics as part of its public health surveillance,” said Ihekweazu (WHO, 2023). This vision does not imply that a country’s public health surveillance system must be within the national public health agency or ministry of health; it could pertain to collaborations between academic partners, national centers or ministries, or a public health agency. The network’s mission is to “engage a mutually supportive global network of pathogen genomic surveillance actors that amplifies and accelerates the work of its members to improve access and equity,” he said. Ihekweazu underscored that the mission relies on inclusive and collaborative member involvement, not on a small group of traditional actors.

As a network of pathogen genomic communities, IPSN will include national and regional public health systems, animal and environmental sectors, policy makers and donors, academic groups, the private sector, civil society, and international standard organizations. Ihekweazu outlined the network’s activities, which include establishing communities of practice to solve common problems, accelerating country scale-up to enable exchange and amplify voices, supporting member projects through catalytic grant funding, advocating for prioritization of pathogen genomics, and convening partners to share progress and innovations. Created less than two years ago, IPSN currently has more than 200 partners from 71 countries. The network is developing a toolkit of IPSN-supported products (e.g., a global investment use case) to bolster advocacy narratives. A country capacity framework will present use cases on scaling to facilitate knowledge sharing as countries plan for capacity scale-up. The toolkit will also include partner-driven products, including frameworks on monitoring and evaluation and on workforce competencies that build shared knowledge about required basic skills and potential resources for outsourcing. Products driven by WHO’s global genomic surveillance strategy will feature global guidelines and a lab-costing tool that countries can use in calculating pathogen-agnostic lab costs, which inform scaling and investment costs. Ihekweazu

emphasized that many of these activities are driven by the community, with WHO curating, coordinating, and supporting efforts.

Current progress toward the five core IPSN activity areas include donor approval for accelerator grants with awardee selection under way, establishment of a functioning data communities of practice for environmental and wastewater surveillance under formation, and development of the country capacity framework, said Ihekweazu. He asserted that accelerating pathogen genomics in public health is contingent upon accelerating collaboration and intersectionality among multiple sectors and partners.

EXPLORING THE POTENTIAL OF PATHOGEN GENOMICS

Presentations during the panel session featured case examples of pathogen genomics applications and their effects on outbreak prevention, detection, and response.

Effects of Technique on Outbreak Detection Timelines

Worobey charted the timeline of using metagenomic sequencing to identify SARS-CoV-2 and contrasted this with mpox and H5N1 detection. In December 2019, a Huanan Market delivery man experienced the onset of symptoms that five days later led to hospitalization (Worobey, 2021). Testing negative for all suspected causes of pneumonia, the man was moved to the intensive care unit on December 22. Two days later the hospital sent a sample from the patient’s lungs to Vision Medicals, a commercial sequencing company. On December 26 Vision Medicals identified SARS-CoV-2, a pathogen closely related to a bat viral isolate and distinct from the SARS-CoV-1 virus identified in 2003, via metagenomic sequencing. By December 27 Vision Medicals had generated a near-complete genome of SARS-CoV-2 and notified the hospital with the results. That same day the doctor who treated the delivery man saw a second case at a different hospital, enabling providers to recognize the first cluster of multiple cases. Worobey underscored that the SARS-CoV-2 genome had already been sequenced on the day that doctors began realizing a new illness was spreading. On January 11 a virologist released the Wuhan-Hu-1 genome, defying a Chinese government gag order on sharing sequences. Worobey noted that the Chinese government announced human-to-human transmission of SARS-CoV-2 and established a cordon sanitaire on January 20, 2020.

In contrast to SARS-CoV-2, an outbreak of mpox in Nigeria was recognized in 2017 but did not capture widespread attention until it reached other countries in 2022, said Worobey. By using gene editing techniques to trace the virus, he and colleagues determined that mpox was likely

transmitted from animals to humans in 2015 (O’Toole et al., 2023). Seven years passed from the time of spillover to when pathogen genomics was leveraged in decision making to prevent outbreak spread. He described how the H5N1 outbreak in cattle occurred in the United States, which has ample investment and capacity to apply advanced techniques. In February 2024 veterinarians collected milk samples after dairy cattle demonstrated reduced yield. For several weeks, researchers conducted testing for approximately 200 viruses and did not confirm H5N1 until March 25. On April 22 the U.S. Department of Agriculture released raw sequence data, which Worobey and colleagues assembled into genomes, revealing that a single outbreak had likely been spreading since November 2023. Worobey highlighted that the distinguishing feature in detection of SARS-CoV-2 was that metagenomic sequencing was used to quickly detect the novel virus. He emphasized that metagenomic sequencing tests all RNA and DNA in a sample, enables immediate detection of an unknown genome, and eliminates the time-intensive process of testing for the presence of individual viruses. Worobey contended that although metagenomic sequencing is costly, it is more powerful than other techniques and should be used in cases of unexplained illness in humans and animals.

