Preparing for Nontraditional Biothreats

Abstract

Biological threats consist of traditional, emerging, enhanced, and advanced threats, but current biodefense approaches focus almost entirely on a subset of traditional threats. There is an urgent need to develop a comprehensive, rational, and systematic plan to address and mitigate the broader risk space. Myriad possible strategies exist, but an ideal strategy will extend beyond a list of agents. This article proposes a functionality-based approach based on systematic identification of key functional elements, essentially focusing on mechanisms of what constitutes a threat: The key threat element is addressed directly instead of extensive characterization of ancillary details. Examples might include a potent toxin, long-term environmental stability, or a specific protein causing morbidity/mortality. By identifying the critical components leading to disease, limited time, efforts, and resources can be focused to address the greatest risks. Further, as future threats will likely contain critical aspects of known agents, this approach will potentially address a large area of uncharacterized risk space. Thus, focused research can buy down a large area of risk space while still addressing traditional threats and mission needs. Application of this strategy will move the field away from agent-based lists toward a more comprehensive hazard analysis and will position biodefense and health communities to prepare for the threats of the future.

Biological threats consist of traditional, emerging, enhanced, and advanced threats, but current biodefense approaches focus almost entirely on a subset of traditional threats. There is an urgent need to develop a comprehensive, rational, and systematic plan to address and mitigate the broader risk space, and an ideal strategy will extend beyond a list of agents. This article proposes a functionality-based approach based on systematic identification of key functional elements, essentially focusing on mechanisms of what constitutes a threat.

Many US government departments and agencies are actively engaged in biodefense, developing and implementing plans to protect against biological threats. Government efforts are directed by Homeland Security Presidential Directives (HSPD) 10 and 18, which describe US plans for biodefense in the 21st century and the development of medical countermeasures against weapons of mass destruction.^1,2 These directives are carried out by various government departments and agencies, with efforts historically focusing on a list of traditional agents such as Bacillus anthracis (anthrax), Variola major virus (smallpox), and Yersinia pestis (plague). For example, the US Department of Homeland Security (DHS) focuses on a list of agents that were selected by scientific and national security subject matter experts based on a variety of inputs including information from the historic US offensive weapons program and current scientific understanding of these agents.^3,4 Other government departments and agencies have their own lists, such as the Federal Select Agent Program select agents and toxins list, and these lists of agents are used to guide biodefense investments according to each department or agency's specific missions and responsibilities.^5-8 While this paradigm works well for known threat agents, such an approach gives minimal attention to the totality of the biological risk space. In fact, it arguably increases susceptibility to attack with anything not on “the list” by ignoring or underrepresenting all other threats.

In addition to agents historically included on the lists, HSPD-10 and HSPD-18 also champion the importance of “preventing and controlling future biological weapons threats.”¹ They describe 4 categories of future biological threats: traditional agents, enhanced agents, emerging agents, and advanced agents (see Figure 1 for definitions). Underrepresenting the additional risk space beyond traditional agents could facilitate technological surprise, while overrepresentation diverts precious biodefense funding away from current priorities. Thus, while the US government recognizes the threat posed by these nontraditional biothreats (ie, anything not on the current lists), “a comprehensive and integrated approach is needed to prevent the full spectrum of biological threats.”⁹

Figure 1.