Diagnostic Development and Deployment Considerations

Lias highlighted test design considerations in optimizing translation from research environments to clinical use. FDA’s Office of In Vitro Diagnostic Devices regulates clinical laboratory tests, including tests for pathogen detection for clinical point-of-care and home environments. The COVID-19 pandemic highlighted challenges in developing and deploying such tests and opportunities to reach all U.S. populations. She maintained that test developers should consider the context in which tests will be used to increase the speed with which innovative tests reach the market and are adopted. Design inputs included in FDA quality system requirements pertain to how a test will be used, by whom, and in what environment. For instance, tests designed to run on commonly available commercial laboratory platforms are designed differently than tests intended for use by untrained individuals. Lias described that there are constitutional differences in tests that require refrigeration and tests that need to be able to travel in unrefrigerated vehicles. Developers must also account for inevitable trade-offs in specificity and sensitivity and that one feature may be prioritized depending on the intended use. A test designed to screen a high volume of individuals to ensure none of them carries a certain pathogen requires a high level of sensitivity. However, test developers who have not worked in the clinical diagnostic field may be unaware of contextual factors that could reduce test specificity, said Lias. In contrast, other scenarios

might warrant balancing sensitivity and specificity to optimize accuracy. Such considerations should be incorporated into the design of tests for emerging pathogen threats, she asserted. Increased communication between translation and research communities could aid in test design that enables robust validation and regulatory authorization, Lias maintained.

Test design and manufacturing considerations affect the time and cost involved in producing diagnostics at scale, said Lias. For example, adding complexity—such as high-throughput or multiplex design—to a molecular single-target test will increase production time and cost, as well as complicating validation studies. The development of certain tests, such as immunoassays or reagent platforms to test for new pathogens, can be particularly time-intensive, Lias said. After manufacturing, a test must obtain regulatory authorization before it can be sold to laboratories, through which FDA assures that tests are appropriately designed, manufactured, validated, and maintained to be clinically and analytically valid. Lias emphasized that laboratories, health care providers, and patients should understand the capabilities and limitations of any diagnostic.

Lias noted that many necessary decisions were made in a rapidly changing landscape at the beginning of the COVID-19 pandemic regarding test development and deployment, and that addressing current gaps could enable a more effective diagnostic response in future crises. For example, R&D of technologies for pathogen-agnostic testing could expedite the identification of emerging threats. Prepositioning of pathogen detection tests could enable rapid deployment and scaling. From a regulatory perspective, this would entail early discussion between FDA and test developers about pre-review of validation data for tests and platforms for rapid deployment. During the pandemic, uncoordinated test development involved hundreds of companies and labs competing for raw materials, equipment, and customers, generating shortages and confusion, said Lias. Regulatory authorization of well-validated tests with high throughput, high manufacturing capacity, and compatibility with processing instruments with a large installed base could accelerate test availability. Agile payment and reimbursement strategies that facilitate timely patient access to accurate diagnostic testing early in a pandemic are also crucial considerations, Lias added.

Genomic Surveillance for Diseases in Routine Infection Prevention and Control

Peacock underscored the value of accessible genomics for local need through its use in routine infection prevention and control (IPC) practice. She highlighted that 1 in 31 U.S. patients contracts one or more infections while receiving health care (CDC, 2024c). Many of these infections are caused by multidrug-resistant organisms (MDRO) spread between patients

through intermediary contact—often by way of health care workers—or within the environment that sometimes result in outbreaks. In the United Kingdom, the current process for hospital outbreak detection involves reviewing lists of patients with MDROs and, in situations in which two cases test positive for the same MDRO, searching for any overlaps in time and space between patients. Subsequent investigation by ward visit and initial assessment is followed by bacterial typing. Peacock emphasized that this process is time-consuming, suggesting that it should be replaced by routine sequencing of all nosocomial pathogens sent to the lab. Outbreak probability could be determined based on the degree of genetic relatedness between pairs of isolates and the temporal-spatial analysis performed using machine learning, Peacock added.