Categories of Future Biological Threats, as defined in HSPD-18²

Despite a clear need, a comprehensive plan to address future threats is lacking for a number of reasons, including (1) the immense breadth of possible traditional, emerging, enhanced, or advanced biological threats; (2) the difficulty of predicting and preparing for future risks; (3) limited resources to prepare for, prevent, and respond to all possible hazards; (4) the rapidly developing technological landscape underpinning the creation of enhanced and advanced threats; and (5) development of a comprehensive plan falls outside the mission of any individual agency. Developing such a plan requires bounding a vast and expanding possibility space. Resources simply do not exist to address all possible threats (even if all threats could be identified). Therefore, any successful strategy must identify and address the greatest risks. If we define risk as the product of probability and consequence, then risk space can be graphically visualized with probability on one axis and consequence on the other (Figure 2). Probability includes a number of factors about the biological hazard, including likelihood of acquisition (can it be isolated from the environment, bought online, or synthetically created?), ease of use (are there significant scientific or technical hurdles to production or weaponization?), and capabilities of the adversary (do they have scientific expertise, access to equipment, or sufficient funding?). Similarly, consequences span a wide range across morbidity and mortality but can also include economic effects or social destabilization. Within the vast possibility space represented on the graph, bounds must be set to ensure prevention, detection, response, and recovery efforts are commensurate with risk.

Figure 2.

Risk Space. The risk space for a bioterrorism attack is the product of probability (y-axis) and consequence (x-axis). Nominal data for 5 different agents (A-E) are graphed to represent the risk of multiple scenarios (ie, enhancements or advancements), where each colored dot represents the risk of 1 scenario and the point cloud allows visualization of a range of possible hazards. Color images available online at www.liebertpub.com/hs

Strategies to Address Biological Hazards

Multiple strategies were considered to support a structured framework and approach to addressing biological hazards in a rational and effective manner. Each strategy was evaluated for application to traditional threats (such as those on the lists), as well as for effectiveness against enhanced, advanced, and emerging threats. Current strategies are not mutually exclusive; in fact, an ideal approach will incorporate elements of multiple strategies. Evaluated strategies include the agent-based approach, the capabilities-based approach, the intelligence-based approach, and futures prediction approach.

Agent-Based Approach

As described above, a list of agents of concern is compiled, a ranking system is devised and implemented, and biodefense research is prioritized according to this list. As these agents represent traditional threats, this approach is likely to be highly successful if an adversary chooses an agent on the list and uses it in a manner anticipated by biodefense planners. However, the agent-based list does not include alternative agents or scenarios, nor does it address any aspect of emerging, enhanced, or advanced threats unless they are specifically included on the list.

Capabilities-Based Approach

This approach examines current and near-term US capabilities in addressing threats, as well as the capabilities of potential adversaries (effectively assigning priority based on intelligence covering adversary resources and abilities). Investments are then prioritized by what can be accomplished with existing capabilities. This strategy would likely be very effective in the short term, with increased results and lower costs. However, it is somewhat shortsighted, sacrificing mid- and long-term strategy and investments in future capabilities for quick wins.

Intelligence-Based Approach

The government intelligence community tracks biological weapons of mass destruction issues worldwide and communicates that information to relevant stakeholders. Current intelligence gathering depends at least partially on agent-based lists and capabilities approaches, where the agents and specific technologies or capabilities serve as triggers for what information should be collected. Intelligence-based approaches address current state and non-state actor issues, but the information generally has limited insight into future efforts. Moreover, intelligence can change rapidly, decreasing its utility in long-range strategic decisions such as development of new medical countermeasures or the size and composition of the Strategic National Stockpile.¹⁰ Therefore, while intelligence plays an important part in biodefense, it cannot be the sole driver of a comprehensive strategy.

Futures Prediction Approach

There are numerous business forecasting models, as well as popular approaches such as social media monitoring, technology forecasting, or tracking venture capital or grant funding to predict where biology will go in the future. Significant investments are being made by the US government in futures prediction (such as the IARPA ForeST project¹¹), but these capabilities have yet to be fully developed. While these are potentially very valuable approaches, they all inherently rely on being able to predict the future, and predictions are often wrong.