Peacock maintained that this paradigm shift toward routine sequencing would improve the accuracy of outbreak investigation, improve clinical decision making, save money, and possibly grow the economy by enabling faster return to work for patients and reduced hospital costs. She described her research in 2018–2019 to determine whether routine sequencing of methicillin-resistant Staphylococcus aureus (MRSA) in a hospital altered IPC decision making (Blane et al., 2024). The study involved identifying all MRSA-positive patients admitted to the hospital or seen in the hospital clinic or emergency department, then sequencing samples from more than 800 cases. Peacock and colleagues conducted geospatial analysis for each case pair and met with the hospital IPC team to discuss findings and record actions. They identified 72 genetic clusters in the isolates, with 41 percent of all cases linked to a cluster. Geospatial analysis determined that geospatial overlaps occurred in 34 of the 72 clusters. Over the course of the study, they found that 17 of the clusters were hospital outbreaks, leading to IPC actions for 14 of the clusters. They also identified 38 pseudo-outbreaks—groupings that appeared to be clusters but were not—and agreed to de-escalate further IPC investigation of six of these groups. Moreover, they identified two clusters driven by transmission in the community, one occurring in a care home referring patients to the hospital and the other among a group of individuals using intravenous drugs and seeking treatment for bacteremia. The teams created feedback and action loops with the goal of reducing the number of MRSA cases occurring in both settings, Peacock noted.

In this study, sequencing MRSA led to changes in IPC decision making in more than 80 percent of confirmed MRSA outbreaks in hospitals, enabled rational de-escalation to pseudo-outbreaks, and facilitated identification of clinically important but unsuspected community outbreaks, said Peacock. Replicating these benefits via routine sequencing in the clinic would entail a shift in the public and political narrative, so she suggested that advocates begin with the anticipated effects of routine sequencing in the laboratory and make the case for this shift in clinical practice. Addi-

tionally, case studies for payers would generate a cost-benefit analysis and characterization of cost-effectiveness. Peacock emphasized the importance of funding and noted that FDA approval of genomics as a diagnostic or clinically validated test could advance change.

Pandemic Prevention Needs and Barriers

Graham outlined needs and barriers in improving response to pandemic threats. He described how pandemic threats almost always begin as regional problems, noting that emerging infections often originate in low-income countries but are detected in high-income countries. Given that early recognition and response with medical countermeasures is the main determinant of outcome, he called for diversifying and expanding the public health workforce. He outlined other pandemic prevention needs, including a One Health approach to surveillance of people, animals, and environment; a comprehensive prototype pathogen approach for developing medical countermeasures; and increased distribution of manufacturing and research capacity. Graham asserted that in the absence of equitable capacity distribution and improved systems for deployment and implementation, regional problems will not be addressed until they become pandemic threats. Noting the cost savings associated with multiplex testing, he commented on the need for multipurpose technologies—such as disease surveillance technologies repurposed to research and discovery, as well as clinical use applications—that could be deployed in low-income countries to serve multiple purposes. However, despite practical, medical, and moral imperatives to act, practices and policies have not kept pace with the technological advances to address pandemic threats, he added. Graham emphasized the need to increase distribution of research and manufacturing capability across both high- and low-income countries, by persuading policymakers and funders of the potential benefits.

DISCUSSION

Future Pathogen Genomics Applications

Madoff asked panelists to predict how pathogen genomics will advance in coming years. Worobey replied that he is working to develop rapid assays with the sensitivity of polymerase chain reaction (PCR) tests that return results within 10 minutes but do not require sample processing, a technician, or a lab. Ihekweazu remarked that sample exchange is often dependent on informal social professional networks, suggesting that professional institutional relationships could serve as mechanisms for sample and knowledge exchange. Lias described her vision of “plug-and-play” diagnostic platforms

that are designed for multiple purposes and can be quickly customized as specific needs arise, noting that demand for point-of-care diagnostics is increasing. Highlighting the availability of a point-of-care molecular test for hepatitis C virus that generates results within an hour, Lias emphasized the value of such tests for future infection diagnosis, treatment, and control. Peacock pointed to the development of sequencing capability for foodborne pathogens, tuberculosis, HIV, hepatitis, and other diseases within the past decade; she was optimistic that the next decade will see full integration of sequencing into diagnostic laboratories. At an international level, this effort will require economists to develop business models that make a strong case for sufficient funding, she added. Graham suggested that sequencing technology should be implemented around the equator—where biodiversity is highest and countries tend to be low-income countries—with support from a robust in-house workforce of bioinformaticians. Moreover, true partnerships between high- and low-income countries should extend beyond sample and data sharing and recognize the human capital, ambition, and energy in the workforces in low-income countries, said Graham.