Functionality-Based Approach

Given the complexity of biodefense, an effective strategy must address both traditional and nontraditional biological threats using the strengths from multiple approaches to mitigate each individual approach's weakness. I propose a strategy to understand the key critical aspects that make an agent a threat in order to provide timely insight on today's threats (the current agent-based list and current capabilities), while also providing a foundation for a more thorough understanding of tomorrow's issues (developing capabilities, evolving intelligence, and futures prediction). Many agents of concern have one or a limited number of key critical components that make them hazards, such as environmental stability (B. anthracis), low infectious dose (F. tularensis), or extreme toxicity (botulinum toxin). Biodefense research can be focused on individual components of an agent as knowledge gaps instead of agents as a whole, rather than extensive investments in ancillary aspects that are irrelevant if critical components have been neutralized.

HIV is a good example (even though it is not a traditional bioterrorism threat included on the list), as it has been extensively researched, and detection, prevention, and response options have vastly improved since the virus was discovered. The virus has 2 critical aspects that allow it to cause AIDS: immunosuppression and integration into the host genome. Immunosuppression can be accomplished in multiple ways, including apoptosis of bystander cells, viral killing of infected cells, and the cytotoxic T-cell response. Regulation of each of these pathways is complex, and blocking one pathway is still not sufficient to prevent immunosuppression. Targeting this critical aspect, then, would not provide a simple and direct way to address HIV risk. However, HIV integration into the host genome is accomplished largely through 2 viral proteins acting sequentially: reverse transcriptase and HIV integrase. Inhibition of either protein will prevent viral integration into the host genome, which will also block virus persistence, production, and spread, and will thereby decrease immunosuppression. Integration therefore represents a key critical aspect. Scientists have been able to target integrase and reverse transcriptase and effectively develop drugs against HIV and AIDS, and these drugs are used clinically in antiretroviral therapies.^12,13 Downstream effects and processes (including virion production and yield) become less relevant, since the infection pathway does not progress past initial infection. As such, investments characterizing downstream processes might be a lower priority, since HIV can be neutralized by addressing reverse transcription and integration.

Another example is measles. The critical component of measles is the infectivity of the virus, arising from the viral hemagglutinin (MVH) and fusion (MVF) glycoproteins.¹⁴ The highly infectious nature of measles leads to rampant spread even though the virus does not have a broad host range, is not highly lethal, and can be effectively prevented with vaccination. While the measles vaccine elicits a broad immune response, researchers have demonstrated that most of the effect of vaccination is due to antibody recognition of MVH and MVF.¹⁵ Therefore, the human immune system effectively identifies host cell attachment via MVH as a key critical component and produces antibodies that block infection. Targeting MVH, either experimentally or via the immune system, prevents initial attachment to host cells, which blocks entry, infectivity, and all downstream effects. From a biodefense perspective, anything downstream of infectivity becomes a lower priority since it is prevented from happening. For both HIV and measles, understanding the key critical components of these viral agents allows for the mitigation of disease through investments in a single aspect rather than addressing all possible aspects of an agent or disease.

Addressing a critical component of a bioterrorism threat agent allows the focusing of limited resources on the most affected areas and maximizes the value of every dollar invested. However, the real value of this functionality-based approach is that it covers an entire area of risk space. Targeting MVH to address measles encompasses not only the “traditional” threat but potentially any emerging, enhanced, or advanced version of measles. The rationale is that the key critical component of an agent is already highly evolved for a specific task (such as attachment and entry). Indeed, mutations leading to vaccine escape (ie, enhanced or emerging agents) generally occur outside the MVH receptor binding site and are thus only peripheral to the critical component of viral entry.¹⁶ It is unlikely that beneficial gain-of-function mutations will occur or be engineered into these critical components, as evidenced by the low degree of mutation in the MVH receptor binding site. Most mutations or engineering, conversely, would weaken the agent, making it less of a threat.^15,17,18 Any enhancements would likely be in other aspects, such as making the agent more virulent or more stable. In the case of measles, these enhancements would still be effectively prevented by targeting MVH, so the key critical component approach addressing traditional MVH would also work for enhanced and emerging risks.

Advanced risks, on the other hand, are more difficult to predict, as these are novel pathogens or materials. The likelihood of success for an adversary developing a completely novel pathogen is incredibly small (and therefore low risk); it is much more likely that the successful development of a novel threat would be based on the critical components of existing agents. For instance, replacing the entry proteins of nearly any virus (even VSV, which is the prototypical pseudotyping entry system) with MVH results in a more infectious virus.¹⁹ Thus, creating a new agent using MVH for attachment would be highly effective and technically tractable. MVH infectivity is largely independent of virion content, and entry is effective for artificial delivery systems such as pseudotyped liposomes, where nonagent threats or bioregulators can be delivered.^20-22 However, regardless of the virus or payload, research addressing MVH will be applicable to any agent or system employing MVH for entry and will prevent infection and downstream consequences, protecting against traditional, enhanced, emerging, and advanced bioterrorism risks. While measles is used here as an academic example, the same concept of key critical components being utilized or combined holds true for the agents on the current biothreat list, as well as a plethora of other threats.

Essentially, the functionality approach addresses one aspect of identified threats (the key critical component) to buy down a whole class of hazards (those scenarios that leverage the same or similar key components). At the very least, the approach can be used for agents currently on the list and will focus limited resources on those aspects of known high-risk agents. The ultimate goal, however, is not just to focus on known agents, but rather to address the greatest threats independent of agent-based lists. These threats may include: What is the most toxic? What is the most contagious? or What is the easiest to obtain? Such a shift in perspective will focus on threats rather than simply threat agents and will further allow mechanistic and analytical studies of why something is a threat instead of more phenomenological observations of predefined threat agents. In the long term, this approach can lead to predictive models of broad threat categories anticipating (and preparing for) both known and unknown threats.

Implementing the Functionality-Based Approach

In order to validate the effectiveness of a functionality approach, proof-of-concept experiments need to be performed on current threat agents. Doing so will prompt a close examination of the key critical components of current threat agents and promote discussions about why these agents are prioritized, what makes an agent a bioterrorism concern, what constitutes key critical aspects of each agent, and how those aspects might be combined. If the first phase is successful, a second phase can expand beyond traditional agents to encompass emerging, enhanced, and advanced threats. Such an approach will address more risk with the same resources and potentially be safer for researchers by allowing work to be performed on portions or aspects of an agent instead of fully infectious versions (eg, isolated viral attachment proteins incapable of causing disease rather than complete infectious viruses).

As with all approaches discussed here, the functionality approach is not perfect. It cannot address all possible risk, just as it cannot predict the future. It is possible and even likely that a potential adversary will choose something that was not prioritized or develop something entirely new. Biodefense, however, is not about predicting the future or about negating all possibilities. Rather, it is about mitigating as much risk as possible to be as prepared as we can be with the resources we have. By applying a functionality-based approach, the vast enhanced and advanced risk space can be reasonably collapsed to those critical mechanistic properties that make an agent a threat. More important, each critical component addressed covers an expanded threat space encompassing potential emerging, enhanced, and advanced threats and does so without having to know exactly what those threats are. Not only does this approach move biodefense closer to preventing the threats of today, but it simultaneously addresses the threats of tomorrow. The implementation of this functionality-based approach will support a government-wide effort toward meeting the mandates of HSPD-10 and HSPD-18, where engagement between government departments and agencies and the research community will accomplish the shared goal “to advance the health security of all people.”⁹

Acknowledgments

The author thanks Matthew Moe and Lloyd Hough for ideas, discussions, and critical feedback on the manuscript. He also thanks the Emerging Leaders in Biosecurity Initiative (ELBI) staff and fellows for feedback and advice. The author's AAAS Science and Technology Policy fellowship is administered by the Oak Ridge Institute for Science and Education (ORISE). The author is also an ELBI fellow, administered by the UPMC Center for Health Security. This manuscript and the ideas contained herein are the opinions of the author and do not necessarily reflect the policies and views of AAAS, ORISE, ELBI, or any other organization with which the author is affiliated.