Development and Use of Multipurpose Technologies and Diagnostics

An attendee commented on the value of multipurpose technologies and asked how to address barriers to the development and use of such innovations. Lias remarked that commercial laboratories and test developers make market-driven decisions that may be incongruent with public health perspectives. Christopher Braden, deputy director of the Centers for Disease Control and Prevention (CDC) National Center for Emerging and Zoonotic Infectious Diseases, weighed in on the need to incentivize the development of diagnostics that generate rapid results, can be used in proximity to the patient to mitigate investigation lag time, and produce the genomic data needed to support public health surveillance, detection, and response. Jill Taylor, senior advisor for scientific affairs at the Association of Public Health Laboratories, noted the challenge that reimbursement structures pose to use of multiplex panels and asked how to advocate for Centers for Medicare & Medicaid Services (CMS) coverage of such diagnostics. Lias stated that engaging experts, including from CMS, on the regulations governing CMS coverage decisions is helpful in determining how best to approach the issue. Worobey added that multiplex tests, which fall on a spectrum between pathogen-specific single-plex tests and pathogen-agnostic metagenomics, are likely to play a larger role in identifying pathogens causing respiratory infections, sepsis, and other illnesses.

Another attendee pointed to the importance of rapid test results and asked how to accelerate data delivery to inform response. Peacock replied that industrial partnerships could contribute to developing technical solu-

tions capable of receiving a sample, sequencing it, and issuing a result within 24 hours. Noting that matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) is rapid and inexpensive and has revolutionized bacterial identification, Peacock suggested issuing a challenge to industry to develop an equivalent technology for sequencing. Lias remarked that industry will respond to market forces and sufficient demand. Worobey stated that incentives for academia to undertake “moonshot” projects can generate innovation—such as the real-time SARS-CoV-2 sequencing efforts—from high- and low-income countries.

Value of Prevention

Worobey remarked that pandemics will continue emerging in the future and carry multi-trillion-dollar costs, risk millions of lives, and have substantial effects on agriculture; thus, prevention efforts merit orders of magnitude more investment. For instance, bioinformatics and sequencing teams could be established to detect pathogens and react in real time. To shift the perspectives of decision makers toward prevention, Graham suggested using metrics other than dollars to measure the value of public health and the extent of the threat posed by pandemics.

Adoption of Routine Genomic Surveillance in Clinical Settings

Alli Black asked Peacock about the effects of the MRSA study on clinical and public health practice. Peacock replied that although some European hospitals have adopted a sequence-first approach to MDROs, the rollout of routine sequencing at the study site was disrupted by the COVID-19 pandemic. Her subsequent research pivot to SARS-CoV-2 resulted in delayed publishing of the MDRO study until 2024, so it may be too early to observe real-world outcomes from the study’s findings. In the United Kingdom, the adoption of routine sequencing by the National Health Service would be most impactful, she stated, given that National Health Service laboratories are regional and public health laboratories are more centralized. Peacock suggested that amplifying messaging about how routine sequencing can improve patient care, prevent unnecessary deaths, and halt outbreaks—all of which can decrease hospital use and length of stay, thereby reducing cost—could help make the case for adopting routine sequencing in the National Health Service.

Data Collection and Storage Considerations

A participant asked how single-plex, multiplex, and agnostic testing methods could be integrated into existing surveillance systems. Addition-

ally, he remarked on the technical logistics of sharing petabytes of data and the need for creative approaches to data compression, transfer, and analysis that could advance metagenomic surveillance applications. Lias replied that the Science and Health Initiative to Combat Infectious and Emerging Life-Threatening Diseases is working to standardize diagnostic data coding to enable database linking and further analysis of real-world data. Differences in data collection and entry frequently pose challenges to collating datasets, she said. Worobey added that the quality of metadata—such as documentation of tissue type, time, and location of sample collection—affects the utility of sequencing data. Peacock noted that in some cases, data may not require or meet minimal standards of usefulness to warrant storage, such as sequencing data that lack necessary metadata. Moreover, in cases where a validated test with adequate sensitivity, specificity, and predictive values meets the need at hand, genomic data may not provide any additional value.

Considerations for Data Sharing

Ihekweazu underscored the difficulty in achieving consensus on a pathogen access and benefits sharing system, suggesting that proponents should elevate the principle of solidarity and highlight the benefits of pathogen data sharing. Graham highlighted data-sharing gaps within the United States as evidenced by the delays in detecting the H5N1 outbreak in cattle. A participant noted that the COVID-19 pandemic amplified issues of data accessibility and asked how WHO is approaching this disconnect and bureaucratic challenges associated with data sharing. Ihekweazu replied that surmounting barriers to data sharing will require addressing components including governance, data platforms, and knowledge building. He added that establishing a global data governance arrangement could facilitate broader data sharing. Additionally, efforts to build knowledge capability—including basic bioinformatics knowledge—that were catalyzed by the COVID-19 pandemic should continue between crises. Ideally, integrating the various components of data sharing should coalesce into a unified framework that advances pandemic preparedness, said Ihekweazu.